Search tips
Search criteria 


Logo of jexpmedHomeThe Rockefeller University PressThis articleEditorsContactInstructions for AuthorsThis issue
J Exp Med. 2010 June 7; 207(6): 1173–1182.
PMCID: PMC2882827

Heterogeneity and hierarchy within the most primitive hematopoietic stem cell compartment


Hematopoietic stem cells (HSCs) have been extensively characterized based on functional definitions determined by experimental transplantation into lethally irradiated mice. In mice, HSCs are heterogeneous with regard to self-renewal potential, in vitro colony-forming activity, and in vivo behavior. We attempted prospective isolation of HSC subsets with distinct properties among CD34−/low c-Kit+Sca-1+Lin (CD34KSL) cells. CD34KSL cells were divided, based on CD150 expression, into three fractions: CD150high, CD150med, and CD150neg cells. Compared with the other two fractions, CD150high cells were significantly enriched in HSCs, with great self-renewal potential. In vitro colony assays revealed that decreased expression of CD150 was associated with reduced erythroblast/megakaryocyte differentiation potential. All three fractions were regenerated only from CD150high cells in recipient mice. Using single-cell transplantation studies, we found that a fraction of CD150high cells displayed latent and barely detectable myeloid engraftment in primary-recipient mice but progressive and multilineage reconstitution in secondary-recipient mice. These findings highlight the complexity and hierarchy of reconstitution capability, even among HSCs in the most primitive compartment.

Long-term multilineage reconstitution of lethally irradiated mouse bone marrow has been used as a gold standard to define the functionality of hematopoietic stem cells (HSCs). Based on this, HSCs have been studied extensively, yielding great contributions to the field of stem cell biology. The development of strategies for prospective isolation of HSCs has made in vivo clonal analysis of HSCs possible over the last decade. Analyses of individual purified HSCs or limiting doses of whole bone marrow transplanted into irradiated animals have revealed marked functional heterogeneity in HSCs with regard to repopulating activity, self-renewal activity, and in vitro colony-forming activity (Abkowitz et al., 2000; Müller-Sieburg et al., 2002, 2004; Uchida et al., 2003; Takano et al., 2004; Ema et al., 2005, 2006; Sieburg et al., 2006; Dykstra et al., 2007). Different patterns in lineage reconstitution by individual HSCs have also been observed (Uchida et al., 2003; Müller-Sieburg et al., 2004; Sieburg et al., 2006; Dykstra et al., 2007). Of interest is that the donor-derived myeloid/lymphoid ratio in reconstituted mice reportedly indicates the degree of self-renewal potential in transplanted HSCs (Müller-Sieburg et al., 2004; Dykstra et al., 2007). If a variety of HSCs exist, HSCs may exhibit and define hierarchical organization within the most primitive hematopoietic compartment. Alternatively, heterogeneity of HSCs may be generated during development of the hematopoietic system and remain fixed thereafter. To further address questions of functional diversity and hierarchy in HSCs, HSC subsets with distinct properties must first be isolated prospectively.

CD34−/low c-Kit+Sca-1+Lin (CD34KSL) cells in mouse bone marrow are highly enriched in adult HSCs (Osawa et al., 1996; Sudo et al., 2000; Matsubara et al., 2005; Morita et al., 2006). To identify candidate cell-surface markers that could prospectively identify functionally distinct HSCs, we screened a large number of antibodies and identified those with heterogeneous staining patterns on CD34KSL cells. These candidates were then tested for functional differences in vivo. In this way, we found that expression of CD150 (Kiel et al., 2005) could be used to enrich for long-term repopulating cells (LTRCs) with distinct reconstitution kinetics patterns. CD34KSL cells were subdivided into CD150high, CD150med, and CD150neg fractions, and the functions of these cells were compared at the clonal level using single-cell transplantation and cultures.

While performing this study, we noticed the existence of a very rare type of LTRCs. These cells were named “latent HSCs” and were operationally defined as cells that exhibited significant engraftment only after 12 wk or more in the primary recipient yet showed multilineage reconstitution in the secondary recipient. Given that these cells could fully reconstitute the secondary recipients, they are assumed to have high reconstitution potential. However, they do not satisfy (at the level of primary transplantation) some of the criteria that currently define HSCs.

Our results indicate that differences in reconstitution activity are caused by intrinsic differences among cells and that classes of HSCs with distinct in vivo behavior can be prospectively isolated. We also report the discovery of a very slowly engrafting, myeloid-biased HSC whose full potential can be seen only after secondary transplantation. The existence of latent HSCs requires reconsideration of the concept of HSCs and leads us to propose a revision of the current criteria for HSCs.


Fractions in CD34KSL cells

As in previous studies (Abkowitz et al., 2000; Müller-Sieburg et al., 2002, 2004; Uchida et al., 2003; Sieburg et al., 2006; Dykstra et al., 2007), our single-cell transplantation experiments yielded data showing that CD34KSL cells display heterogeneous reconstitution patterns in primary- and secondary-recipient mice (Ema et al., 2005). For instance, there are HSCs that sequentially reconstitute myeloid, B lymphoid, and T lymphoid lineages in primary-recipient mice, followed by multilineage reconstitution that is maintained in secondary-recipient mice; HSCs that predominantly reconstitute myeloid lineage in primary-recipient mice, followed by greater levels of multilineage reconstitution in secondary-recipient mice; and HSCs that predominantly reconstitute lymphoid lineage in primary-recipient mice, followed by markedly reduced levels of reconstitution in secondary-recipient mice. To achieve greater purification of HSCs, and to identify cell-surface markers that distinguish HSCs with heterogeneous reconstitution behaviors, CD34KSL cells were costained with a comprehensive list of additional antibodies (Table S1) and analyzed by flow cytometry (Fig. 1; and Figs. S1 and S2).

Figure 1.
Markers with heterogeneous expression in CD34KSL cells. KSL cells were stained with FITC-conjugated anti-CD34 antibody and additional PE-conjugated antibodies as shown. Shown are flow cytometric profiles for the markers that had heterogeneous ...

In this study, 118 cell-surface markers listed in Table S1 were examined. Among them, 81 markers appeared not to be expressed on CD34KSL cells. Most of these were not expressed on CD34+KSL cells either, with some notable exceptions, as shown in Fig. S1. Markers with heterogeneous expression in CD34+KSL cells include CD48 and CD244, members of the Slam family recognized as negative markers for HSCs (Kiel et al., 2005), and CD135, used to isolate lymphoid-primed multipotent progenitors (LMPPs; Adolfsson et al., 2001, 2005). These markers, as well as CD11a (LFA1), CD62L (L-selectin), and CD138 (Syndecan1), can be used for better discrimination of CD34 from CD34+ cells within the KSL population. Whether N-cadherin is expressed on HSCs remains somewhat controversial (Zhang et al., 2003; Kiel et al., 2007, 2008; Haug et al., 2008). In our analysis, CD34KSL cells did not express N-cadherin, as additionally demonstrated in Fig. S1.

On the other hand, 22 markers were expressed on CD34KSL cells, as depicted in Fig. S2. Although most such markers were expressed on both CD34 and CD34+KSL cells, expression of CD27, CD201, and c-Mpl seemed down-regulated on some CD34+KSL cells. This is consistent with earlier reports in which CD201 and c-Mpl were used as positive markers for HSC purification (Balazs et al., 2006; Yoshihara et al., 2007). Only 15 markers showed variable expression on CD34KSL cells. We were particularly interested in this group of markers because their use permitted us to test fractions of CD34KSL cells to determine whether they might represent functionally distinct subsets of HSCs. These candidates were AA4.1, CD1d, CD11b, CD18, CD31, CD38, CD49b, CD51, CD61, CD86, CD103, CD147, CD150, FcγR, and Tie2 (Fig. 1).

Based on expression of these antigens, CD34KSL cells were sorted into positive and negative fractions by flow cytometry. In the case of CD38, CD34KSL cells were separated into CD38high and CD38med fractions, which amounted to 22.4 ± 3.8% and 56.1 ± 5% of CD34KSL cells, respectively (n = 5; mean ± SD); the CD38neg fraction was barely detectable. In the case of CD150, CD34KSL cells were separated into CD150high, CD150med, and CD150neg fractions (Fig. 1). These represented 26 ± 5.6%, 24 ± 2.9%, and 14.6 ± 3.6% of CD34KSL cells, respectively (n = 8; mean ± SD); CD150 expression levels varied widely among CD34KSL cells.

We next used transplantation to compare the in vivo function of fractions of CD34KSL cells. Each of a group of lethally irradiated mice received 10 cells from these individual fractions along with 2 × 105 competitor cells. Recipient mice were analyzed between 16 and 20 wk after transplantation. As shown in Fig. 2, long-term (16 wk or more) repopulating activity was detected in both the positive and negative fractions in all cases, suggesting that none of these markers is particularly useful for excluding non-LTRCs from CD34KSL cells. However, we found significant differences in the percentage of chimerism among fractions separated on the basis of expression of CD38, CD147, CD150, and Tie2 (Fig. 2).

Figure 2.
Long-term reconstitution by fractions of CD34KSL cells. CD34KSL cells were subdivided into fractions positive and negative for additional markers. 10 cells from each fraction were transplanted into each member of a group of lethally ...

We observed particularly interesting reconstitution kinetics patterns in the cases of CD150 and CD38. CD150highCD34KSL cells exhibited low levels of reconstitution in the early phase after transplantation, but reconstitution levels gradually increased thereafter (15 out of 19; Fig. 3). CD150medCD34KSL cells exhibited relatively high levels of reconstitution in the early phase, and these levels were maintained (11 out of 19). CD150negCD34KSL cells exhibited low levels of reconstitution in the early phase without change thereafter (9 out of 14), with the exception of one case in which the level of reconstitution gradually increased. Lineage compositions in reconstituted mice significantly differed between recipients of CD150high and recipients of CD150negCD34KSL cells (P < 0.0001 by the t test; Fig. 3). When CD150highCD34KSL cells were transplanted, the myeloid lineage was more reconstituted than the lymphoid lineage. In contrast, when CD150negCD34KSL cells were transplanted, the lymphoid lineage was more reconstituted than the myeloid lineage. CD150medCD34KSL cells gave rise to a seemingly intermediate pattern between those obtained using CD150high cells and those obtained using CD150neg cells. In addition, the percentages of chimerism 16 wk after transplantation with CD150medCD34KSL cells were significantly greater than those with CD150high or CD150negCD34KSL cells (Fig. 2). CD38high and CD38medCD34KSL cells exhibited reconstitution kinetics similar to CD150high and CD150negCD34KSL cells, respectively (unpublished data).

Figure 3.
Reconstitution kinetics with CD150high, CD150med, and CD150neg CD34KSL cells. Each of a group of lethally irradiated mice received 10 CD150high cells, 10 CD150med cells, or 10 CD150negCD34KSL cells. (left) The change in the percentage ...

The similarity in reconstitution pattern between CD38high and CD150high CD34KSL cells prompted us to examine simultaneously the expression of these two markers on CD34KSL cells. As shown in Fig. S3, most CD38highCD34KSL cells expressed CD150 strongly. These data let us focus on CD150 as a marker that might permit identification of distinct HSC subsets.

Identification of distinct LTRCs

To compare the frequencies and functional properties of LTRCs within CD150high, CD150med, and CD150negCD34KSL cells, single-cell transplantation was performed. Cells were individually transplanted into each of a total of 40 lethally irradiated mice together with 2 × 105 competitor cells in two independent experiments. 6–7 mo later, secondary transplantation was performed by injecting 5 × 106 bone marrow cells from reconstituted recipient mice into three to five lethally irradiated mice.

During these experiments, we made an important observation. Some LTRCs showed very low levels of myeloid reconstitution in primary-recipient mice but exhibited much higher reconstitution levels in secondary-recipient mice. In this study, therefore, we considered reconstitution to have been achieved when the percentage of donor chimerism was 0.3% or more at one time point or more after transplantation, regardless of which lineage was reconstituted.

Results of primary and secondary transplantation with single cells are shown in Fig. 4. In the three 40-mouse cohorts that were primary recipients of CD150high, CD150med, and CD150negCD34KSL cells, LTRC activity was detected in 16, 13, and 13 mice, respectively. The similar rates of reconstitution suggest that LTRCs are equally present among CD150high, CD150med, and CD150negCD34KSL cells.

Figure 4.
Single-cell transplantation. (A–C) Single CD150highCD34KSL (A), CD150medCD34KSL (B), and CD150negCD34KSL cells (C) were transplanted into 40 lethally irradiated mice together with 2 × 105 competitor cells. Blood ...

Results of secondary transplantation differed remarkably, however. After secondary transplantation, LTRCs were detected in all of the recipient mice transplanted from primary-recipient mice that had initially received CD150highCD34KSL cells (Fig. 4 A). In most recipient mice, hematopoiesis was sustained 5 mo after secondary transplantation. Unfortunately, three primary-recipient mice in this cohort could not be used as donors for secondary transplantation because they died early.

Among the primary recipients of CD150med and CD150neg CD34KSL cells, one and two mice died, respectively. LTRCs were detected in 4 out of 13 and in 2 out of 12 cohorts of secondary-recipient mice whose donors had initially received CD150med and CD150negCD34KSL cells, respectively. Consistent with data from 10-cell transplantation experiments (Fig. 3), CD150highCD34KSL cells predominantly reconstituted the myeloid lineage, whereas CD150negCD34KSL cells predominantly reconstituted the lymphoid lineage (Fig. 4).

With respect to in vivo reconstitution kinetics, percentages of chimerism progressively increased throughout primary and secondary transplantation in 10 out of 13 cases initiated with single CD150highCD34KSL cells. When CD150medCD34KSL cells were transplanted, this progressive increase occurred in only 1 out of 13 cases, and when CD150negCD34KSL cells were transplanted, no such instance was observed out of 12 cases. From these data, we concluded that CD150highCD34KSL cells are greatly enriched in highly self-renewing LTRCs.

Noteworthy is that a latent and a myeloid-limited type of LTRCs existed among CD150highCD34KSL cells. Cells of the latent type represented 1 out of every 10 CD150highCD34KSL cells (approximately one in a million total bone marrow cells) and produced virtually undetectable levels of blood cells for several months after transplantation. Beginning at 12 wk or later, these LTRCs effected a low level of myeloid reconstitution (Fig. 4 A, cases #3, #5, #6, and #16). However, these cells showed progressive and robust repopulating activity in secondary-recipient mice. Cells of the myeloid-limited type represented 1 out of 20 CD150highCD34KSL cells and exhibited a low level of myeloid reconstitution without lymphoid reconstitution during primary and secondary transplantation (Fig. 4 A, cases #7 and #15).

Figs. S4–S6 show reconstitution kinetics in individual secondary-recipient mice of bone marrow cells reconstituted with single CD150high, CD150med, and CD150negCD34KSL cells. When high levels of reconstitution took place in primary-recipient mice, high levels of reconstitution were achieved in all secondary-recipient mice (e.g., cases #1 and #2; Fig. S4). As previously reported (Müller-Sieburg et al., 2002; Dykstra et al., 2007), reconstitution patterns were very similar among secondary-recipient mice in each group. However, in the case of latent LTRCs, various levels of reconstitution were observed for secondary-recipient mice in each group (e.g., cases #6 and #16; Fig. S4). Furthermore, reconstitution patterns differed remarkably among recipient mice. Although only myeloid-lineage reconstitution was detected in primary-recipient mice of latent LTRCs, B and T lymphoid lineages became readily detectable in most secondary-recipient mice.

Dykstra et al. (2007) recently proposed that four types of LTRCs exist based on their observations in single-cell transplantation. According to their classification, α cells show a myeloid/lymphoid contribution ratio of >2:1, β cells show a balanced ratio between 1:4 and 2:1, γ cells show a myeloid/lymphoid contribution ratio of <1:4, and δ cells show a ratio of <1:4, with < 1% myeloid reconstitution. As shown in Fig. S7, 75% of CD150highCD34KSL LTRCs qualified as α cells and the remaining cells qualified as β cells. On the other hand, 100% of CD150negCD34KSL LTRCs qualified as γ or δ cells. Among CD150medCD34KSL LTRCs, 15% were assigned to each of the α, β, and γ cell categories, with the remaining cells assigned as δ cells. Collectively, these data suggest a continuum of expression of CD150 among CD34KSL cells, where CD150 expression is tightly associated with repopulation kinetics and with self-renewal potential.

Identification of balanced LTRCs

On single-cell transplantation, CD150medCD34KSL cells did not show any unique properties as compared with CD150high or CD150negCD34KSL cells. Also, the percentage of balanced LTRCs, like β cells, in all single-cell transplants was less than that in previous studies (Ema et al., 2005; Dykstra et al., 2007). In the single-cell experiments described in the previous section, we set a rather narrow gate for the CD150medCD34KSL cells to see the difference more clearly, as in Fig. 1 (bottom). Although a relatively large number of cells lay between the CD150med and CD150high CD34KSL fractions, these cells were not sorted. We assumed that balanced LTRCs are mostly in this population. To address the effects of gating, we performed two independent single-cell transplantation experiments. In these experiments, gates were set to sort CD34KSL cells into four fractions: CD150high, CD150int, CD150low, and CD150neg, as shown in Fig. S8 A. LTRCs were detected in 11, 12, 10, and 9 recipient mice of CD150high, CD150int, CD150low, and CD150negCD34KSL cells, respectively. Results from the CD150high and CD150neg transplants were nearly identical to the initial single-cell transplantation experiments (Fig. S7 B). On the other hand, CD150intCD34KSL LTRCs exhibited high levels of chimerism (Fig. S8 B), and the majority (7 out of 12) were classified as β cells, revealing that large numbers of balanced LTRCs were present just beneath the CD150highCD34KSL cells. These data provide further evidence that functionally distinct LTRCs can be identified according to their levels of CD150 expression.

Colony-forming ability in distinct LTRCs

Single-cell liquid cultures were performed to compare in vitro colony-forming abilities among CD150high, CD150med, and CD150negCD34KSL cells. Cells were individually cultured (48 cells from each fraction) for 14 d in the presence of stem cell factor, thrombopoietin, IL-3, and erythropoietin.

Frequencies of colony formation were similar among these fractions (Fig. 5). The colony types formed, however, differed markedly. On average, 60% of CD150highCD34KSL cells formed neutrophil/macrophage/erythroblast/megakaryocyte (nmEM) colonies, whereas only 7% of CD150negCD34KSL cells did so. In contrast, 66% of CD150negCD34KSL cells formed neutrophil/macrophage colonies, whereas 12% of CD150highCD34KSL cells did so. CD150medCD34KSL cells seemed to exhibit activity in between that of CD150high and CD150negCD34KSL cells. These results suggest that the loss of erythroblast/megakaryocyte differentiation potential is associated with loss of CD150 expression within CD34KSL cells.

Figure 5.
Colony formation by single cells. Single-cell cultures were performed using CD34KSL, CD150highCD34KSL, CD150medCD34KSL, and CD150negCD34KSL cells. 48 CD34KSL, CD150highCD34KSL, CD150medCD34 ...

Hierarchical order in distinct LTRCs

From transplantation data, we predicted that CD150high cells reside in the uppermost portion of the hematopoietic hierarchy. To assess this, we evaluated the reversibility of CD150 expression. Three cohorts of three mice that had been reconstituted with 10 CD150high, CD150med, or CD150negCD34KSL cells were analyzed 8 mo after transplantation. Flow cytometric analysis was performed on bone marrow cells from these mice to assess CD150 expression on reconstituted CD34KSL cells. To distinguish test donor-derived cells from competitor- and recipient-derived cells, only Ly5.1+Ly5.2 cells were gated and analyzed. As shown in Fig. 6, test donor-derived CD150highCD34KSL cells were detected in each of the three mice transplanted with CD150highCD34KSL cells. The CD150 expression pattern in the reconstituted CD34KSL cells mimicked that seen in freshly isolated CD34KSL cells. Very few CD150highCD34KSL cells were detected in one out of three mice transplanted with CD150medCD34KSL cells. Test donor-derived KSL cells were not detected in three mice transplanted with CD150negCD34KSL cells (unpublished data). These data support our hypothesis that CD150high cells rank highly among CD34KSL cells.

Figure 6.
Regeneration of CD150highCD34KSL cells after transplantation. (A) Three recipient mice reconstituted with 10 CD150high or CD150medCD34KSL cells were analyzed 8 mo after transplantation. CD34 and CD150 expression by KSL bone marrow cells ...


Single-cell culture and transplantation are powerful assays for the characterization of individual HSCs (Ema et al., 2000, 2005, 2006; Takano et al., 2004; Seita et al., 2007). Our experience with >2,000 single-cell transplantations of CD34KSL cells leads us to expect engraftment in 20–50% of recipients. Since establishment of this experimental system, we have been searching for markers useful to further enrich for HSCs. During the course of these studies, we have found, even among CD34KSL cells, extensive heterogeneity in reconstitution activity that could be identified only by clonal analysis. This study challenges received ideas on the prospective isolation and identification of HSC subsets. We identified three distinct subsets of LTRCs with regard to repopulating activity, self-renewal potential, in vitro colony-forming activity, and in vivo kinetics after transplantation.

CD150highCD34KSL cells were highly enriched in LTRCs, with greater self-renewal potential than cells in the other two subsets (Figs. 3 and and4).4). The repopulating activity increased gradually from primary through secondary transplantation. Myeloid reconstitution significantly dominated lymphoid reconstitution. On average, 87.5% of CD150highCD34KSL cells formed colonies in vitro. More than half of the colonies included all nmEM elements. These data imply an association among high self-renewal potential, myeloid-biased reconstitution, and nmEM differentiation potential.

CD150negCD34KSL cells were highly enriched in LTRCs, with limited self-renewal potential (Figs. 3 and and4).4). The repopulating activity decreased from primary through secondary transplantation. Lymphoid reconstitution significantly dominated myeloid reconstitution. Only limited myeloid reconstitution was observed in some cases. On average, 85% of CD150negCD34KSL cells formed colonies in vitro, and 73% of these colonies consisted of only neutrophil/macrophage or macrophage elements. These data imply an association among low self-renewal potential, lymphoid-biased reconstitution, and loss of erythroid and megakaryocytic potential.

Kent et al. (2009) recently reported that HSCs with higher repopulating activity are enriched in the CD150+ subset of EPCR+CD48CD45+ bone marrow cells, whereas those in the CD150 subset have limited self-renewal activity. These data all support our hypothesis that CD150 expression levels can be used prospectively to separate distinct types of LTRCs.

In this study, we found latent LTRCs among CD150highCD34KSL cells that exhibited a very late onset of repopulation (>4 mo after transplant) despite exposure to the “cytokine storm” presumed to rage in the early phase of reconstitution. Cells similar to LTRCs of the latent type have been described by some groups (Ortiz et al., 1999; Dykstra et al., 2007). The existence of such LTRCs challenges the current experimental definition of HSCs: multilineage reconstitution in primary recipients 12–16 wk after transplantation. Of importance is that, in addition to this late onset of engraftment, these cells do not always show multilineage reconstitution at a particular point after transplantation. Because of these unique properties, these cells are not likely defined as HSCs by current experimental criteria. Furthermore, we have found mice in which engraftment was not observed until 8 mo after transplant. We have also experienced cases in which multilineage engraftment was confirmed only in the secondary recipients, with no engraftment at all in the primary recipients (unpublished data).

In addition, we have found “myeloid-limited LTRCs” among CD150highCD34KSL cells. These cells exhibited low-level but continuous myeloid reconstitution for at least 40 wk, with no detectable lymphoid differentiation potential. We speculate that they may be downstream of latent HSCs, because latent HSCs exhibit myeloid-predominant reconstitution in primary-recipient mice and express the same markers as those expressed by myeloid-limited LTRCs. CD150highCD34KSL cells also contain at least two other types of LTRCs. One exhibits balanced reconstitution during primary and secondary transplantations (Fig. 4 A, cases #1 and #12). The other exhibits myeloid-predominant reconstitution, with high repopulating activity and without delayed reconstitution, during primary and secondary transplantations (Fig. 4 A, cases #2, #8, and #10). The use of additional markers may permit distinction of these functionally distinct LTRCs from among the CD150highCD34KSL population. In addition, there may be other novel types of HSCs that were not detected by the current assay system. Given such heterogeneity even among the most primitive HSC compartment, the criteria for HSC identification must be redefined. Considering the data presented in this paper, multilineage reconstitution in a secondary transplant recipient appears to be a useful baseline definition for conventional HSCs. This definition would probably include HSCs and define myeloid-limited LTRCs. If this definition is applied, 11 CD150high, 2 CD150med and 1 CD150negCD34KSL LTRCs in the initial single-cell transplantation experiments are considered as conventional HSCs. However, this definition requires expensive and time-consuming experiments. Therefore, novel assays that permit rapid and efficient detection of all sorts of HSCs need to be developed.

Of interest is how such HSCs with diverse reconstitution capacity and kinetics are generated. We have shown that HSC dormancy occurs via hibernation (Yamazaki et al., 2006). Although latent HSCs appear to be found primarily within CD150highCD34KSL cells, because they are so infrequent it is as yet unclear whether they differ from other HSCs in their expression of FOXO transcriptional factors, cell-cycle regulators, and other molecules assumed to be involved in hibernation. Alternatively, different niches may differently regulate HSC hibernation and responses to cytokines. Of particular interest will be to determine to what extent the latent HSCs found in this study are relevant to the recently reported infrequently cycling HSCs among CD150+CD34KSL cells (Wilson et al., 2008; Foudi et al., 2009). Recent reports of cell-cycle regulation by TGF-β in HSCs (Yamazaki et al., 2008; Challen et al., 2010) raise the possibility of its involvement in the unique behavior of latent HSCs. Understanding the mechanism that controls the latent HSC cell cycle may contribute to exploration of the pathogenesis and treatment of some forms of leukemia in which quiescent leukemic stem cells may be highly resistant to current chemotherapy.

As the heterogeneity of HSCs with respect to their differentiation potentials becomes better understood, hematopoietic differentiation pathways from HSCs to mature blood cells also become more complicated than previously thought (Adolfsson et al., 2005; Arinobu et al., 2007). In particular, early differentiation pathways remain poorly understood. Based on our data from this study, we propose a hierarchical organization model in which CD150high cells with high self-renewal potential give rise to CD150med cells with moderate self-renewal potential and, finally, to CD150neg cells with very limited self-renewal potential (Fig. S9).

When CD150high, CD150med, and CD150negCD34KSL LTRCs in secondary recipients were classified based on the schema in Dykstra et al. (2007), it was interesting to note that their classes shifted toward more lymphoid potential than in primary recipients (Fig. S10). The latent HSCs also exhibited varying reconstitution patterns among secondary recipients in the same groups. These data support our hierarchical organization model.

LMPPs have been shown to have B lymphoid, T lymphoid, and neutrophil/macrophage differentiation potentials without erythroblast/megakaryocyte differentiation potential (Adolfsson et al., 2005). In this study, we showed that CD150neg cells have very similar potential but greater repopulating activity. Thus, it seems likely that CD150neg cells give rise to LMPPs. Reduced expression of CD150 is associated with reduction of self-renewal potential and loss of erythroblast/megakaryocyte differentiation potential, leading to neutrophil/macrophage and lymphoid-lineage specification. Although CD150 expression levels can be used to mark self-renewal potential and myeloid predominance, as yet we have no idea why HSCs with high self-renewal activity express high levels of CD150, nor do we know the function of CD150. Our data imply that self-renewal potential, biased differentiation manner, and reconstitution kinetics are predetermined by an intrinsic program within HSCs. Future studies should clarify the physiological and pathological roles of these HSC subsets in the hematopoietic system.



C57BL/6 mice congenic for the Ly5 locus (B6-Ly5.1 mice) were bred and maintained at Sankyo Labo Service. B6-Ly5.2 mice were purchased from Japan SLC. All procedures were approved by the Animal Care and Use Committee, Institute of Medical Science, University of Tokyo.

Screening of antibodies.

Antibodies used for screening are listed in Table S1. Bone marrow cells were obtained from 8–10-wk-old B6-Ly5.1 mice and were suspended in PBS. Low-density cells (<1.077 g/ml) were isolated by density gradient centrifugation and were stained with each additional antibody for screening. For staining of FITC- or Alexa Fluor 488–conjugated additional antibodies, CD34KSL cells were identified by PE-conjugated anti–Sca-1 (D7; eBioscience), PE-Cy7–conjugated anti–c-Kit (2B8; eBioscience), Alexa Fluor 647–conjugated anti-CD34 (RAM-34; eBioscience), and an allophycocyanin (APC)-Cy7–conjugated or a biotinylated anti–lineage marker antibody cocktail (anti-CD4, -CD8, -CD11b, -CD45R, -CD127, –Gr-1, and -TER119 antibodies; eBioscience). The biotinylated anti–lineage marker cocktail was developed using streptavidin-conjugated APC-Cy7.

For staining of PE-conjugated antibodies, as when PE-conjugated anti-CD38 (90) or PE-conjugated anti-CD150 (TC15-12F12.2; BioLegend) was used, CD34KSL cells were identified by FITC-conjugated anti-CD34 (RAM-34), PE-Cy7–conjugated anti–Sca-1, APC-conjugated anti–c-Kit, and APC-Cy7–conjugated or biotinylated anti–lineage marker antibodies. The biotinylated antibodies of interest were developed with streptavidin-PE. PE-conjugated anti–rabbit IgG was used for visualization of unconjugated antibody. To test anti-CD4, -CD8, -CD11b, -CD45R, -CD127, –Gr-1, or -TER119 antibodies, each antibody was excluded from the lineage marker cocktail. To prepare cells for sorting, low-density cells were depleted of lineage marker–positive cells using magnetic beads (Miltenyi Biotec).

Flow cytometric analysis and sorting were performed on a MoFlo (Beckman Coulter) equipped with solid-state (488-nm) and HeNe (633-nm) lasers, or a FACSVantage SE (BD).


Competitive repopulation assays with 2 × 105 whole bone marrow cells from Ly5.1/Ly5.2-F1 mice were performed in B6-Ly5.2 mice lethally irradiated with two doses of 4.75 Gy, delivered 4 h or more apart. 1 or 10 cells were transplanted per recipient mouse. Peripheral blood cells from the recipient mice were analyzed at 1, 2, 3, 4, and 5 mo after transplantation. After erythrocyte lysis, cells were stained with FITC-conjugated anti–Ly5.2, PE-Cy7–conjugated anti-B220, PE-conjugated anti-CD4 and -CD8, APC-conjugated anti–Mac-1 and –Gr-1, and biotinylated anti-Ly5.1 antibodies. The biotinylated antibody was developed with streptavidin–Alexa Fluor 594. Before flow cytometric analysis, cells were suspended in PBS containing 1 µg/ml propidium iodide. Six-color flow cytometric analysis was performed using a FACSVantage SE equipped with argon (488-nm) and dye (599-nm) lasers. The percentage of chimerism was defined as follows: (percentage of Ly5.1+ test donor cells) × 100/(percentage of Ly5.1+ test donor cells + percentage of F1 competitor cells). If chimerism was 0.3% or more at any time point after transplantation, regardless of which lineage was reconstituted, mice were considered to be reconstituted with test donor cells.

Secondary transplantation was performed by transferring 5 × 106 bone marrow cells from femurs and tibias of the primary-recipient mice into each of three to five lethally irradiated B6-Ly5.2 mice. Peripheral blood cells from the secondary-recipient mice were analyzed at 1, 2, 3, 4, and 5 mo after transplantation.

LTRCs were defined as cells that exhibited 0.3% or more chimerism 16 wk or more after transplantation. Latent HSCs were defined as HSCs that began to exhibit 0.3% or more chimerism beginning at 12 wk or more after primary transplantation and exhibited multilineage reconstitution after secondary transplantation.

Single-cell colony assays.

Cells were sorted clonally into 96-well plates containing 200 µl S-clone SF-O3 medium (Sanko Junyaku) supplemented with 10% fetal bovine serum, 1% bovine serum albumin, 10 ng/ml mouse stem cell factor, 10 ng/ml human thrombopoietin, 10 ng/ml mouse IL-3, 1 U/ml human erythropoietin (PeproTech), and 5 × 10−5 M β-mercaptoethanol. After culture for 14 d, colonies were cytocentrifuged onto glass slides and stained with Hemacolor (Merck). Cells comprising colonies were identified by light microscopy using morphological criteria.

Online supplemental material.

Fig. S1 shows flow cytometric profiles of the markers that were expressed by CD34+KSL cells but not by CD34KSL cells. Fig. S2 shows flow cytometric profiles of the markers that were expressed by CD34KSL cells. Fig. S3 shows coexpression of CD150 and CD38 in CD34KSL cells. Figs. S4–S6 show the percentage of chimerism of individual secondary-recipient mice. Fig. S7 shows the classification of LTRCs in CD150 subsets. Fig. S8 shows that the CD150intCD34KSL fraction is enriched in balanced LTRCs. Fig. S9 shows the hierarchal organization model within the primitive HSC compartment. Fig. S10 shows the classification of LTRCs in secondary transplantation. Online supplemental material is available at


We thank A.S. Knisely for critical review of the manuscript.

This work was supported by grants from the Japanese Ministry of Education, Culture, Sports, Science, and Technology.

The authors have no conflicting financial interests.


Abbreviations used:

CD34−/low c-Kit+Sca-1+Lin
hematopoietic stem cell
lymphoid-primed multipotent progenitor
long-term repopulating cell
neutrophil/macrophage/ erythroblast/megakaryocyte


  • Abkowitz J.L., Golinelli D., Harrison D.E., Guttorp P. 2000. In vivo kinetics of murine hemopoietic stem cells. Blood. 96:3399–3405 [PubMed]
  • Adolfsson J., Borge O.J., Bryder D., Theilgaard-Mönch K., Astrand-Grundström I., Sitnicka E., Sasaki Y., Jacobsen S.E. 2001. Upregulation of Flt3 expression within the bone marrow Lin(−)Sca1(+)c-kit(+) stem cell compartment is accompanied by loss of self-renewal capacity. Immunity. 15:659–669 10.1016/S1074-7613(01)00220-5 [PubMed] [Cross Ref]
  • Adolfsson J., Månsson R., Buza-Vidas N., Hultquist A., Liuba K., Jensen C.T., Bryder D., Yang L., Borge O.J., Thoren L.A., et al. 2005. Identification of Flt3+ lympho-myeloid stem cells lacking erythro-megakaryocytic potential: a revised road map for adult blood lineage commitment. Cell. 121:295–306 10.1016/j.cell.2005.02.013 [PubMed] [Cross Ref]
  • Arinobu Y., Mizuno S., Chong Y., Shigematsu H., Iino T., Iwasaki H., Graf T., Mayfield R., Chan S., Kastner P., Akashi K. 2007. Reciprocal activation of GATA-1 and PU.1 marks initial specification of hematopoietic stem cells into myeloerythroid and myelolymphoid lineages. Cell Stem Cell. 1:416–427 10.1016/j.stem.2007.07.004 [PubMed] [Cross Ref]
  • Balazs A.B., Fabian A.J., Esmon C.T., Mulligan R.C. 2006. Endothelial protein C receptor (CD201) explicitly identifies hematopoietic stem cells in murine bone marrow. Blood. 107:2317–2321 10.1182/blood-2005-06-2249 [PubMed] [Cross Ref]
  • Challen G.A., Boles N.C., Chambers S.M., Goodell M.A. 2010. Distinct hematopoietic stem cell subtypes are differentially regulated by TGF-beta1. Cell Stem Cell. 6:265–278 10.1016/j.stem.2010.02.002 [PMC free article] [PubMed] [Cross Ref]
  • Dykstra B., Kent D., Bowie M., McCaffrey L., Hamilton M., Lyons K., Lee S.J., Brinkman R., Eaves C. 2007. Long-term propagation of distinct hematopoietic differentiation programs in vivo. Cell Stem Cell. 1:218–229 10.1016/j.stem.2007.05.015 [PubMed] [Cross Ref]
  • Ema H., Takano H., Sudo K., Nakauchi H. 2000. In vitro self-renewal division of hematopoietic stem cells. J. Exp. Med. 192:1281–1288 10.1084/jem.192.9.1281 [PMC free article] [PubMed] [Cross Ref]
  • Ema H., Sudo K., Seita J., Matsubara A., Morita Y., Osawa M., Takatsu K., Takaki S., Nakauchi H. 2005. Quantification of self-renewal capacity in single hematopoietic stem cells from normal and Lnk-deficient mice. Dev. Cell. 8:907–914 10.1016/j.devcel.2005.03.019 [PubMed] [Cross Ref]
  • Ema H., Morita Y., Yamazaki S., Matsubara A., Seita J., Tadokoro Y., Kondo H., Takano H., Nakauchi H. 2007. Adult mouse hematopoietic stem cells: purification and single-cell assays. Nat. Protoc. 1:2979–2987 10.1038/nprot.2006.447 [PubMed] [Cross Ref]
  • Foudi A., Hochedlinger K., Van Buren D., Schindler J.W., Jaenisch R., Carey V., Hock H. 2009. Analysis of histone 2B-GFP retention reveals slowly cycling hematopoietic stem cells. Nat. Biotechnol. 27:84–90 10.1038/nbt.1517 [PMC free article] [PubMed] [Cross Ref]
  • Haug J.S., He X.C., Grindley J.C., Wunderlich J.P., Gaudenz K., Ross J.T., Paulson A., Wagner K.P., Xie Y., Zhu R., et al. 2008. N-cadherin expression level distinguishes reserved versus primed states of hematopoietic stem cells. Cell Stem Cell. 2:367–379 10.1016/j.stem.2008.01.017 [PubMed] [Cross Ref]
  • Kent D.G., Copley M.R., Benz C., Wöhrer S., Dykstra B.J., Ma E., Cheyne J., Zhao Y., Bowie M.B., Gasparetto M., et al. 2009. Prospective isolation and molecular characterization of hematopoietic stem cells with durable self-renewal potential. Blood. 113:6342–6350 [PubMed]
  • Kiel M.J., Yilmaz O.H., Iwashita T., Yilmaz O.H., Terhorst C., Morrison S.J. 2005. SLAM family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells. Cell. 121:1109–1121 10.1016/j.cell.2005.05.026 [PubMed] [Cross Ref]
  • Kiel M.J., Radice G.L., Morrison S.J. 2007. Lack of evidence that hematopoietic stem cells depend on N-cadherin-mediated adhesion to osteoblasts for their maintenance. Cell Stem Cell. 1:204–217 10.1016/j.stem.2007.06.001 [PubMed] [Cross Ref]
  • Kiel M.J., Acar M., Radice G.L., Morrison S.J. 2009. Hematopoietic stem cells do not depend on N-cadherin to regulate their maintenance. Cell Stem Cell. 4:170–179 [PMC free article] [PubMed]
  • Matsubara A., Iwama A., Yamazaki S., Furuta C., Hirasawa R., Morita Y., Osawa M., Motohashi T., Eto K., Ema H., et al. 2005. Endomucin, a CD34-like sialomucin, marks hematopoietic stem cells throughout development. J. Exp. Med. 202:1483–1492 10.1084/jem.20051325 [PMC free article] [PubMed] [Cross Ref]
  • Morita Y., Ema H., Yamazaki S., Nakauchi H. 2006. Non-side-population hematopoietic stem cells in mouse bone marrow. Blood. 108:2850–2856 10.1182/blood-2006-03-010207 [PubMed] [Cross Ref]
  • Müller-Sieburg C.E., Cho R.H., Thoman M., Adkins B., Sieburg H.B. 2002. Deterministic regulation of hematopoietic stem cell self-renewal and differentiation. Blood. 100:1302–1309 [PubMed]
  • Müller-Sieburg C.E., Cho R.H., Karlsson L., Huang J.F., Sieburg H.B. 2004. Myeloid-biased hematopoietic stem cells have extensive self-renewal capacity but generate diminished lymphoid progeny with impaired IL-7 responsiveness. Blood. 103:4111–4118 10.1182/blood-2003-10-3448 [PubMed] [Cross Ref]
  • Ortiz M., Wine J.W., Lohrey N., Ruscetti F.W., Spence S.E., Keller J.R. 1999. Functional characterization of a novel hematopoietic stem cell and its place in the c-Kit maturation pathway in bone marrow cell development. Immunity. 10:173–182 10.1016/S1074-7613(00)80018-7 [PubMed] [Cross Ref]
  • Osawa M., Hanada K., Hamada H., Nakauchi H. 1996. Long-term lymphohematopoietic reconstitution by a single CD34-low/negative hematopoietic stem cell. Science. 273:242–245 10.1126/science.273.5272.242 [PubMed] [Cross Ref]
  • Seita J., Ema H., Ooehara J., Yamazaki S., Tadokoro Y., Yamasaki A., Eto K., Takaki S., Takatsu K., Nakauchi H. 2007. Lnk negatively regulates self-renewal of hematopoietic stem cells by modifying thrombopoietin-mediated signal transduction. Proc. Natl. Acad. Sci. USA. 104:2349–2354 10.1073/pnas.0606238104 [PubMed] [Cross Ref]
  • Sieburg H.B., Cho R.H., Dykstra B., Uchida N., Eaves C.J., Muller-Sieburg C.E. 2006. The hematopoietic stem compartment consists of a limited number of discrete stem cell subsets. Blood. 107:2311–2316 10.1182/blood-2005-07-2970 [PubMed] [Cross Ref]
  • Sudo K., Ema H., Morita Y., Nakauchi H. 2000. Age-associated characteristics of murine hematopoietic stem cells. J. Exp. Med. 192:1273–1280 10.1084/jem.192.9.1273 [PMC free article] [PubMed] [Cross Ref]
  • Takano H., Ema H., Sudo K., Nakauchi H. 2004. Asymmetric division and lineage commitment at the level of hematopoietic stem cells: inference from differentiation in daughter cell and granddaughter cell pairs. J. Exp. Med. 199:295–302 10.1084/jem.20030929 [PMC free article] [PubMed] [Cross Ref]
  • Uchida N., Dykstra B., Lyons K.J., Leung F.Y., Eaves C.J. 2003. Different in vivo repopulating activities of purified hematopoietic stem cells before and after being stimulated to divide in vitro with the same kinetics. Exp. Hematol. 31:1338–1347 10.1016/j.exphem.2003.09.001 [PubMed] [Cross Ref]
  • Wilson A., Laurenti E., Oser G., van der Wath R.C., Blanco-Bose W., Jaworski M., Offner S., Dunant C.F., Eshkind L., Bockamp E., et al. 2008. Hematopoietic stem cells reversibly switch from dormancy to self-renewal during homeostasis and repair. Cell. 135:1118–1129 10.1016/j.cell.2008.10.048 [PubMed] [Cross Ref]
  • Yamazaki S., Iwama A., Takayanagi S., Morita Y., Eto K., Ema H., Nakauchi H. 2006. Cytokine signals modulated via lipid rafts mimic niche signals and induce hibernation in hematopoietic stem cells. EMBO J. 25:3515–3523 10.1038/sj.emboj.7601236 [PubMed] [Cross Ref]
  • Yamazaki S., Iwama A., Takayanagi S.I., Eto K., Ema H., Nakauchi H. 2008. TGF-beta as a candidate bone marrow niche signal to induce hematopoietic stem cell hibernation. Blood. 113:1250–1256 [PubMed]
  • Yoshihara H., Arai F., Hosokawa K., Hagiwara T., Takubo K., Nakamura Y., Gomei Y., Iwasaki H., Matsuoka S., Miyamoto K., et al. 2007. Thrombopoietin/MPL signaling regulates hematopoietic stem cell quiescence and interaction with the osteoblastic niche. Cell Stem Cell. 1:685–697 10.1016/j.stem.2007.10.020 [PubMed] [Cross Ref]
  • Zhang J., Niu C., Ye L., Huang H., He X., Tong W.G., Ross J., Haug J., Johnson T., Feng J.Q., et al. 2003. Identification of the haematopoietic stem cell niche and control of the niche size. Nature. 425:836–841 10.1038/nature02041 [PubMed] [Cross Ref]

Articles from The Journal of Experimental Medicine are provided here courtesy of The Rockefeller University Press