HSPCs are marathon and not sprint relay athletes

Lambert Busque

In this issue of Blood, Koelle et al demonstrate that hematopoietic output originates from multiple multipotent clones that are stable over time, with little evidence of clonal succession.1

The hematopoietic compartment has the daunting task of daily producing an enormous quantity of various blood cells. This daily miracle relies upon a very limited number of super athlete cells called hematopoietic stem cells (HSCs). The HSC reigns on a hierarchical system where self-renewal capacity and differentiation toward specific lineages are finely regulated.2 An in-depth understanding of hematopoiesis has obvious implications for leukemogenesis, stem cell transplantation, gene therapy, and biology of aging. Studying the clonal dynamics of hematopoiesis is challenging because the cells of interest are elusive and their progeny not readily identifiable. Fortunately, recent technological advances have supported the development of different models and strategies that provided new insights in the biology of HSCs.

In their article, Koelle et al transplanted 4 rhesus macaques with autologous CD34+ cells transduced with green fluorescent protein–labeled lentiviral vector libraries carrying high-diversity oligonucleotide barcodes after myeloablative transplantation. The diversity of the barcodes was sufficient to ensure that each barcode marked a single engrafting HSC and progenitor cell (HSPC). After transplantation, iterative peripheral blood samples were collected (up to 49 months); lineage-specific cells (granulocytes, monocytes, B and T cells) were isolated and submitted to sequencing for barcode identification and quantification. A significant proportion of hematopoiesis (≈9%-50%) was derived from the labeled CD34+ cells in each animal, was stable over time, and was derived from 3000 to 5000 different clones.

The authors document waves of short-lived cell-type–restricted clones immediately after transplantation which are replaced by long-lived ones after 2 to 4 months. They detected thousands of these long-lived clones which variably contributed to hematopoiesis. A fraction of them contributed disproportionally to hematopoiesis. This was true for all cell types. Remarkably, these long-lived clones were quantitatively stable over the observation period. There was no evidence of clonal succession, exhaustion, or expansion. Only the T-cell fraction saw the emergence of several new large clones, probably responding to adaptive immune response to antigenic stimulation. These results are in line with those reported by Biasco et al, who performed in vivo tracking of the clonal dynamics of 4 patients with Wiskott-Aldrich syndrome treated with lentiviral gene therapy.3 However, they partially contradict the results obtained by Sun et al, who labeled murine native hematopoiesis using Sleeping Beauty transposase.4 In the granulocytic fraction, Sun et al documented bursts of successive clones with no long-term stability and little correlation with B or T cells. These discordant results may be explained by the size of the model, the difference between native and posttransplantation hematopoiesis, and the different technological approaches used to label HSPCs. These discrepancies will be difficult to resolve, the study of native hematopoiesis in humans is limited to noninvasive clonal markers such as X-chromosome inactivation (XCI) ratio analysis.5 That being said, the stable marathon-like behavior concept for HSPCs is appealing.

The results presented by Koelle et al raise an interesting paradox between early replicative advantage and long-term maintenance of hematopoiesis. After the initial posttransplant period characterized by clonal diversity, stable clones emerge. The relative clone size is different. Because the barcode is unique and corresponds to a single HSPC, this indicates that some HSPCs replicated more actively than others during this early phase, generating larger clones. The paradox resides in the fact that, right after this replicative period, no change in clone size is observed. There is institution of a status quo. Minimally, this indicates that there is a genetic diversity among HSPCs,6 and it also raises the question of the equilibrium between expansion and maintenance of the clone. This equilibrium is probably dependent on the complex interactions between genetic, epigenetic, telomere biology, and HSC niche determinants.7 The identification of factors that promote clonal expansion and break this equilibrium is of high clinical relevance. It may be key to understanding why some aging individuals develop clonal hematopoiesis,8 which has been associated with risk of hematological cancer progression.9

Koelle et al also allowed us to reevaluate the proximity of the end lineage cell type. As expected, granulocytes and monocytes are highly correlated and originate from the same HSPCs. Initially, the correlation between B cells and the myeloid fraction was weaker, but it continuously increased with time. In fact, the correlation between B cells and myeloid cells was greater than the correlation between B cells and T cells. This was due to the emergence of significant T-cell clones potentially associated with immune response. Interestingly, we have obtained identical correlations between cell subtypes studying a large cohort of female subjects using an XCI method.10 Taken together, this indicates that: steady-state hematopoiesis is derived by multipotent progenitor cells, the human blood hierarchy paradigm may be slightly different than initially thought, and the paradigm may need to be revisited.

The recently developed clonal tracking methodology used by Koelle et al is complex and costly. However, it is robust and highly quantitative. As demonstrated in the article, it has the potential to unravel several mysteries regarding HSC biology in health and in disease.


  • Conflict-of-interest disclosure: The author declares no competing financial interests.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.