Page 120 - 2021_07-Haematologica-web

A. O’Neill et al.
significantly lower expression, indicative of their reduc- tion in megakaryopoietic potential (Figure 4C). Analysis of the expression of the neutrophil/lymphoid marker (Gfi1) revealed that the expression of Gfi1 was higher in the CD150- HSPC population than in LT-HSC, and its expression was reduced in CD41+ HSPC (Figure 4D). These data suggest that the CD150- HSPC may be lym- phoid-primed at the expense of megakaryopoietic poten- tial, while the CD41+ HSPC may become megakaryocyte- primed at the expense of lymphoid potential.
Loss of quiescence is not responsible for loss of cell number
Previous studies have shown that THPO plays a role in maintaining HSC quiescence.4 In order to investigate this, we first looked at the quiescence levels of the HSC popu- lations in Thpo-/- mice (Figure 5A). The frequency of qui- escence was reduced in all three HSPC populations. Interestingly, quiescence was reduced in CD150- HSPC despite them having limited dependence on THPO/MPL signaling. Previous studies have suggested that megakary- ocytes in the bone marrow are responsible for maintain- ing HSCs in quiescence through the megakaryocyte-spe- cific cytokine CXCL4/PF4.18 We looked at CXCL4/PF4 levels in the bone marrow of Thpo-/- mice and found that CXCL4/PF4 concentration is decreased on loss of THPO (Figure 5B). This suggests that the loss of megakaryocytes in KO mice reduces bone marrow CXCL4/PF4 concentra- tion. To assess whether administration of recombinant CXCL4/PF4 could rescue quiescence in Thpo-/- mice, CXCL4/PF4 was injected daily for 7 days. Treated KO mice showed a significant increase in the proportion of cells in G0 in all three HSPC populations (Figure 5C-E). The quiescence levels in treated KO mice were not signif- icantly different from those of wild-type untreated mice, suggesting that CXCL4/PF4 treatment rescues quiescence back to wild-type levels in all three HSPC populations. Taken together these data would suggest that the HSPC require CXCL4/PF4 from megakaryocytes for mainte- nance of quiescence. Although quiescence is rescued by CXCL4/PF4 treatment the number of LT-HSCs and CD150- HSPC remained unchanged (Figure 5F and G). This would suggest that the loss of cell numbers is inde- pendent of loss of quiescence. To confirm this we treated Thpo-/- mice with the MPL receptor agonist romiplostim for 5 days via tail vein injection. On day 5 platelet counts in Thpo-/- mice had recovered to the level in wild-type untreated mice, suggesting that thrombopoietic potential – and by extension megakaryopoietic potential – is not lost in the absence of THPO signaling and can return to wild-type levels upon rescue of MPL signaling (Online Supplementary Figure S5A).
Discussion
We have shown that loss of THPO signaling in bone mar- row leads to a reduction of cell numbers throughout the hematopoietic compartment, primarily in cells that express high levels of MPL and show potential for megakaryopoiet- ic differentiation. We show that the CD150- HSPC popula- tion has low MPL expression and reduced megakaryopoiet- ic potential, findings that are consistent with previous reports that these cells are lymphoid-biased.14,16 Previous reports have defined the CD150- HSPC and CD41+ HSPC
populations as HSC, although strictly speaking these popu- lations contain predominantly hematopoietic progenitor cells, rather than true HSCs. In spite of this, it is important to note that these populations are both capable of repopu- lating the bone marrow of lethally irradiated mice, though not long-term, indicating a capacity for self-renewal. Both CD41+ HSPC and CD150- HSPC populations also show high levels of quiescence similar to LT-HSC and express cell surface markers very similar to those of LT-HSC, suggesting a very close relationship between the three populations. Indeed, previous studies have suggested that CD41+ HSPC and CD150- HSPCs represent the earliest branch point in the hematopoietic hierarchy, with CD41+ HSPC being the earliest myeloid branch and CD150- HSPC being the earliest lymphoid branch.14 The data in this study support this view.
Despite their low levels of MPL expression and limited dependence on THPO for cell proliferation both in vivo and in vitro, CD150- HSPC show reduced quiescence in the absence of THPO in the bone marrow. Previous studies showed that loss of megakaryocytes in bone marrow results in reduced HSC quiescence,19,20 while a subsequent study showed that this effect is due to loss of megakary- ocyte-derived CXCL4/PF4.18 We show that in the case of THPO KO mouse models, the loss of quiescence in HSPC results from the loss of CXCL4/PF4 signaling due to reduced megakaryocyte number in the bone marrow and that this can be rescued by administration of exogenous CXCL4/PF4. Interestingly, CXCL4/PF4 administration in wild-type mice does not produce an increase in quiescence, suggesting that quiescence is not dependent on CXCL4/PF4 signaling alone. Previous studies have shown that CXCL12 from CXCL12-abundant reticular (CAR) cells in the bone marrow also plays a role in HSPC quiescence.21 Other studies have provided evidence of dimerization of CXC ligands, including CXCL12.22 One theory is that CXCL4/PF4 and CXCL12 form heterodimers that induce quiescence in HSC and that without an increase in CXCL12, increased CXCL4/PF4 cannot induce further qui- escence. Further research may clarify whether CXCL4/PF4-dependent quiescence is co-dependent on other signaling pathways.
Our in vitro and in vivo data suggest that MPL expression in HSC correlates closely with megakaryocytic differentia- tion potential, indicating that THPO in the bone marrow is responsible for maintaining megakaryocytic differentiation potential in HSPC. A recent study from our own group has shown that increased THPO/MPL signaling leads to increased proliferation and megakaryocytic differentiation as well as mitochondrial activation in HSCs, suggesting that THPO drives cell division, rather than suppressing it.23 There are reports that THPO induces self-renewal division in HSC24 and it is possible that loss of this self-renewal divi- sion is responsible for the loss of cell numbers of megakaryopoietic HSPC within the bone marrow. Here we show that THPO is required for proliferation of HSPC with megakaryopoietic potential while previous studies showed that it is required for the maturation of megakary- ocytes from MkP. Together, this would suggest that THPO plays a dual role in maintaining the megakaryocyte popu- lation and that both loss of megakaryocyte-producing HSPC and impairment of MkP maturation lead to the acute reduction of megakaryocytes seen in KO models. In humans, injection of THPO leads to an autoimmune response against the exogenous protein and even against endogenous THPO. For this reason, the artificial recombi-
1890
haematologica | 2021; 106(7)