Hematopoietic stem/progenitor cells (HSC/HPCs) are retained in the bone marrow (BM) niche via receptor-ligand interactions and mobilized from the BM after proteolytic degradation of these retention complexes. Yet, the proteinases and retention signals involved remain incompletely identified. Here, we studied the role of the plasminogen proteinase system with its plasminogen activators PA) tPA and uPA, and active plasmin (Pli) in chemo- and G-CSF-induced mobilization of HPCs and HSCs.Therefore, 5-fluorouracil (5-FU) or G-CSF were administered to mice lacking plasminogen (Plg−/−), tPA (tPA−/−), uPA (uPA−/−), both activators (tPA−/−uPA−/−), uPAR (uPAR−/−), PAI-1 (PAI-1−/−), or α2-antiplasmin (α2-AP−/−).5-FU treatment in WT mice elevated Pli activity in BM plasma 5-fold, killed 10% of WT mice, and increased number/proliferation of HSC/HPCs in the surviving mice, with full hematopoietic recovery after 3 weeks. In contrast, up to 75% of 5-FU-treated Plg−/−and tPA−/−uPA−/− demised in the early phase of recovery, with reduced number/proliferation of HSC/HPCs and delayed hematopoietic recovery. Following G-CSF, deficiency of Plg or inhibition of Pli by tranexamic acid reduced HPC expansion and HSC translocation in the BM, resulting in impaired HPC/HSC mobilization, by up to 55% and 75%, respectively.Further analysis using uPA−/− and tPA−/− mice revealed that uPA was critical for 5-FU-induced mobilization, whereas tPA was crucial for G-CSF-induced mobilization. In addition, analysis of mice lacking MMP-2, -3, -9, and -12 revealed that 5-FU- and G-CSF-induced mobilization required predominantly MMP-9 and -3, respectively. MMP-3 and MMP-9 activities upon mobilization were reduced in Plg−/− mice, suggesting that Pli activates these MMPs. In the absence of Plg, degradation of fibronectin in the BM and production of soluble Kit ligand (but not SDF1 α) were impaired, indicating additional downstream targets of Pli.uPAR is a membrane-anchored receptor for uPA, which is cleaved into a soluble form (suPAR) by Pli and other proteinases. Interestingly, uPAR was expressed on BM-derived HPC/HSCs, and uPAR deficiency reduced their retention within the BM niche in vitro and in vivo. Furthermore, uPAR−/− mice showed poor HSC/HPC mobilization in response to 5-FU and G-CSF, while suPAR administration in WT mice amplified G-CSF-induced mobilization. Moreover, suPAR levels in the BM were increased in WT but not in Plg−/− mice during mobilization, indicating, all together, that uPAR might be a novel retention signal for HSC/HPCs in the BM niche.Finally, increased Pli activity in PAI-1−/− and α2-AP−/− mice, and WT mice treated with tenecteplase (i.e. recombinant tPA variant used for clinical thrombolysis) enhanced G-CSF-induced mobilization. Importantly, initial results suggest that thrombolytic treatment of individuals after acute myocardial infarction also seemed to stimulate HPC mobilization, extrapolating our findings to man.In conclusion, these genetic and pharmacological data reveal, for the first time, a novel role for uPAR as a retention signal for HSC/HPCs in the BM and suggest that strategies to increase PA or Pli activity might offer novel therapeutic opportunities for HSC/HPC mobilization.
Read full abstract