Paroxysmal nocturnal hemoglobinuria (PNH) clones expand in the setting of immune-mediated marrow failure, and functional and molecular studies suggest an escape mechanism. Expansion of the PNH clone in aplastic anemia has been ascribed to insusceptibility to immune attack, resistance to apoptosis, or to an intrinsic proliferative advantage. Urokinase plasminogen activator receptor (sUPAR), normally linked by the glycosylphosphoinositil (GPI) anchor to the cell membrane, is increased in its soluble form in patients with PNH compared to normal; levels correlate with the number of GPI-negative neutrophils. Plasma levels of sUPAR are also increased in patients with solid tumors, where it reportedly plays a role in checking metastasis; in tumor cells, sUPAR increases MAP kinase and the anti-apoptotic protein ATK in cells lacking GPI- anchored membrane UPAR, but paradoxically downregulates these proteins in cells with membrane anchored UPAR (
J Biol Chem, 2003;278:46692
). Here we demonstrate in vitro release of suPAR from PNH hematopoietic cells: supernatants obtained from BM PNH CD34 cells cultured in liquid media with standard growth factors in 3 experiments showed suPAR levels 20-fold greater than those obtained from normal CD34 cells, when measured by ELISA (p<0.01), a finding, not accounted for by cell death. Platelets were a substantial source of suPAR, and uPAR was expressed on the surface of ADP-activated CD59-positive but not CD59-negative platelets from PNH patients. By an ELISA specific for soluble uPAR, substantial quantities of suPAR were released following collagen-activation of PNH but not of normal platelets. In five short-term culture experiments, when recombinant sUPAR was added to media, the numbers of PNH (CD59-negative) colonies increased, while the CD59-positive normal colonies decreased. Extracellular signal-related kinase (ERK) was increased in PNH cells but not in the patients’ GPI-positive cells. For CD59-negative mononuclear cells (BMMNC) from five patients co-cultured with UPAR, annexin binding, a marker of early apoptosis, was decreased by at least 50% (p<0.01) with no change in the annexin in the CD59-positive BMMNC. These changes were dose-dependent and were blocked by addition of anti-sUPAR. sUPAR decreased colony formation by 75% in four patients with AA without a PNH clone, while BM from healthy controls showed significant decreases in the colony size. BM was sorted into GPI-negative and -positive fractions using CD59 as a marker; GPI- negative and positive cells were plated separately and then in a 1:1 mixture. Normal (CD59-positive) colonies grew substantially better when plated alone than when cocultured with CD59 BMMNC, consistent with the effect of a soluble inhibitor released from PNH cells, that inhibited the growth of GPI-positive cells. We hypothesize that UPAR plays an exocrine function, stimulating growth of PNH cells and inhibiting their apoptosis, while having the opposite effect on GPI-positive cells. These functions of UPAR might provide the GPI-negative cell a survival and growth advantage over normal cells, particularly in a BM undergoing immune attack.
