CXCR4 mediates immature hematopoietic cell retention in marrow. As a G-protein coupled receptor, it is expected to rapidly adapt to continuously presented ligand concentrations. Thus how immature hematopoietic cells manage to retain sensitivity to its ligand, SDF-1, is unclear. Earlier studies demonstrated that perturbed CXCR4 internalization leads to abnormalities of mature and immature hematopoiesis. G-protein coupled receptor kinase 6 (GRK6) negatively modulates a different aspect of CXCR4 signalling, i.e. ligand-induced receptor desensitization. We therefore sought to explore immature murine hematopoiesis in mice deficient for GRK6.
GRK6+/− mice were gifted by Robert Lefkowitz (Durham, NC). To generate GRK6−/− and WT littermates, GRK6+/− were bred and housed at the comparative medicine facility at Goethe University Frankfurt under non-SPF conditions. Young (8–12 wk) and old (60–100 wk) mice were studied. For transplantation, WT recipients were radio-conditioned with 950 cGy prior to i.v. transplantation of BM or spleen cells from WT and/or GRK6−/− donors. Stress hematopoiesis was induced with phenylhydrazine, 5-fluorouracil or sub-lethal irradiation. Mobilization was done with G-CSF or AMD3100, using standard dosing regimens. Where indicated, mice were splenectomized in deep anaesthesia using aseptic technique. All methods were approved by the IACUC. CBCs were done with an automatic hemacytometer. Progenitor cells were enumerated with cytokine-replete semi-solid media cultures. Surface and intracellular markers were studied by flow cytometry.
GRK6−/− mice were fertile, had normal litter sizes, and pups were born at expected ratios. Weight and length were indistinguishable. GRK6−/− steady-state hematopoiesis was remarkable for lymphocytopenia in peripheral blood, more pronounced in young than in old mice. Circulating CFU-C were similar. Marrow cellularity and CFU-C contents and frequency increase with age was much more pronounced in WT mice. Marrow contents of LSK cells was the same for WT and GRK6−/− mice, young or old. Spleen size and leukocyte cellularity were similar. CFU-C contents of GRK6−/− spleens was almost two-fold greater. Accordingly, transplantation of similar numbers of WT or GRK6−/− CD45.2 spleen cells in competition with CD45.1 WT BM cells resulted in markedly higher donor chimerism for GRK6−/− than for WT cells. Despite this, splenectomy did not result in excessive numbers of circulating CFU-C in GRK6−/− mice. The immunophenotype of c-kit+ BM cells from WT and GRK6−/− mice was similar. Functionally, BM CFU-C were characterized by greater migration towards SDF-1 in transwell migration assays, as well as a greater sensitivity to lower doses of SDF-1. Erk1/2 phosphorylation was markedly stronger in SDF-1 treated GRK6−/− c-kit+ BM cells than in WT controls. This did not, however, translate into altered (improved) homing of transplanted cells to marrow or spleen, nor to altered engraftment kinetics. The same lymphocytopenia that was observed in naïve GRK6−/− mice was also seen 12 weeks after transplantation of WT mice with GRK6−/− cells, indicating cell-intrinsic effects. Serial competitive engraftment assays uncovered a modest engraftment defect; the kinetics of the loss of GRK6−/− contribution to hematopoiesis suggested defects in early niche interaction, but not in retention. Several additional models of stress hematopoiesis were also tested, including mobilization with G-CSF or the CXCR4 antagonist AMD3100, phenylhydrazine, 5-fluorouracil or sub-lethal irradiation; in all cases the response of GRK6−/− mice and WT mice was the same, i.e. did not uncover any marked defects associated with defective SDF-1 signalling.
Given the prominent role of CXCR4 in immature hematopoiesis and observed in vitro hyperresponsiveness of GRK6−/− cells in response to SDF-1, the hematological phenotype of the GRK6−/− mouse is surprisingly modest. The data suggest that ligand-induced receptor phosphorylation is not a key modulator of SDF-1 responsiveness of hematopoietic stem/progenitor cells in the hematopoietic niche in marrow. Specifically, the data also add to the body of evidence indicating that the strength of in vitro SDF-1 responsiveness of cells does not predict homing efficiency nor engraftment. We conclude that GRK6 is a non-essential molecule for in vivo hematopoiesis.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.