Dynamaic Change Of SDF-1 And HGF During G-CSF-induced Peripheral Blood Stem Cell Mobilization

Background and Objective:Hematopoietic stem cell transplantation is an effective therapy for hematopoietic malignancies, aplastic anemia and some solid tumors. Currently, the majority of transplants are performed using peripheral blood as a source of reconstituting cells. The use of mobilized peripheral blood hematopoietic progenitor cells (PBPCs) is associated with more rapid engraftment, and reduced costs as compared with BM transplantation, all of which have contributed to the decline in utilization of BM as a source of Hematopoietic progenitor cells (HPCs). Despite the now widespread use of mobilized PBPCs, there are serious problems ofquality and quantity in mobilized PBPCs. Granulocyte colony-stimulating factor (G-CSF) is the most commonly used agent in mobilization of peripheral blood stem cells (PBSC). Recent evidence suggested that the chemokine stromal-derived factor-1 (SDF-1) and its receptor (CXC receptor 4,CXCR4) played a key role in G-CSF induced PBSC mobilization. There are different reports about the changes of SDF-1/CXCR4 in G-CSF-induced mobilization. In order to exlore a effective index which can predict the effect of mobilization, we observed dynamically the levels of SDF-1 and investigate the correlation between the levels of SDF-1 and total mobilized CD34+ cell numbers to determine if SDF-1 is the index we need.As a angiogenic factor, hepatocyte growth factor (HGF) is also a stromal-derived multi-functional growth factor, and has mitogenic activity for various cell types and other biology function. Vascular endothelial growth factor (VEGF) has been reported to mobilize HSPC and EPC. We postulate that angiogenic factors might be involved in the G-CSF induced mobilization. HGF is produced by human bone marrow stromal cells and promotes proliferation, adhesion and survival of HSPCs. The serum levels of HGF can predict the stage of haematological malignancy. It is not clear about the relation of HGF and mobilization, wealso observed dynamically the levels of HGF and investigate the correlation between the levels of HGF and total mobilized CD34+ cell numbers in order to clearout if HGF was involved in the G-CSF-induced mobilization.Metheds and Materials:Granulocyte-colony stimulating factor(G-CSF) 5ug kg^d""1 twice daily were administered to 24 normal donors 5 or 6 days. The number of WBC and CD34+ cell were determined during mobilization. Plasma samples and serum samples were collected from the peripheral blood during mobilization. Most normal donors underwent 2 apheresis collections from d5 and all patients were transplanted with >4.0xl06 CD34+ cell/kg. 7 patients with acute lymphoblastic leukemia received chemotherapy followed by daily administration of G-CSF 3u￡ kg~1d~1. 8 patients with acute lymphoblastic leukemia received chemotherapy followed by daily administration of G-CSF 5 fig kg"'d"'. on d5 from G-CSF administration, serum samples were collected, the number of CD34+ cell and white blood cell(WBC) by flow cytometric analysis and hematometry respectively. Flow cytometry: Peripheral blood samples and apheresis collections were obtained. Tubes containing l.OxlO6 monunclear cells were placed in PBS and samples were stained withanti-CD34 PE, anti-CD38 FITC, anti-CD45 Pecy5. Isotype-specific antibodies were used as controls. All antibodies and immunoglobulin isotype controls were from Immunotech. Following 30 min of incubation at 4°C in the dark after erythrocytes were lysed. Stained cells were analyzed by flow cytometry. Results were quantitated by Expo32 software. The plasma levels of SDF-1 and serum levels of HGF were dynamically assayed by Enzyme-linked immonosorbent assay (ELISA). ELISA: the antigen to examine has been in lamellan aperture, added samples, washing plate, added Ag-enzyme complex, washing plate again after incubation added substratum, put in darkness to colorate in room temperature, then examined OD values by enzyme immunoassay. The data are analysed by SPSS 10.0 statistical software, aequals 0.05 were considered significant test level.Result:1 >. The numbers of CD34+ cell and white blood cell(WBC) increase significantly during the mobilization process: dl: WBC 5.87±0.57><109/L CD34+ 0.20±0.54xl06/L ; d5: WBC 44.53±7.66xl09/L CD34+ 89.75±15.61xl06/L.2> There was a significant decrease in plasma levels of SDF-1 during the mobilization process(dl:1781±324pg/ml ; d5 : 907±175pg/ml.p=0.0027).3> On dayl and on day5, the SDF-1 levels all showed inverse correlations with total mobilized CD34+ cell numbers(dl: r^—0.32, p=0.0048; d5: r= -0.57, p=0.0061).4 > Serum HGF levels increased significantly with G-CSF administration (dl=421±217pg/ml; d5= 3981±1863 pg/ml. p=0.0001).5 > On day 1, the HGF levels showed an inverse correlation with total mobilized CD34+ cell numbers(r= —0.44, p=0.04), although the HGF on day5 showed no correlation^ = —0.14, p=0.53). Time course kinetics of HGF showed the trend for time dependency was significant (p=0.0067).6> The extend of HGF increase appeared to be more marked as the G-CSF dosage increased. 15 patients with haematological malignancies received different dosage of G-CSF, the difference of HGF was statistically significant(3ug/kg/d : 1085.2±487.4/ml ; 5ug/kg/d : 3981.2±1863.7pg/ml , p=0.009), as was the trend for G-CSF dosage(p=0.0041).Conclusion:1> There was a significant decrease in plasma levels of SDF-1 during the mobilization process. SDF-1 is an important factor in G-CSF-induced mobilization.2> Moreover, on dayl and on day5, the SDF-1 levels all showed inverse correlations with total mobilized CD34+ cell numbers. The disruption of SDF-1/CXCR4 signaling is required for mobilization of HSC. Low plasma levels of SDF-1 appear to characterize normal donors with good mobilization outcome. The levels of SDF-1 before mobilization can predict the effect of mobilization in clinic practice.3 > Serum HGF levels increased significantly with G-CSF administration. Moreover, on dayl, the HGF levels showed an inverse correlation with total mobilized CD34+ cell numbers. Which indicate that HGF may be involved in G-CSF-induced mobilization.4> The serum HGF appeared to be dose-dependent on the G-CSF administration, suggesting that G-CSF might be a good candidate for an as yet undefined physiological HGF inducer.