Increased plasma EPO and MIP-1α are associated with recruitment of vascular progenitors but not CD34(+) cells in autologous peripheral blood stem cell grafts

Results Thirty-six patients (mean age = 51 years, 42% female) had plasma collected at baseline prior to PBSC mobilization and on the day of PBSC collection. Only erythropoietin (EPO) levels increased significantly on the day of PBSC collection in comparison with baseline plasma levels (2.2-fold increase; p = 0.003). Interleukin-2, -10, epidermal growth factor, interferon-α, and angiopoietin-1 all decreased significantly between baseline and the day of PBSC collection ( p < 0.02). The remaining cytokine levels did not change appreciably ( p = NS). The cohort was divided into “low” graft VPCs (<2.0 × 10 3 /kg) and “high” graft VPCs (≥2.0 × 10 3 /kg) and cytokine levels were compared between the groups. At baseline, increased levels of macrophage inflammatory protein-1α (MIP-1α) were associated with increased graft VPCs ( p = 0.05) while higher EPO concentrations on the day of PBSC collection predicted higher graft VPC levels ( p = 0.02). These cytokines were not associated with CD34(+) cell mobilization. Conclusions The association of different plasma proteins with graft VPC and CD34(+)-cell levels suggests that mobilization of vascular and hematopoietic progenitors occurs through independent mechanisms. Patients with low levels of MIP-1α at baseline may be candidates for interventions aimed at increasing graft VPC levels. Strategies that increase plasma EPO concentrations may be most promising to augment the regenerative properties of PBSC products. Autologous hematopoietic stem cell transplantation (HSCT) represents a common treatment option for patients with blood-related cancers, such as multiple myeloma and lymphomas [1,2] . Peripheral blood stem cells (PBSCs) are collected in advance of transplantation and reinfused after high-dose myeloablative chemotherapy and/or radiation. High-dose treatment is associated with potential toxicity in multiple organs and is typically restricted to younger patients without serious comorbidity. In addition to hematopoietic precursors responsible for engraftment and immune reconstitution, PBSC grafts contain endothelial-like vascular progenitor cells (VPCs), which may facilitate recovery from treatment-related tissue injury. VPCs are angiogenic cells derived from bone marrow [3] and can be increased in the peripheral blood in response to tissue injury [4] and various cytokines, growth factors, and medications [5–8] . Our previous work suggests that PBSCs with higher VPC levels are associated with reduced transplant-related toxicity [9] . Optimizing conditions that increase VPC levels in PBPC grafts could reduce the toxicity associated with high-dose treatment and make transplantation applicable to older patients and for those with comorbidities. In this study, we investigated known inflammatory, hematopoietic, and angiogenic plasma proteins in HSCT patients undergoing autologous PBSC mobilization to identify potential signals associated with improved VPC mobilization. We hypothesized that concentrations of key plasma proteins at baseline or on the date of PBSC collection would correlate with graft VPC levels and may yield potential targets for development of strategies to improve the regenerative capacity of PBSCs. Materials and methods Patients This protocol was approved by our institutional ethics review board and participating patients provided informed consent allowing the use of clinical and laboratory data for research purposes. Peripheral blood plasma samples were collected from patients (n = 36) undergoing autologous HSCT at The Ottawa Hospital between November 2006 and August 2007. In addition, five plasma samples from healthy allogeneic donors were analyzed and served as controls. Plasma collection Plasma samples (7 mL in ethylene diamine tetraacetic acid) were obtained prior to mobilization treatment (baseline) and on the day of PBSC collection. Plasma was obtained following immediate centrifugation at 1500 g , aliquotted and frozen at −80°C until tested. Plasma analyses Plasma samples at baseline and from the day of PBSC collection were analyzed to determine the concentration of 16 known inflammatory and angiogenic markers, including hepatocyte growth factor (Human Serum Adipokine Panel B Linco plex ; Millipore, St-Charles, MO, USA), monocyte chemoattractant protein-3, basic fibroblast growth factor, platelet derived growth factor-β (using Human Multiplex-Cytokine Linco plex kit ; Millipore), vascular endothelial growth factor, interleukin-1 (IL-1) α/β, IL-2, IL-6, IL-10, granulocyte macrophage-colony stimulating factor (GM-CSF), epidermal growth factor, tumor necrosis factor-α (TNF-α), macrophage inflammatory protein-1 (MIP-1) α/β, interferon-γ (using Human Cytokine Linco plex kit; Millipore). Assay kits were read using the Bio-Rad- Bio-Plex suspension array system-open xMAP technology platform (Bio-Rad Life Science, Hercules, CA, USA). Additionally, concentrations of angiopoietin 1, angiopoietin 2, and erythropoietin (EPO) were, measured using enzyme-linked immunosorbent assay kits (R&D Systems, Minneapolis, MN, USA). These three cytokines were unavailable using Luminex technology. Optical densities were quantified using an MRX plate reader (Dynatec Laboratories, El Paso, TX, USA). Each sample was analyzed in duplicate and procedures were performed in accordance with the manufacturer's instructions. VPC enumeration VPC levels were enumerated using the mononuclear cell fraction of PBPC collections using a method adapted from Hill et al. [10] . VPC clusters were grown in culture after plating mononuclear cells on fibronectin coated plates under serum-rich angiogenic media conditions (Endocult Liquid Medium Kit, StemCell Technologies, Vancouver, BC, Canada). Nonadherent cells were collected 48 hours after initial plating where mature adherent endothelial and monocytic cells were removed, followed by replating. VPC clusters were enumerated 72 hours after nonadherent cell removal using an inverted microscope and an early outgrowth VPC cell cluster was defined as a central core of at least 30 “round cells” with at least three attached emanating cells [10] . Clinical parameters of transplant-related toxicity Clinical data were retrieved from The Ottawa Hospital bone marrow transplantation clinical database for all patients undergoing HSCT. Transplant-related toxicity was determined using the Seattle criteria for organ toxicity [11] . Graft CD34 + levels were measured by flow cytometry following PBSC collection according to standard methods [12] (fluorescein isothiocyanate−conjugated CD34; Stemkit, Beckman Coulter, Fullerton, CA USA). Statistical methods Comparison of means was performed using two-tailed Student's t-tests, assuming unequal variance. A p value <0.05 was considered to be statistically significant. Statistical analyses were performed using Microsoft Excel software. Mean values are reported ± standard error of the mean. Results Patients Mean age of our patients was 51 years (range, 20−69 years) and our population was 42% female. See Table 1 for a summary of patient characteristics. A total of 30 patients underwent autologous hematopoietic transplantation. A total six patients did not undergo transplantation due to insufficient collection of CD34 + cells and/or progressive disease. Changes in plasma factor concentrations from baseline to collection We measured mean plasma factor concentrations at baseline and on the date of PBSC collection ( Fig. 1 ) and calculated the fold change in analyte concentrations between these two time points. Concentrations for three analytes (IL-1β, basic fibroblast growth factor, and monocyte chemoattractant protein-3) were too low for accurate analysis and these data were omitted from further analysis. Five cytokines demonstrated significant reductions between baseline and the first day of PBSC collection (see Fig. 2 ). Epidermal growth factor (2.6-fold decrease; p = 0.002), IL-2 (2.1-fold decrease; p = 0.003), IL-10 (1.7-fold decrease; p = 0.01), interferon-γ (2.4-fold decrease; p = 0.001), and angiopoietin 1 (2.8-fold decrease; p = 0.001) all significantly decreased during the PBSC mobilization period. Only plasma EPO levels increased significantly (2.2-fold increase; p = 0.003) between baseline and collection day (see Fig. 2 ). The remaining seven cytokine levels did not change significantly. Plasma factors associated with high graft VPClevels and optimal levels of CD 34 + cells VPC clusters were enumerated in a subset of patients (n = 22) using a standard cell culture assay and identified as described by Hill et al. [10] . A threshold value of 2.0 × 10 3 VPCs/kg was used to separate patients into high (n = 11) or low (n = 11) graft VPC levels based on our previous observations that patients with >2.0 × 10 3 /kg experienced reduced transplant-related toxicity [9] . Mean plasma analyte concentrations were compared between the two groups using baseline plasma samples. Baseline MIP-1α concentrations were 1.8-fold greater ( p = 0.05) in the high VPC group (see Table 2 ). All other analytes tested were not significantly different. Plasma concentrations of cytokine proteins were also measured using samples from the day of first PBSC collection. Collection-day plasma EPO concentrations were 2.9-fold increased ( p = 0.02) in patients with high graft VPC levels ( Table 2 ). Interestingly, only two patients (9%) demonstrated a reduction in EPO levels on collection and both had low graft VPC levels. The remaining cytokines measured were not significantly different between the two groups. Using the mean and median graft VPC levels to divide the cohort into “high” and “low” graft VPCs did not produce any significant change in our observations. Mean plasma analyte concentrations associated with optimal graft CD34 + mobilization ( > 5 × 10 6 CD34 + /kg) were compared with plasma samples in patients with reduced CD34 + levels ( < 5 × 10 6 CD34 + /kg). None of the cytokine concentrations from baseline and day of PBSC collection were associated with graft CD34(+) cell levels. Plasma factors associated with low organ toxicity Organ toxicity data was available for all patients undergoing HSCT (n = 30). Patients were divided into two groups based on “high” (at least one organ system with a score of ≥ 2, n = 16) or “low” transplant-related toxicity (all organ systems having less than grade 2 toxicity, n = 14). No analyte was significantly different in either group when baseline plasma samples were analyzed. At date of PBSC collection, plasma TNF-α concentrations were 1.6-fold increased ( p = 0.03) in the group with high transplant toxicity. Discussion In this study, we characterized a pattern of angiogenic, hematopoietic, and inflammatory protein alterations in the plasma of patients undergoing PBSC collection. In particular, we observed an important increase in EPO concentrations between baseline and the day of PBSC collection. Moreover, patients with increased levels of graft VPCs had higher EPO levels on the day of PBSC collection. Our previous work has highlighted the importance of graft VPC levels in reducing organ toxicity following autologous HSCT [9] and we have reported an association between reduced hemoglobin on the day of PBSC collection and increased transplant-related toxicity [13] . Further, lower levels of graft VPCs are observed in patients with anemia on the day of PBSC collection [13] . Taken together, our work suggests an important role for EPO in the mobilization of VPCs in PBSC grafts. Chemotherapy and cytokine stimulation with GCSF appears to alter angiogenic and inflammatory signals in patients undergoing PBSC mobilization. In particular, we observed that certain cytokine factors were associated with VPC mobilization, but not CD34 + cell mobilization. We do not have sufficient data to comment on cytokine changes that occur with poor CD34 + cell mobilization and this will require additional study. Elevated EPO concentrations at time of PBSC collection correlated with high graft VPC levels, but not graft CD34 content in our study. EPO stimulates the proliferation and differentiation of erythroid progenitors as well as increasing the number of circulating VPCs [14,15] . EPO increases VPC mobilization in peripheral blood and also increases the number of functionally active VPCs [15] . These findings were recently confirmed in a cohort of patients with coronary heart disease, where EPO serum levels were significantly associated with the number and function of circulating VPCs [6] . In that study, Heeschen et al. [6] proposed that reduced EPO serum levels may help identify patients with impaired VPC recruitment capacity. Our observation that increased plasma EPO levels are correlated with high graft VPCs is consistent with the notion that EPO stimulates postnatal neovascularization, in part, by enhancing functional VPC mobilization from the bone marrow. The possibility that EPO could act as a biomarker for the collection of increased numbers of VPCs in PBSC grafts is intriguing and introduces the possibility that exogenous EPO or interventions that increase endogenous EPO levels may increase graft VPC levels [16] . EPO is likely mobilized in response to tissue injury caused by the chemotherapy regimen. Circulating EPO can stimulate nitric oxide synthase in regions of ischemic tissue damage [17] . Local increases in nitric oxide have been reported to increase the number of circulating VPCs [14] . Autologous PBSC grafts with increased VPC levels may have greater capacity for vascular repair and have been associated with reduced toxicity after HSCT [9] . A prospective randomized trial of PBSC mobilization using EPO and GCSF and compared with GCSF alone appears warranted. Baseline MIP-1α concentrations correlated well with high graft levels of VPCs, but were not associated with graft CD34 + cell levels. Our observations suggest that VPC mobilization and hematopoietic stem cell mobilization may occur through independent signaling pathways. In addition to the possible utility of measuring MIP-1α concentrations at baseline to predict graft VPC content, MIP-1α is a chemokine involved in many proinflammatory activities, such as leukocyte chemotaxis that may provide insight regarding the biology of VPC mobilization [18] . MIP-1α has been associated with inhibition of immature erythroid progenitor cell proliferation [19] and, in this regard, may allow EPO to act preferentially to mobilize VPCs in place of erythropoietic stimulation. Additional work is needed to understand potential interactions between MIP-1α and EPO. We recognize that different mobilization protocols involving different chemotherapy agents may influence the collection of graft VPCs and this will require further study in larger more homogeneous populations. We also acknowledge that various methods have been reported for enumerating VPCs. Although VPC populations have been identified using flow cytometric analysis, the precise immunophenotype remains under study. The VPC cluster assay was, therefore, used in the present study because we have previously reported a correlation between VPC clusters in PBPC grafts and toxicity after transplantation [9] . In our cohort of patients, only TNF-α concentrations on the day of PBSC collection were associated with increased clinical toxicity after autologous HSCT. This finding is consistent with previous observations. Holler et al. [20] reported a significant association between elevated TNF-α serum levels in patients with severe transplant-related complications occurring within the first 6 months following allogeneic HSCT. Additionally, increased TNF-α levels prior to HSCT proved to be predictive for the development of transplant related toxicity [21] . TNF-α is largely produced by monocytes/macrophages and has been shown to be one of the predominant cytokines modulating endothelial function [22] . It belongs to a large family of proteins that have beneficial roles, such as inflammatory and protective immune responses, as well as being involved in potentially damaging effects, such as host damage in sepsis and autoimmune diseases [23] . Overall, the protein signals identified in this study suggest that mobilization of vascular and hematopoietic progenitors may occur through independent mechanisms. Further, patients with reduced levels of MIP-1α at baseline may be candidates for interventions aimed at increasing graft VPC levels. Strategies that increase plasma EPO levels would seem most promising for future studies. Our observations require further validation in larger studies to improve our understanding of the mechanisms of VPC mobilization. Acknowledgments This research was supported by the Bayer Partnership Fund and by the BMT Research and Education Fund of The Ottawa Hospital (TOH) Foundation (Ottawa, Canada). We would like to acknowledge S. McDiarmid and Dr. L. Heubsch for assistance in retrieving data from the BMT database at TOH. We thank Linda Hamelin from the BMT program at TOH and nurses affiliated with Canadian Blood Services (CBS) (Ottawa, Canada) for procurement of patient samples. We thank the physicians and health care workers in the Blood and Marrow Transplant Program at TOH for their care of the patients described in this study. Further we are grateful for helpful discussions with Dr. Alp Oran (University of Ottawa). D.S.A. is an adjunct scientist with CBS. L.L. and A.G. were recipients of summer internships with CBS. References 1 W.I. Bensinger Hematopoietic cell transplantation for multiple myeloma Cancer Control 5 1998 235 242 2 C. Martinez O. Salamero L. Arenillas Autologous stem cell transplantation for patients with active Hodgkin's lymphoma: long-term outcome of 61 patients from a single institution Leukemia Lymphoma 48 2007 1968 1975 3 T. Asahara T. Murohara A. Sullivan Isolation of putative progenitor endothelial cells for angiogenesis Science 275 1997 964 967 4 M. Gill S. Dias K. Hattori Vascular trauma induces rapid but transient mobilization of VEGFR2(+)AC133(+) endothelial precursor cells Circ Res 88 2001 167 174 5 H.J. Cho H.S. Kim M.M. Lee Mobilized endothelial progenitor cells by granulocyte-macrophage colony-stimulating factor accelerate reendothelialization and reduce vascular inflammation after intravascular radiation Circulation 108 2003 2918 2925 6 C. Heeschen A. Aicher R. Lehmann Erythropoietin is a potent physiologic stimulus for endothelial progenitor cell mobilization Blood 102 2003 1340 1346 7 M. Vasa S. Fichtlscherer K. Adler Increase in circulating endothelial progenitor cells by statin therapy in patients with stable coronary artery disease Circulation 103 2001 2885 2890 8 D.S. Allan P. Dubé J. Roy L. Busque D.C. Roy Endothelial-like vascular progenitor cells (VPCs) from allogeneic and autologous donors: mobilization features distinct from hematopoietic progenitors Biol Blood Marrow Transplant 13 2007 433 439 9 T. Iqbal T. Rajagopal M. Halpenny Increased graft content of vascular progenitor cells is associated with reduced toxicity following autologous hematopoietic transplantation Exp Hematol 36 2008 506 512 10 J.M. Hill G. Zalos J.P.J. Halcox Circulating endothelial progenitor cells, vascular function, and cardiovascular risk N Engl J Med 348 2003 593 600 11 S.I. Bearman F.R. Appelbaum D. Buckner Regimen-related toxicity in patients undergoing bone marrow transplantation J Clin Oncol 6 1988 1562 1568 12 M. Keeney I. Chin-Yee K. Weir Single platform flow cytometric absolute CD34+ cell counts based on the ISHAGE guidelines. International Society of Hematotherapy and Graft Engineering Cytometry 34 1998 61 70 13 G. Kasbia F. Al-Gahtani J. Tay Reduced hemoglobin on day of peripheral blood progenitor cell collection is associated with low graft content of vascular progenitors and increased toxicity after autologous hematopoietic stem cell transplantation Transfusion 48 2008 2421 2428 14 N. Urao M. Okigaki H. Yamada erythropoietin-mobilized endothelial progenitors enhance reendothelialization via Akt-endothelial nitric oxide synthase activation and prevent neointimal hyperplasia Circ Res 98 2006 1405 1413 15 F.H. Bahlmann K. de Groot J.M. Spandau Erythropoietin regulates endothelial progenitor cells Blood 103 2004 921 926 16 L. Labonté T. Iqbal S. McDiarmid Erythropoietin during PBPC collection: can we reduce toxicity of autologous transplants in multiple myeloma Biol Blood Marrow Transplant 14 2008 132 133 17 D. Burger M. Lei N. Geoghegan-Morphet X. Lu A. Xenocostas Q. Feng Erythropoietin protects cardiomycocytes from apoptosis via up-regulation of endothelial nitric oxide synthase Cardiovasc Res 72 2006 51 59 18 D.N. Cook The role of MIP-1α in inflammation and hematopoiesis J Leukocyte Biol 59 1996 61 66 19 S. Su N. Mukaida J. Wang Inhibition of immature erythroid progenitor cell proliferation by macrophage inflammatory protein-1 alpha by interacting mainly with a C-C chemokine receptor, CCR1 Blood 90 1997 605 611 20 E. Holler H.J. Kolb A. Moller Increased serum levels of tumor necrosis factor alpha precede major complications of bone marrow transplantation Blood 75 1990 1011 1016 21 J.A. Bianco F.R. Appelbaum J. Nemunaitis Phase I–II trial of pentoxifylline for the prevention of transplant-related toxicities following bone marrow transplantation Blood 78 1991 1205 1211 22 J.S. Pober Effects of tumour necrosis factor and related cytokines on vascular endothelial cells Ciba Found Symp 131 1987 170 184 23 T. Hehlgans K. Pfeffer The intriguing biology of the tumour necrosis factor/tumour necrosis factor receptor superfamily: players, rules and the games Immunology 115 2005 1 20