Proteoglycans on bone marrow endothelial cells bind and present SDF-1 towards hematopoietic progenitor cells.

Recognition events between hematopoietic progenitor cells (HPC) and bone marrow endothelial cells (BMEC) initiate homing of HPC to the bone marrow. The chemokine SDF-1 is present on BMEC and plays a crucial role in bone marrow engraftment. We studied the role of proteoglycans (PGs) on BMEC in binding and presentation of SDF-1. SDF-1 mRNA was present in three human BMEC cell lines. Competition experiments showed that 125I-SDF-1 alpha binding to the BMEC cell line 4LHBMEC was inhibited by heparins, heparan sulfate (HS) intestinal mucosa, chondroitin and dermatan sulfate (CS/DS), but not by HS bovine kidney. Pretreatment of 4LHBMEC with glycosaminoglycan (GAG)-degrading enzymes or sodium chlorate demonstrated that SDF-1 bound to both HSPGs and CS/DSPGs in a sulfation-dependent manner, as determined with an SDF-1 antibody recognizing the CXCR4-binding site. 4LHBMEC bound four-fold more SDF-1 than HUVEC. Isolated endothelial PGs did not bind SDF-1 in a filter or microplate-binding assay, suggesting the necessity of membrane association. In flow adhesion experiments, endothelial arrest of CXCR4+ KG-1 and not of CXCR4- KG-1a cells increased significantly when SDF-1 was presented on 4LHBMEC. In conclusion, SDF-1 is produced by BMEC and binds to the BMEC cell surface via HS and CS/DS-GAGs, thereby presenting its CXCR4 binding site to HPC contributing to their arrest.

Extensive early apoptosis in frozen-thawed CD34-positive stem cells decreases threshold doses for haematological recovery after autologous peripheral blood progenitor cell transplantation.

Stem cell doses necessary for engraftment after myelo-ablative therapy as defined for fresh transplants vary largely. Loss of CD34+ cell quality after cryopreservation might contribute to this variation. With a new early apoptosis assay including the vital stain Syto16, together with the permeability marker 7-AAD, CD34+ cell viability in leucapheresis samples of 49 lymphoma patients receiving a BEAM regimen was analysed. After freeze-thawing large numbers of non-viable, early apoptotic cells appeared, leading to only 42% viability compared to 72% using 7-AAD only. Based on this Syto16 staining in the frozen-thawed grafts, threshold numbers for adequate haematological recovery of 2.8-3.0 x 10(6) CD34+ cells/kg body weight determined for fresh grafts, now decreased to 1.2-1.3 x 10(6) CD34+ cells/kg. In whole blood transplantation of lymphoma patients (n = 45) receiving a BEAM-like regimen, low doses of CD34+ cells were sufficient for recovery (0.3-0.4 x 10(6)CD34+ cells/kg). In contrast to freeze-thawing of leucapheresis material, a high viability of CD34+ cells was preserved during storage for 3 days at 4 degrees C, leaving threshold doses for recovery unchanged. In conclusion, the Syto16 assay reveals the presence of many more non-functional stem cells in frozen-thawed transplants than presumed thus far. This led to a factor 2.3-fold adjustment downward of viable CD34+ threshold doses for haematological recovery.

Differences in sulfation patterns of heparan sulfate derived from human bone marrow and umbilical vein endothelial cells.

OBJECTIVE: Heparan sulfates (HS), the polysaccharide side chains of HS proteoglycans, differ in structure and composition of sulfated domains among various tissue types, resulting in selective protein binding. HS proteoglycans on bone marrow endothelial cells (BMEC) could contribute to tissue specificity of the bone marrow endothelium and play a role in the presentation of chemokines such as stromal cell-derived factor-1 (SDF-1) and adhesion of hematopoietic progenitor cells after stem cell transplantations. We characterized differences in HS structure and SDF-1 binding between BMEC and human umbilical vein endothelial cells (HUVEC). MATERIALS AND METHODS: Expression of HS proteoglycans on human bone marrow microvessels was investigated by immunohistochemical staining. Comparison of three human BMEC cell lines with HUVEC and an HUVEC cell line was studied by flow cytometry using antibodies against different epitopes of the HS polysaccharide chain. HS proteoglycans were biochemically characterized after isolation from metabolically labeled cultures of the BMEC cell line 4LHBMEC and HUVEC. Binding of radiolabeled SDF-1 to 4LHBMEC and HUVEC and competition with heparins were investigated. RESULTS: Bone marrow microvessels constitutively expressed HS proteoglycans. Flow cytometric experiments showed differences in HS chain composition between BMEC and HUVEC. Biochemical characterization revealed more O-sulfation of the N-sulfated domains present in cell-associated HS glycosaminoglycans in 4LHBMEC compared to HUVEC. Binding experiments showed that 4LHBMEC bound more 125[I]-SDF-1 per cell than HUVEC. This could be inhibited largely by heparin and O-sulfated heparin and to a lesser extent by N-sulfated heparin. CONCLUSIONS: Cellular HS from BMEC differs in composition from HUVEC. We postulate that the presence of highly sulfated domains in the HS chains from BMEC contributes to tissue specificity of bone marrow endothelium in which HS may be involved in SDF-1 presentation and adhesion of hematopoietic progenitor cells.

Primary human keratinocytes as targets in predicting acute graft-versus-host disease following HLA-identical bone marrow transplantation.

Graft-versus-host-disease (GVHD) remains a major problem following allogeneic bone marrow transplantation (BMT) and manifests itself mainly by damage to epithelial cells of the skin, gut and bile ducts. Reliable tests to predict GVHD are lacking. We developed an assay in which donor T cells are stimulated by patient keratinocytes (KCs), compared that with stimulation by patient peripheral blood mononuclear cells (PBMCs) and studied the relationship to GVHD. In 27 patients undergoing HLA-identical BMT for haematological malignancies, donor T-cell reactivity was determined as the helper T-lymphocyte precursor (HTLp) frequency against host PBMCs (25 patient-donor pairs) and host KCs (20 patient-donor pairs). KCs were obtained by shave biopsies and cultured with interferon (IFN)-gamma to induce HLA class II expression. In assays using patient KCs and donor T cells, anti-CD28 antibody was added to compensate for the lack of co-stimulatory molecules on KCs. Results were related to the occurrence of GVHD. As BMTs were performed with partially T cell-depleted grafts, GVHD was limited to grade 0 (five patients), grade I (seven patients) and grade II (12 patients). No differences were found in donor T-cell reactivity to patient PBMCs, as expressed as HTLp frequency in patients with or without GVHD. However, significant differences (P < 0.01) were found in donor T-cell reactivity to patient KCs when comparing patients with and without GVHD. Donor HTLp frequencies against patient KCs give a better prediction of GVHD than those against patient haemopoietic cells following HLA-identical BMT, which may indicate that at least some minor non-HLA histocompatibility antigens present on KCs are different from those on haemopoietic cells.

Determination of helper T-cell precursor frequencies against non-haemopoietic cells: comparison of co-stimulation provided by anti-CD28 antibody versus the cellular ligand B7-1.

Helper T-cell precursor frequency assays (HTLp-assays) are commonly used in transplantation to examine the frequency of T cells reactive against donor or host alloantigens. In these assays, peripheral blood mononuclear cells (PBMCs) are most often used as stimulator cells. However, cells targeted after transplantation do not always belong to the haematopoietic lineage and may express different alloantigens, especially minor histocompatibility antigens (mHags). Non-haematopoietic cells lack expression of the B7 co-stimulatory molecules needed to activate primary T cells that can be supplied by anti-CD28 (alphaCD28) antibodies or transfection with B7-1 coding sequences. At present, it is not known how these two ways of supplied co-stimulation compare in HTLp assays. B7-1-transfected A431 keratinocytes (A431B7-1) induced higher proliferative responses in allogeneic primary T cells and more interleukin (IL) 2 production than that induced by A431 cells plus alphaCD28, whereas the kinetics of proliferation and IL-2 production were similar. Neither cross-linking of alphaCD28 bound to T cells nor prevention of IL-2 resorption by the anti-IL-2 receptor resulted in improved proliferation or IL-2 production. Results of HTLp assays indicated that A431B7-1 activated on average 7.5 times more alloreactive IL-2-producing T cells than A431 cells plus alphaCD28. We conclude that primary T-cell alloresponses against major histocompatibility complexes (MHCs) and mHags expressed on non-haematopoietic cells can be measured in HTLp assays using supplied co-stimulation, although alphaCD28 yields an intrinsic underestimation of actual frequencies.

Selective depletion of major and minor histocompatibility antigen reactive T cells: towards prevention of acute graft-versus-host disease.

Development of acute graft-versus-host disease (aGVHD) following HLA-identical sibling bone marrow transplantation (BMT) remains a serious complication. A selective depletion of T cells has proved to be effective in preventing aGVHD but is associated with relapse and increased incidence of infection. As aGVHD is directed mainly against epithelial tissues we examined whether it would be feasible to selectively deplete T cells reactive with epithelial cells whilst preserving other specificities. Donor T cells which express HLA-DR, CD25, CD69 and CD71 activation markers after cocultivation with patient keratinocytes were depleted using magnetic cell separation techniques. Depletion of major as well as minor histocompatibility antigen activated T cells revealed a significant (P = 0.004 and P = 0.031, respectively) 10-fold decrease in the frequency of donor T lymphocyte precursors reactive with patient keratinocytes. The frequency reactive with third-party and patient peripheral blood mononuclear cells, including leukaemia cells, remained unchanged, supporting the notion that aGVHD and graft-versus-leukaemia (GVL) may be separable. This alloantigen-specific depletion may be used in matched unrelated as well as HLA-identical sibling BMT for reducing aGVHD whilst conserving GVL.

Human B7-1 is more efficient than B7-2 in providing co-stimulation for alloantigen-specific T cells.

Besides a signal via the T cell receptor/CD3 complex, an additional costimulatory signal is required for optimal T cell activation. This signal can be delivered by interaction of either B7-1 or B7-2 expressed by antigen-presenting cells with CD28 on the T cells. Comparison of the function of B7-1 and B7-2 in different experimental animal systems generated conflicting data on the roles for the co-stimulatory molecules. We therefore investigated whether there are differences between B7-1 and B7-2-mediated co-stimulation in an alloantigen-specific primary T cell response induced by B7-transfected human cell lines of epithelial origin. Both transfected keratinocyte cell lines efficiently induce T cell proliferation and the ratios of stimulator versus responder cells are similar. The kinetics of proliferation and interleukin (IL)-2, IL-4 and interferon-gamma production are also comparable between both transfectant lines. However, despite equal B7 expression levels, it is consistently found that the magnitude of the B7-1-induced T cell proliferation was higher than that of B7-2. Comparison of precursor frequencies of helper T lymphocytes responsive with either B7-1 or B7-2 revealed that the frequency of B7-1-responsive T cells was higher than that of B7-2, and that the frequency of cells activated by a combination of B7-1 and B7-2 did not differ significantly from that of B7-1 alone. We therefore conclude that the B7-2-responsive T cells are part of the B7-1-responsive population, and that B7-1 on keratinocytes is more efficient in providing co-stimulation for alloantigen-specific T cells.