Alloantigen Presenting Capacity, T Cell Alloreactivity and NK Function of G-CSF-Mobilized Peripheral Blood Cells
Total Page:16
File Type:pdf, Size:1020Kb
Bone Marrow Transplantation, (1998) 22, 631–637 1998 Stockton Press All rights reserved 0268–3369/98 $12.00 http://www.stockton-press.co.uk/bmt Alloantigen presenting capacity, T cell alloreactivity and NK function of G-CSF-mobilized peripheral blood cells D Rondelli1, D Raspadori1, C Anasetti2, G Bandini1,FRe1, M Arpinati1, M Stanzani1, A Morelli3, C Baccini3, A Zaccaria4, RM Lemoli1 and S Tura1 1Institute of Hematology and Medical Oncology ‘Sera`gnoli’, University of Bologna; 3Toxicology and Pharmacology Unit, 4Division of Hematology, ‘S Maria delle Croci’ Hospital, Ravenna, Italy; and 2Division of Clinical Research, Fred Hutchinson Cancer Research Center, Seattle, WA, USA Summary: Hematopoietic stem cells mobilized into the peripheral blood with granulocyte colony-stimulating factor (G-CSF) In this study we addressed whether the proportion and or granulocyte–macrophage colony-stimulating factor the function of antigen presenting cells (APC), T and NK (GM-CSF) have been demonstrated to be as effective as lymphocytes are modified in the apheresis product of six marrow stem cells in reconstituting all blood lineages after healthy donors who received a stem cell mobilizing treat- allogeneic transplantation.1–3 However, leukaphereses col- ment with glycosylated G-CSF at 10 g/kg/day × 5 days lected from normal donors receiving G-CSF contain large s.c. Flow cytometry analysis showed comparable percent- numbers of T cells,4 potentially increasing the risk of acute ages of HLA-DR+,CD19+, CD86+, CD80+ and CD1a+ cells graft-versus-host disease (aGVHD). Nevertheless, previous in preG-CSF-peripheral blood mononuclear cells (preG- reports demonstrated that the incidence and the severity of PBMC) and after mobilization in G-PBMC, whereas the aGVHD after allogeneic transplantation with G-CSF mobil- proportion of CD14+ monocytes significantly increased in ized peripheral blood mononuclear cells (G-PBMC) are not -G-PBMC (3 ؎ 1% vs 17 ؎ 8%, P = 0.003). Analysis of higher than after allogeneic transplantation with steady lymphocyte subsets in preG-PBMC and G-PBMC showed state marrow cells.3,5–7 similar proportions of CD3+, CD4+, CD8+ and CD28+ T Experimental studies addressed whether G-CSF may cells, but a significantly lower percentage of CD16+ modulate alloimmune responses by directly modifying T vs 4 ؎ 1%, P = 0.01), CD56+ (15 ؎ 6% vs cell function or by inducing cell-mediated suppression of 7% ؎ 11) -P = 0.008), CD57+ (16 ؎ 9% vs 5 ؎ 2%, T cell alloreactivity. In particular, it was previously demon ,2% ؎ 5 P = 0.04), CD25+ (19 ؎ 2% vs 9 ؎ 6%, p = 0.009) and strated that aGVHD is more likely due to the activation CD122+ (5 ؎ 2% vs 2 ؎ 1%, P = 0.05) cells in G-PBMC. of type-1 donor cells (secreting IL-2 and IFN-␥)8 and that Unfractionated preG-PBMC and G-PBMC were pretreatment of mice with G-CSF polarizes donor T cells irradiated and tested in primary mixed leukocyte culture toward type-2 cytokine production (IL-4 and IL-10), thus (MLC) with two HLA-incompatible responders and reducing aGVHD.9,10 Mielcarek et al11 showed that G-CSF induced efficient alloresponses in four of six cases, whereas mobilization increases the frequency of CD14+ monocytes G-PBMC stimulated poorly in the remaining two cases. in the blood of normal donors and that a large number of Also, in allo-MLC with irradiated G-PBMC we detected CD14+ cells in the leukapheresis could suppress T cells lower amounts of IFN-␥ (P = 0.04) and of IL-2 (P = 0.06) mainly in a non-contact way, suggesting that this effect than in allo-MLC with preG-PBMC. Furthermore, freshly may contribute to a lower incidence of aGVHD after G- isolated preG-PBMC and G-PBMC from each donor PBMC allogeneic transplantation.11 Also, we recently dem- exerted comparable allogeneic responses to HLA-incom- onstrated that G-PBMC induce a lower allogeneic CD4+ T patible irradiated mononuclear cells in all cases. However, cell response than purified CD34+ blood cells in primary G-PBMC showed no NK activity against K562 target cells MLC.12 at any effector:target ratio tested. These data suggest that Cytolytic cell precursors are present in G-PBMC,13 how- normal G-PBMC may prevent Th1 alloresponses, main- ever previous data showed that G-CSF could downregulate tain efficient alloreactivity to HLA mismatched antigens the NK activity by inhibiting TNF-␣ and IFN-␥ production and have impaired NK activity. by NK cells.14 Nevertheless, Neubauer et al15 reported a Keywords: peripheral blood stem cells; APC; IFN-␥;T higher proportion of CD8+, CD56+ and p75+ cells and a cell; NK activity more efficient IL-2-induced lymphokine-activated killer (LAK) activity in G-PBMC of patients undergoing an auto- logous transplantation than in normal controls. In the present study we analyzed the proportion of anti- gen presenting cells (APC), T and NK cells, as well as the immunological functions of these cell populations in Correspondence: Dr D Rondelli, Institute of Hematology and Medical Oncology ‘Sera`gnoli’, University of Bologna, Policlinico S. Orsola, via the blood of healthy donors before and after treatment Massarenti, 9, 40138 Bologna, Italy with G-CSF. Our results suggest that unfractionated G- Received 4 March 1998; accepted 19 June 1998 PBMC may prevent HLA-incompatible type-1 T cell allo- Alloimmune effects of mobilized PBMC D Rondelli et al 632 responses. In addition, although maintaining the capacity Table 1 Expression of APC markers on mononuclear cells from heal- of responding to alloantigen in MLC, they do not exert thy donors before (PreG-PBMC) and after (G-PBMC) stem cell mobilizing NK activity. treatment with G-CSF PreG-PBMC (%) G-PBMC (%) P value Materials and methods HLA-DR 21 Ϯ 726Ϯ 9NS CD14 3 Ϯ 117Ϯ 8 0.003 Cell preparation CD86 5 Ϯ 15Ϯ 3NS CD80 Ͻ1 Ͻ1 Blood and apheresis samples were obtained with informed CD1a NEG NEG consent from healthy adult peripheral blood stem cell CD19 4 Ϯ 14Ϯ 2NS donors. Blood samples were collected before initiating stem Data represent the mean Ϯ s.d. of six separate cases. cell mobilization with glycosylated G-CSF (lenograstim; = = Ͻ Rhone-Poulenc Rorer, Milan, Italy) at a dose of NS not significant; NEG 0.5%. 10 g/kg/day s.c. × 5–6 days and apheresis samples were obtained from the first apheresis on day 5 of treatment. Blood samples from HLA-DR incompatible healthy vol- ounteers were also collected. All samples were separated 18 h before harvest on day 6. The stimulation index (SI) by centrifugation over Ficoll/Hypaque (Nycomed Pharma, was calculated for each individual experiment as: Oslo, Norway). Light density cells were washed twice in = + RPMI medium enriched with 10% fetal calf serum (FCS) SI c.p.m. (T cell responders stimulators)/c.p.m. (T (Seromed Biochrome, Berlin, Germany). Freshly isolated cell responders). mononuclear cells were then used for cytofluorimetric and functional assays. NK activity Flow cytometry The NK assay was performed using preG-PBMC or G- PBMC and third-party mononuclear cells as effector cells Flow cytometric analysis was carried out on preG-PBMC and the K562 cell line as the target in a 4 h 51Cr release and G-PBMC by direct immunofluorescence. Human assay. Briefly, 1 × 106 target cells were incubated for 1 h monoclonal antibodies (mAbs) used in this study are the at 37°C with 100 Ci Na51Cr, washed twice, mixed in trip- following: CD3 fluorescein isthyocianate (FITC), CD4 licate with effector cells at final effector:target (E:T) ratios FITC, CD8 FITC, CD16 FITC, CD56 phycoerythrin (PE), of 25:1, 12:1 and 6:1 in round-bottomed 96-well plates for CD57 FITC, CD80 (anti-B7–1) PE, CD25 (anti-p55 IL-2 4 h at 37°C, and then centrifuged at 400 g for 10 min. An receptor) FITC, CD28 PE, CD14 FITC and CD19 FITC aliquot (100 l) of the supernatant was collected and coun- (Becton Dickinson, San Jose´, CA, USA); CD1a PE and ted in a gamma scintillation counter. The percentage CD86 (anti-B7–2) PE (PharMingen, San Diego, CA, USA); of 51Cr release was expressed as: CD122 (anti-p75 IL-2 receptor) FITC (Janssen, Brussels, Belgium). Isotype controls IgG1 FITC and PE and IgG2a % lysis = (E − S)/(M − S) × 100 PE are from Becton Dickinson and IgG2b PE is from PharMingen. Data acquisition and analysis were assessed (where E = mean c.p.m. release in the presence of effector by Lysis II software (Becton Dickinson). cells; S = mean c.p.m. spontaneously released by target cells incubated with medium alone; M = mean c.p.m. of resuspended target cells). Primary MLC Freshly isolated preG-PBMC or G-PBMC were either irradiated with 3000 cGy and used as stimulators, or used Cytokine assays as responders in primary allogeneic mixed lymphocyte cul- tures (MLC). Each donor preG- or G-PBMC were tested Levels of IL-2, IFN-␥, IL-4 and IL-10 in MLC supernatants with the same two third-party mononuclear cells from two were determined by commercially available enzyme-linked HLA incompatible healthy volounteers. In selected experi- immunosorbent assay (ELISA) kits (IL-2: Immunotech, ments over 90% T cells were isolated from preG- and G- Marseille, France; IFN-␥, IL-4 and IL-10: Endogen, Bos- PBMC and from third party mononuclear cells by rosetting ton, MA, USA) and a standard curve was generated to with sheep red blood cells, as previously described.16 determine the cytokine concentration in the sample. Detec- Irradiated stimulator and responder cells were resus- tion limits of the assays were 5 pg/ml for IL-2, Ͻ2 pg/ml pended in medium containing RPMI-1640, 25 mm Hepes, for IFN-␥, Ͻ2 pg/ml for IL-4 and Ͻ3 pg/ml for IL-10.