Transplantation (2000) 26, 371–376  2000 Macmillan Publishers Ltd All rights reserved 0268–3369/00 $15.00 www.nature.com/bmt Dendritic cells can be successfully generated from CD34؉ cord cells in the presence of autologous cord blood plasma

FE Borra`s1, NC Matthews1, R Patel1 and C Navarrete1,2

1Department of Histocompatibility and Immunogenetics, NBS, London and The South East, North London Centre, London; and 2Department of Immunology, Royal Free and University College Medical School, London, UK

Summary: gens to T cells.3 Mature DCs have several morphological and phenotypical properties which make them potent APCs. Dendritic cells (DCs) are currently being considered as Dendritic cells express high levels of HLA class I and II adjuvants in . Depending on their (especially DR), and co-stimulatory molecules such as source and culture conditions, they show different fea- CD40, CD80 and CD86 which bind to their counter- tures and maturation states. Dendritic cells can be gen- parts, CD40L, CD28 and CTLA-4 respectively.4 Myeloid ؉ erated from monocytes and CD34 haematopoietic stem DCs activated by different stimuli such as CD40L, produce cells, from both adult and cord blood. Here, we report and secrete IL-12, a potent T cell stimulatory cytokine ؉ the generation of mature DCs from enriched CD34 which induces the production of IFN␥ by T cells and NK cord blood (CB) cells using autologous cord blood cells.5,6 plasma (ACBP) as a source of serum and fac- ؉ The unique ability of DCs to prime naive T cells and to tors. In the presence of ACBP, CD34 cells proliferated present antigens effectively has led to their use as ‘natural and differentiated resulting in a population of cells with adjuvants’ in experimental tumour immunotherapy.7 At a dendritic phenotype as assessed by morphology and present, efforts are being made to generate DCs ex vivo flow cytometry analyses. The DC population obtained from different sources. Dendritic cells can be obtained from using ACBP showed higher levels of HLA class II mol- peripheral blood monocytes primed by GM-CSF and IL-4 ecules, co-stimulatory molecules including CD40, CD80 to differentiate into DCs8–10 and from CD34ϩ progenitor or CD86, and the marker CD83, com- cells from either bone marrow or cord blood, requiring pared with those generated in adult blood serum (ABS). GM-CSF and TNF␣ for their differentiation.11,12 Other Furthermore, the DCs generated in the presence of cytokines including SCF, Flt3-L and TGF␤1 have also been ACBP were more potent stimulatory cells in the mixed lymphocyte:dendritic cell reactions (MLDCR), com- extensively used. factor increases the number of initial pluripotent stem cells in CD34ϩ-derived DC cul- pared to cells generated in ABS. Similar results were 13 obtained using homologous cord blood plasma (HCBP). tures, whereas Flt3-L increases the number of DCs in lymphoid and non-lymphoid tissues in mice14 and promotes These results show that ACBP can support the gener- ϩ ation of DCs from CD34؉ progenitor cells when only an extensive amplification of human CD34 progenitor 15 ␤ GM-CSF and TNF␣ are used as differentiating cyto- cells in culture. TGF 1, in the presence of GM-CSF and kines. Bone Marrow Transplantation (2000) 26, 371–376. IL-4, induces differentiation of monocytes to Langerhans 16 Keywords: dendritic cells; cord blood; autologous cells. IL-4 has also been used to inhibit the monocyte plasma; CD34ϩ differentiation pathway and to favour the DC pathway when CD34ϩ cells are used.8,17 In addition to the cytokines, the source of serum used to obtain differentiated DCs in vitro is also important. Experi- Dendritic cells (DCs) play a crucial role in the induction ments to date have used different sources of serum proteins of the immune response. These cells, which are present in and factors including human adult blood serum (ABS), low numbers all over the body, act as professional antigen human plasma and fetal calf serum (FCS). We have studied presenting cells (APC) and have the unique feature of being and compared the ability of autologous cord blood plasma 1 able to prime naive T cells. (ACBP) to promote the differentiation of DCs from cord Antigen presentation and T cell activation are features blood enriched CD34ϩ cells. Dendritic cells differentiated of mature DCs, whilst immature DCs are more specialised in ACBP or ABS cultures were tested for the expression 2,3 in capturing and processing antigens. After capturing of HLA class II (DR and DQ) co-stimulatory molecules antigen, these immature cells differentiate and migrate to (CD40, CD80 and CD86) and the CD83 and CD1a DC the lymphoid tissues where they present the processed anti- markers. The allostimulatory capacity of cord blood CD34ϩ derived DCs was also assessed using the mixed lympho- Correspondence: Dr C Navarrete, Department of Histocompatibility and cyte:dendritic cell reaction (MLDCR). Immunogenetics, NBS, North London Centre, Colindale Avenue, London, NW9 5BG, UK Received 31 January 2000; accepted 14 May 2000 Generation of cord blood DCs with autologous plasma FE Borra`s et al 372 Materials and methods used: CD1a (Ortho Diagnostic Systems, Raritan, NJ, USA), CD40, CD83 (Immunotech, Marseille, France) CD14, Cell culture CD16, CD19, CD25, CD80, CD86, HLA DR, HLA DQ ϩ (Pharmingen, San Diego, CA, USA). After 15 min at RT, Human CD34 progenitor cells were enriched from cord the cells were washed and analysed in a FACSCAN flow blood obtained from normal full-term deliveries collected cytometer using standard CellQuest acquisition software following Ethics Committee approval and consent from all (Becton Dickinson, San Jose, CA, USA). All samples were mothers. Prior to Ficoll separation, plasma was obtained by gated using forward and side scatter to exclude dead cells. a brief centrifugation process. All serum or plasma samples Both the percentage of cells expressing CD markers and ° were heat inactivated at 56 C for 30 min. The plasma- also the fluorescence intensity defined by the geometric (G) depleted blood was diluted 1:1 in RPMI and the mono- mean of the CD expressed by the cells were taken into nuclear cells separated by a Ficoll–Hypaque density centri- account in the analyses performed. Statistical analyses were fugation (Nycomed, Birmingham, UK). The mononuclear performed by Student’s t test. A value of P Ͻ 0.05 was cell fraction was then subjected to a negative selection pro- considered significant. cess. Briefly, the cells were first incubated with a CD3- depleting antibody cocktail, plus a CD34 enrichment cock- tail for 15 min at room temperature (RT). This enrichment Mixed lymphocyte:dendritic cell reaction (MLDCR) cocktail contains a mixture of antibodies to deplete cells In order to assess the allostimulatory capacity of CD34ϩ- expressing CD2, CD3, CD14, CD16, CD19, CD24, CD56, derived DCs, these cells were irradiated (25 Gy) and cul- CD66b and (Stemcell Technologies, Van- tured with purified CD4ϩ T cells (50000 cells/well) at dif- couver, Canada). This incubation was followed by the ferent cell ratios. Triplicate wells were set up. CD4ϩ T cells addition of a magnetic colloid for a further 15 min. Finally, were obtained by a negative selection process with a T cell the cells were separated over a magnetic column (Stemcell enrichment cocktail (Stemcell Technologies) as above. The ϩ Technologies). Using this approach, the yield of CD34 purity of this population was Ͼ90% in all experiments. cells obtained was between 0.1 to 1% of the total mono- After 4 days, 3H-thymidine (1 ␮Ci/well) was added and the nuclear cells, depending on the sample. The purity of the cells were incubated for a further 12 to 18 h. T cell prolifer- cell separation used in all the experiments was higher than ation was measured on a beta-counter (Packard Biosci- 75%, and in fact the majority of the samples analysed con- ences, Berks, UK). The results are presented as the mean ϩ tained over 90% CD34 cells (data not shown). c.p.m. values of triplicate cultures. All the experiments ϩ The enriched CD34 cells were washed twice in PBS comparing ACBP and ABS were made in parallel, using and approximately 25 to 30 × 105 cells/ml were plated in the same T cells as responders. IMDM culture medium (Gibco BRL, Paisley, UK) contain- ing, for comparative purposes, either 5% pooled (n = 3) ABϩ ABS, or 10% ACBP or 10% HCBP (obtained from Results a different donor) and different mixtures of cytokines. Fresh cytokines and medium were added to the cell cultures on Phenotype of differentiated DCs after 14 days in culture days 4 and 8 by carefully replacing half of the culture Cord blood-enriched CD34ϩ cells were cultured in ACBP medium. On day 14, unless otherwise indicated, the cells or ABS in medium containing GM-CSF, SCF, IL-4, Flt3- were collected by gentle pipetting and further processed. ligand and TNF␣ (rich mixture of cytokines). After 7 days The quantity of serum proteins at working concentrations in culture some morphological differences were observed. was measured by direct reading in a spectrophotometer at Cells in ABS showed an increased capacity to spread and 280 nm (GeneQuant II, Pharmacia Biotec, Uppsala, attach to the culture plate whilst this feature was not Sweden) and were found to be similar in all cultures (data observed in ACBP cultures (data not shown). After 14 days, not shown). cell recovery was significantly (P Ͻ 0.01) greater in ABS cultures than in ACBP cultures (520000 Ϯ 190000 vs Cytokines 150000 Ϯ 75000 total cells, n = 4). On day 14, the pheno- typic features of these cells were measured by flow cytome- All cytokines used in this study were of human recombinant try. The results showed that over 90% of the cell population origin (R&D Systems, Oxford, UK) and used as indicated expressed HLA class II molecules (DR ϩ DQ), and other at the following concentrations: GM-CSF 20 ng/ml, IL-4 DC associated markers such as CD40, CD1a, CD80, CD83 20 ng/ml, TNF␣ 20 ng/ml, SCF 100 ng/ml, Flt3 ligand 100 and CD86 were expressed by different proportion of cells ng/ml. Throughout the text we have used the terms rich (Figure 1a). Cells cultured in either ABS or ACBP were and poor mixture of cytokines to define those cultures con- able to differentiate into DCs showing a similar percentage taining either all the cytokines mentioned above (rich of cells expressing these markers, although cells cultured mixture) or GM-CSF and TNF␣ only (poor mixture). in ACBP showed a slight but significant increase in the percentage of cells expressing CD83 and CD86 markers. Immunostaining and flow cytometry The intensity of expression (mean) was similar in both ACBP and ABS culture derived cells (Figure 1b). These Cultured cells were collected and washed twice in PBS results suggest that under these conditions both ACBP and Dulbecco’s. The following FITC or PE conjugated anti- ABS could equally support the differentiation of CD34ϩ bodies to the specific markers and isotype controls were cells into DCs.

Bone Marrow Transplantation Generation of cord blood DCs with autologous plasma FE Borra`s et al 373 GM-CSF + SCF + IL4 + Flt3-L + TNFa GM-CSF + TNFa 100 100 a a

80 80 ** * * 60 60 * * * 40 40 % +ve cells % +ve cells

20 20

0 0 CD40 CD25 CD83 CD86 CD14 CD40 CD25 CD83 CD86 CD14 CD80 CD80 CD1A CD34 CD1A CD34 HLA-DR HLA-DQ HLA-DR HLA-DQ 2500 b ACBP 2500 b ABS ACBP 2000 ** ABS 2000 1500

1000 ** ** 500 500 Fluorescence intensity (mean) 0 Fluorescence intensity (mean) 0 CD40 CD25 CD83 CD86 CD14 CD80 CD1A CD34 HLA-DR HLA-DQ CD40 CD25 CD83 CD86 CD14 CD80 CD1A CD34 Figure 1 DC phenotype after 14 days in culture in the presence of either HLA-DR HLA-DQ ACBP (white boxes) or ABS (filled boxes). Cells were differentiated with Figure 2 DC phenotype after 14 days in culture in the presence of either GM-CSF, IL-4, SCF, Flt3-L and TNF␣. Results are expressed as the mean ACBP (white boxes) or ABS (filled boxes). Cells were differentiated with values of four independent experiments. (a) The percentage of positive GM-CSF and TNF␣ only. Results are expressed as the mean values of cells (%+ve in the Y axis); (b) the intensity of fluorescence (*P Ͻ 0.05). three independent experiments. (a) The percentage of positive cells (%+ve in the Y axis), and (b) the intensity of fluorescence (*P Ͻ 0.05, **P Ͻ 0.01). In order to investigate whether this effect could be due to the composition of cytokines used, cord blood-enriched CD34ϩ cells were cultured in ACBP or ABS in GM-CSF and TNF␣ only (poor mixture of cytokines). In this case On the other hand, cells obtained in the presence of ABS IL-4 was not added to the cultures. The cell recovery under were a mixture of ‘less mature’ DCs and monocyte– these conditions was similar to that obtained above, 665000 macrophage cells as reflected by the lower expression of Ϯ 149000 for ABS vs 191000 Ϯ 117000 for ACBP (total CD83 and the high expression of CD14. cells, n = 3, P Ͻ 0.01). However, in this case, the DC To further confirm the effect of ACBP on DC differen- phenotype of the cells was greatly enhanced in the ACBP tiation, the phenotype of cord blood CD34ϩ cells differen- cultures, as reflected by the higher percentage of cells tiated in either a rich or poor mixture of cytokines, but expressing CD1a, CD83 and CD86 (Figure 2a). In contrast, using ACBP as source of serum factors, was compared. The CD14 expression was greatly enhanced in ABS cultures, percentage of positive cells and mean values of cells whereas it remained at low levels in ACBP cultures. Fur- obtained using ACBP in either the rich or poor mixture of thermore, a great number of the cells obtained in ABS had cytokines were the same (see white boxes in Figures 1 and the same shape, size and appearance as monocyte– 2), indicating that the use of additional cytokines had no macrophages showing an ability to attach to the culture significant effect in the differentiation of cord blood CD34ϩ wells (data not shown). The effect of ACBP on DC differ- cells into DCs. entiation was also reflected by the mean values of DR, DQ, The ability of HCBP to generate DCs was also tested. and CD40 expression which were higher in ACBP cell cul- The results showed that under the same conditions cells tures, while the CD14 mean value was higher in ABS cul- cultured in HCBP had DC characteristics similar to those tures (Figure 2b). These results suggest that ACBP was by obtained in ACBP (data not shown). Thus, the ability to itself able to induce DC differentiation from CD34ϩ cord generate DCs from enriched CD34ϩ cord blood cells cul- blood cells in cultures containing GM-CSF and TNF␣ only. tured in the poor mixture of cytokines (GM-CSF and TNF␣

Bone Marrow Transplantation Generation of cord blood DCs with autologous plasma FE Borra`s et al 374 only) can be supported by either homologous or autologous powerful adjuvants for inducing immune responses to cord blood plasma. tumours18 as it has been postulated that the defective immune responses to tumours could be due to deficient MLDC reactions antigen presentation of the relevant tumour antigens. It is possible that this deficiency may be due to a lack of DC The results obtained in the MLDC reactions showed that differentiation as a result of the tumour process. Menetrier- the DCs generated in ACBP had a higher capacity to induce Caux et al19 have described the role of IL-6 and M-CSF CD4ϩ T lymphocyte proliferation compared to those gener- derived from tumour cells in the inhibition of differen- ated in ABS. The capacity of DCs to cause allostimulation tiation of DCs from CD34ϩ precursors. In addition, IL-10 was assessed using either cord or adult purified T cells. and vascular endothelial growth factor (VEGF) both DCs generated in the presence of ACBP showed a similar secreted by tumour cells have been shown to inhibit the profile of allostimulation either they were cultured in the development of monocyte-derived DCs.20 Other studies presence of rich or poor mixture of cytokines (Figure 3, have suggested that an acidic environment in the tumour panels a and b). By contrast, DCs generated in ABS only surrounding areas may contribute to the lack of monocyte- showed allostimulation capacity when cultured in the pres- derived DC maturation and function.21 Therefore, the possi- ence of rich mixture of cytokines, whereas those generated bility to promote DC differentiation may be a useful in the presence of the poor mixture of cytokines were approach to develop competent DCs to induce immune unable to induce a strong allostimulatory response (Figure responses to tumours or infections. 3, panel c and d). Also, the allostimulatory capacity of DCs One of the main sources of in vitro-derived DCs are generated in ACBP was slightly higher than that obtained monocytes. The main advantage of this model is that mono- with the ABS-derived DCs. Cells generated in the presence cytes are a large source of easily obtainable cells that can of homologous plasma showed similar stimulation indices be directed to a DC phenotype. The other major source of to autologous plasma DCs (data not shown). However, in DCs are the CD34ϩ progenitor cells. In this study we all cases, higher stimulation capacities were apparent at describe the generation of mature DCs from CD34ϩ cord high T cell:DC ratios. blood cells in the presence of ACBP and using GM-CSF and TNF␣ only as differentiation cytokines. The majority of the protocols described so far have used Discussion either fetal calf serum (FCS) or adult blood serum (ABS) to differentiate DCs.11,15,21,22 Rosenzwajg et al22 reported a The generation of dendritic cells from different cell sources method to generate DCs in vitro from CD34ϩ cord blood is considered to be of major importance as possible immun- precursors using FCS. However, the use of FCS in DC cul- otherapy treatment. The number and maturity of the DCs tures may limit their potential use in clinical immunother- generated may determine their potential clinical use. In the apy.23 Due to the time required to generate DCs in culture, last few years DCs have been identified as potentially the exposure to serum antigens could induce significant

2500 a 8000 b 7000 2000 6000 1500 5000 4000 1000 3000 (c.p.m.) (c.p.m.) 2000 500 1000 Thymidine incorporation Thymidine incorporation 0 0 1000 100 10 1 1000 100 10 1 Adult T cell: APC ratio Cord T cell: APC ratio 6000 1600 c d 1400 5000 1200 4000 1000 800 3000 600 (c.p.m.) (c.p.m.) 2000 400 1000 200 Thymidine incorporation 0 Thymidine incorporation 0 1000 100 10 1 1000 100 10 1

Adult T cell: APC ratio Cord T cell: APC ratio

Figure 3 Allostimulatory capacity of irradiated DCs. (a and b) Cells obtained in the presence of ACBP; (c and d) cells obtained in the presence of ABS. Black square are cells obtained in a rich mixture of cytokines, whereas white circles represent cells obtained with GM-CSF and TNF␣ only (poor mixture of cytokines). The mean values of a single representative experiment are shown. Adult PBMC allostimulation control is shown by a white triangle. Background value is subtracted in all the plots.

Bone Marrow Transplantation Generation of cord blood DCs with autologous plasma FE Borra`s et al 375 variables. On the other hand, the use of homologous AB tors using a poor mixture of cytokines (GM-CSF and TNF␣ serum could have detrimental side-effects in the recipient, only). Further experiments are now underway to determine e.g. transmission of infections. These limitations could then the antigen processing and presentation capacities of these be overcome by using autologous plasma (AP) as a source cells. Their potential clinical application for immunother- of and serum factors. AP has the additional advan- apy protocols requires further investigation. tage that it is of the same origin as the cells, so it is likely to be less hazardous. Nicola et al24 have shown that by using AP they were able to generate mature DCs from adult References CD34ϩ cells without the addition of IL-4 or IL-13. Never- theless, they used a rich mixture of cytokines (GM-CSF, 1 Steinmann RM. The dendritic cell system and its role in TNF␣, SCF and FLt3-L) to generate those DCs. The results immunogenicity. Annu Rev Immunol 1991; 9: 271–296. presented in this study show that in the presence of ACBP, 2 Romani N, Koide S, Crowley M et al. Presentation of exogen- DCs can be generated from cord blood CD34ϩ cells using ous protein antigens by dendritic cells to T cell clones: intact GM-CSF and TNF␣ only. Thus, ACBP could substitute or protein is presented best by immature epidermal Langerhans even improve the effect of IL-4 added to the ABS cultures cells. J Exp Med 1989; 169: 1169–1178. to block the monocytic differentiation pathway and to pro- 3 Steinmann RM, Hoffman L, Pope M. Maturation and 8,17 migration of cutaneous dendritic cells. J Invest Dermatol mote DC. Although cell yields are higher using ABS, it 1995; 105: 2S–7S. is still possible to obtain approximately six times the initial 4 Pick WF, Majdic O, Kohl P et al. Molecular and functional number of cultured cells using ACBP. The higher cell characteristics of dendritic cells generated from highly purified yields obtained with ABS were due, at least in part, to the CD14ϩ peripheral blood monocytes. J Immunol 1996; 157: higher amount of apoptosis detected in ACBP cultures (data 3850–3859. not shown). 5 Cella M, Scheidegger D, Palmer-Lehmann K et al. Ligation One of the cytokines involved in promoting the mono- of CD40 on dendritic cells triggers production of high levels cytic differentiation pathway is M-CSF.25 M-CSF is of interleukin 12 and enhances T cell stimulatory capacity: T- secreted by uterine glandular epithelial cells26 as well as T help via APC presentation. J Exp Med 1996; 184: 747–752. other cell types, supports the growth of placental tropho- 6 Koch F, Stanzl U, Jennewein P et al. High level IL-12 pro- 27 26 duction by murine dendritic cells: up-regulation via MHC blasts and is increased in serum during pregnancy. In class II and CD40 molecules and down-regulation by IL-4 and fact, M-CSF is also present in cord blood plasma at high IL-10. J Exp Med 1996; 184: 741–746. concentrations (unpublished observations). However, at this 7 Hart DNJ. Dendritic cells: unique leukocyte populations point it is not yet clear whether the effect of ACBP in the which control primary immune response. Blood 1997; 90: differentiation of DCs has any relationship to the presence 3245–3287. of M-CSF in cord blood plasma. On the other hand, pre- 8 Sallusto F, Lanzavecchia A. Efficient presentation of soluble vious data indicates that tumour-derived M-CSF in combi- antigen by cultured human dendritic cells is maintained by nation with interleukin 6 (IL-6) blocks the DC differen- granulocyte–macrophage colony-stimulating factor plus tiation pathway.19 Further investigations on the role of cord interleukin 4 and downregulated by tumour necrosis factor blood plasma M-CSF on DC differentiation and on the alpha. J Exp Med 1994; 179: 1109–1118. 9 Romani N, Reider D, Heuer M et al. Generation of mature cytokine content of cord blood plasma and adult serum are dendritic cells from human blood. An improved method with underway to clarify these observations. special regard to clinical applicability. J Immunol Meth 1996; With regard to the allostimulatory capacity, the ACBP 196: 137–151. generated DCs were more effective than the DCs generated 10 Reddy A, Sapp M, Feldman M et al. A monocyte conditioned in ABS. Moreover, DCs generated in the presence of ACBP medium is more effective than defined cytokines in mediating either in a rich or poor mixture of cytokines were similarly the terminal maturation of human dendritic cells. Blood 1997; 90: 3640–3646. effective. Interestingly, DCs generated in the presence of ϩ ABS and poor mixture of cytokines, were unable to induce 11 Herbst B, Kohler G, Mackensen A et al. CD34 peripheral a strong alloresponse. One of the features of this population blood progenitor cell and monocyte derived dendritic cells: a comparative analysis. Br J Haematol 1997; 99: 490–499. is the expression of CD14 and lower expression of CD83, ϩ 12 Caux C, Vanbervliet B, Massacrier C et al. CD34 hematopo- suggesting a monocyte–macrophage phenotype. These ietic progenitors from human cord blood differentiate along observations support the view that ACBP could substitute two independent dendritic cell pathways in response to GM- or even improve the effect of IL-4 added to the ABS cul- CSF ϩ TNF␣. J Exp Med 1996; 184: 695–706. tures to block the monocytic differentiation pathway and to 13 Saraya K, Reid CD. Stem cell factor and the regulation of promote DC differentiation.8,17 dendritic cell production from CD34ϩ progenitors in bone The use of autologous plasma to generate DCs from adult marow and cord blood. Br J Haematol 1996; 93: 258–264. monocytes and adult CD34ϩ cells have been previously 14 Maraskovsky E, Brasel K, Teepe M et al. In vivo adminis- reported.21,28 Ratta et al28 have found that expansion of tration of flt3 ligand results in generation of large numbers of CD34ϩ cell-derived DCs was further enhanced when auto- dendritic cells in the lymphoid tissue of mice. J Exp Med logous serum rather than FCS or serum-free medium was 1996; 184: 1953–1962. 21 15 Arrighi JF, Hauser C, Chapuis B et al. Long-term culture of used in the cultures. In addition, Eljaafari et al have dem- human CD34ϩ progenitors with Flt3 ligand, thrombopoietin onstrated that autologous plasma is more efficient than and SCF induces extensive amplification of a CD34− CD14− homologous serum in DC differentiation from adult mono- and a CD34− CD14ϩ dendritic cell precursor. Blood 1999; 93: cytes. This study has showed that ACBP is able to support 2244–2252. the differentiation of DCs from cord blood CD34ϩ progeni- 16 Geissmann F, Prost C, Monnet JP et al. Transforming growth

Bone Marrow Transplantation Generation of cord blood DCs with autologous plasma FE Borra`s et al 376 factor ␤1, in the presence of granulocyte/macrophage colony- human dendritic cell differentiation from cord blood CD34ϩ stimulating factor and interleukin 4, induces differentiation of progenitor cells. Exp Hematol 1998; 26: 63–72. human peripheral blood monocytes into dendritic Langerhans 23 Rosenzwajg M, Gluckman JC. Problems raised by the in vitro cells. J Exp Med 1998; 187: 961–966. generation of dendritic cells for use in immunotherapy. Exp 17 Romani N, Gruner S, Brang D et al. Proliferating dendritic Hematol 1999; 27: 387–388. cell progenitors in human blood. J Exp Med 1994; 180: 83–93. 24 Nicola M, Longoni P, Magni M, Gianni AM. The influence 18 Schuler G, Steinmann RM. Dendritic cells as adjuvants for of interleukin (IL)-4, IL-13 and Flt3 ligand on human dendritic immune-mediated resistance to tumors. J Exp Med 1997; 186: cell differentiation from cord blood CD34ϩ progenitor cells. 1183–1187. The importance of the source of serum. Exp Hematol 1998; 19 Menetrier-Caux C, Montmain G, Dieu MC et al. Inhibition of 27: 386–387. the differentiation of dendritic cells from CD34ϩ progenitors 25 Sherr CJ. Colony stimulating factor-1 receptor. Blood 1990; by tumor cells: role of interleukin 6 and macrophage colony- 75: 1–12. stimulating factor. Blood 1998; 92: 4778–4791. 26 Praloran V, Coupey L, Donnard M et al. Elevation of serum 20 Steinbrink K, Jonuleit H, Muller G et al. Interleukin 10-treated M-CSF concentrations during pregnancy and ovarian hypersti- human dendritic cells induce a melanoma-antigen-specific mulation. Br J Haematol 1994; 86: 675–677. anergy in CD8ϩ T cells resulting in failure to lyse tumor cells. 27 Saito S, Saito M, Enomoto M et al. Human macrophage col- Blood 1999; 93: 1634–1642. ony stimulating factor induces differentiation of trophoblasts. 21 Eljaafari A, Duperrier K, Mazet S et al. Generation of stable Growth Factors 1993; 9: 11–19. monocyte-derived dendritic cells in the presence of high con- 28 Ratta M, Rondelli D, Fortuna A et al. Generation and func- centrations of homologous or autologous serum: influence of tional characterisation of human dendritic cells derived from extra-cellular pH. Hum Immunol 1998; 59: 625–634. CD34 cells mobilised into peripheral blood: comparison with 22 Rosenzwajg M, Camus S, Guigon M, Gluckman JC. The bone marrow CD34ϩ cells. Br J Haematol 1998; 101: 756– influence of interleukin (IL)-4, IL-13 and Flt3 ligand on 765.

Bone Marrow Transplantation