Erythropoietin enhances immunostimulatory properties of immature dendritic cells

Abstract

Dendritic cells (DCs) are the most potent antigen-presenting cells and play a crucial role by modulating the T cell immune response against infective agents, tumour antigens and alloantigens. The current study shows that differentiating bone marrow (BM)-derived DCs but not fully differentiated DCs are targets of erythropoietin (EPO). Indeed, DCs emerging from rat bone marrow, but not splenic DCs, express the EPO receptor (Epo-R) and respond to EPO stimulation displaying a more activated phenotype with increased CD86, CD40 and interleukin (IL)-12 expression levels and a higher allostimulatory capacity on T cells than untreated DCs. Moreover, results here presented show that EPO up-regulates Toll-like receptor (TLR)-4 in differentiating DCs rendering these cells more sensitive to stimulation by the TLR-4 ligand lipopolysaccharide (LPS). Indeed, DCs treated with EPO and then stimulated by LPS were strongly allostimulatory and expressed CCR7, CD86, CD40, IL-12 and IL-23 at higher levels than those observed in DCs stimulated with LPS alone. It is tempting to speculate that EPO could act as an additional danger signal in concert with TLR-4 engagement. Thus, EPO, beyond its erythropoietic and cytoprotective effects, turns out to be an immune modulator.

Introduction

Erythropoietin (EPO), produced by the kidney in response to hypoxia, regulates the survival, proliferation and differentiation of erythroid progenitor cells [1]. The expression of EPO receptors in non-erythroid tissues such as the brain, kidney, heart and vascular endothelium has been associated with the discovery of novel biological functions of endogenous EPO signalling in non-haematopoietic tissues [2,3]. Thus, in addition to its essential role in the regulation of mammalian erythropoiesis, EPO has emerged as a major tissue-protective survival factor in various non-haematopoietic organs and this cytoprotective effect has been associated primarily with a reduction in apoptotic cell death [4,5]. Moreover, recent studies, although with discordant results, have raised the possibility that EPO affects the immune cell function. Yuan et al. have shown that EPO treatment, beyond the neuroprotective effect, reduced inflammation in a model of experimental autoimmune encephalomyelitis (EAE) in mice [6]. Inflammation reduction was attributed to a block of dendritic cell (DC) expansion and T cell proliferation and a concomitant expansion of regulatory T cells (T_regs). Conversely, opposite results were reported by Neumann et al., who showed that both human monocyte-derived DCs and mouse bone marrow (BM)-derived DCs treated with EPO acquired a mature phenotype with a higher expression of CD80, CD86 and major histocompatibility complex (MHC) class II [7,8] and produced a greater quantity of interleukin (IL)-12 than untreated DCs.

The effects of EPO on DCs and T cells, if any, may have great impact on the immune response of patients chronically treated with this drug to correct anaemia. On one hand, mature DCs, in view of their potent immunoactivating properties, may be beneficial in the context of cancer by efficiently presenting tumoral antigens and licensing T cells to attack tumoral cells. On the other hand, mature DC action can be deleterious in the case of renal transplantation, as DCs stimulate recipient T cells to reject the graft.

We designed this study with the aim of shedding light on the complex immunomodulatory action of EPO by studying the effect of EPO on dendritic cell phenotype and allostimulatory activity in vitro.

Materials and methods

Animals and reagents

Brown Norway (BN, RT1ⁿ) rats were used for preparation of BM and spleen-derived DCs. Lymph node cells were obtained from BN or Lewis (LW, RT1^l) rats. Antibodies specific for rat determinant included antibodies for CD11c [clone 8A2, fluorescein isothiocyanate (FITC)-conjugated; Serotec, Oxford, UK], CD86 [clone 24F, phycoerythrin (PE)-conjugated; BD Pharmingen, San Diego, CA, USA], MHC-II [clone OX 6, allophycocyanin (APC)-conjugated; Serotec], CD25 (clone OX39; BD Pharmingen), CD134 (clone OX40; Serotec) and EPO receptor (Epo-R, clone M-20; Santa Cruz Biotechnology, Santa Cruz, CA, USA).

GMP-manufactured sterile syringes of EPO dedicated for patients were obtained from Amgen (Darbepoetin alfa, Aranesp, Thousand Oakc, CA, USA). We have used Darbepoetin previously and successfully in vivo in a model of renal ischaemia/reperfusion injury in the rat [9]. Recombinant rat EPO (endotoxin-free) was from Sigma-Aldrich (St Louis, MO, USA) and was used after reconstitution with phosphate-buffered saline (PBS).

BM-derived DCs

BM-DCs were obtained as described previously [10]. BM cells were cultured in RPMI-1640 supplemented with 10% rat serum, 100 U/ml of granulocyte-macrophage colony-stimulating factor (GM-CSF) and 20 ng/ml IL-4 (Insight Biotechnology, Wembley, UK). Every second day, 75% of the medium was replaced. From day 7 different EPO doses (0·05 ng/ml, 5 ng/ml and 500 ng/ml) were added to the culture (Fig. 1, setting 1: EPO-DCs). Additional experiments were carried out with recombinant rat EPO (50 and 500 ng/ml) instead of Aranesp. At day 11 cells were harvested, analysed by fluorescence activated cell sorter (FACS) (FACSAria; Becton Dickinson, Mountain View, CA, USA) and used as stimulators for primary mixed leucocyte reaction (MLR). In selected experiments, at day 8, 20 µg/ml of lipopolysaccharide (LPS) (Sigma Aldrich) was added to either EPO-treated (500 ng/ml, Fig. 1, setting 3: EPO/LPS-DCs) or untreated cells. Additional experiments in which BM cells were stimulated with LPS at day 7 and then treated with EPO (500 ng/ml) at day 8 were also performed (Fig. 1, setting 2: LPS/EPO-DCs).

Fig. 1.

Experimental design for erythropoietin (EPO) and lipopolysaccharide (LPS) treatment of bone marrow cell culture. Setting 1: EPO was administered at days 7 and 9 of culture [EPO-dendritic cells (DCs)]. Setting 2: LPS was added at day 7 and EPO was administered at days 8 and 9 of culture (LPS/EPO-DCs). Setting 3: EPO was administered at days 7 and 9, LPS was added at day 8 of culture (EPO/LPS-DCs). In settings 1, 2 and 3, cells were harvested at day 11 for mixed leucocyte reaction (MLR), fluorescence activated cell sorter (FACS) and real-time polymerase chain reaction (PCR) experiments. In setting 1, cells were harvested even at days 7 and 8 for real-time PCR of Toll-like receptor (TLR)-4 expression.

Splenic DCs

Spleens were minced and incubated with type IV collagenase (1 mg/ml; Sigma-Aldrich) for 30 min at 37°C. At the end of incubation, tissues were passed through a stainless steel sieve and centrifuged, and erythrocytes were removed by osmotic lysis. Remaining cells were incubated for 2 h and half (at 37°C with 5% CO₂) in culture medium [RPMI-1640 10% fetal calf serum (FCS)]. At the end of the incubation non-adherent cells were removed by extensive washing with warm PBS. The adherent cells were cultured overnight in RPMI-1640 10% FCS at 37°C with 5% CO₂ in the presence or absence of EPO (500 ng/ml). At the end of overnight culture, floating cells were collected as enriched DCs. To obtain a pure splenic DC population, the enriched population was sorted for CD11c-positive cells (purity > 90%) using FACSAria (Becton Dickinson).

MLR

MLR was performed using DCs from all the above-mentioned experimental conditions and lymph node (LN) cells (LW or BN for allogeneic or syngeneic combinations, respectively). The DC : LN ratio was 1:100. Cultures were maintained in RPMI-1640/FCS 20% in 5% carbon dioxide in air at 37°C for 4 days. In selected experiments MLRs were performed in the presence of EPO (500 ng/ml). T cell proliferation was evaluated by [³H]-thymidine incorporation (1 µCi/well) and expressed as counts per minute (cpm). Darbepoetin vehicle was 0·05 mg/ml polysorbate 80 in PBS. Both untreated and vehicle-treated DCs were used as control; the immunostimulatory effect was similar (12 413 ± 1853 and 12 138 ± 1377 cpm, respectively, n = 3). In selected experiments enzyme-linked immunospot assay (ELISPOT) for IFN-γ was performed after 3-day MLR (Abcam). The resulting spots were counted on a computer-assisted immunospot image analyser (Aelvis Elispot Scanner system) and expressed as number of spots/100 000 responder cells.

Additional MLR experiments were carried out to evaluate IL-12 production by DCs. In detail, DCs obtained from all the experimental conditions were stimulated by allogeneic T cells (1:50) for 3 days, then medium was harvested and IL-12 release was measured by enzyme-linked immunosorbent assay (ELISA), according to the manufacturer's instructions (Invitrogen, Carlsbad, CA, USA).

FACS analysis

Cell surface immunophenotypic analysis was performed by flow cytometry using FACSAria (Becton Dickinson).

Cells were incubated with optimal concentrations of FITC, PE or APC-conjugated or unconjugated primary antibodies for 20 min at 4°C in PBS containing 2% FCS and then washed twice in the same buffer. For EPO-R staining, cells were incubated with unconjugated primary antibodies and then with secondary antibodies, biotin-conjugated goat anti-rabbit immunoglobulin (Ig)G followed by APC-cyanin 7 (Cy7)-conjugated streptavidine. All stainings included negative control with control isotype IgG.

Western blot analysis

Reduced and denatured proteins from freshly isolated BM cells, differentiating DCs at days 7 and 8 of BM culture and splenic DCs were resolved by sodium dodecyl sulphate (SDS)-polyacrylamide gel electophoresis, as described previously [11]. Proteins were transferred to polyvinylidine fluoride (PVDF) membrane, blocked with 5% milk and incubated with anti Epo-R antibody. ECL Advance (Amersham Biosciences, Milano, Italy) was used for detection.

Quantitative real-time PCR

DCs were harvested at days 7, 8 or 11 of BM culture and total RNA was obtained by extraction with PicoPure RNA isolation kit (Arcturus, Mountain View, CA, USA). RNA was treated with DNase and reverse-transcribed to cDNA by Superscript II (Invitrogen). Quantitative real time-PCR was performed on the Applied Biosystems 7300 real-time PCR system (PE Applied Biosystems, Foster City, CA, USA) with Power Syber Green Master Mix and primers specific for rat IL-6, IL-12p35, IL-23p19, CD40, CCR7 and Toll-like receptor (TLR)-4. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) served as housekeeping gene. A complete list of primers is given in Table 1. For TLR-4 we analysed the RNA of cells harvested at days 7 and 8, while for IL-6, IL-12p35, IL-23p19, CCR7 and CD40 we considered the RNA of cells harvested at day 11 of BM culture. We used the ΔΔ threshold cycle technique to calculate cDNA content in each sample using as reference (calibrator) the cDNA expression in untreated DCs harvested at day 7 for TLR-4 and day 11 for the other primers. Results were expressed as arbitrary units (AU), taking the expression in the calibrator as 1.

Table 1.

Primers used for real-time reverse transcription–polymerase chain reaction (RT–PCR)

Target	Sequence	nM
IL-6
Forward	5′-GATGCTTCCAAACTGGATATAACCA-3′	300
Reverse	5′-AAGACCAGAGCAGATTTTCAATAGG-3′	300
IL-12p35
Forward	5′-GCACACTGGAGGCCTGCTT-3′	300
Reverse	5′-GCCAGGCAACTCTCATTCTTG-3′	300
IL-23p19
Forward	5′-ACACACCAGTGGGACAAATGG-3′	300
Reverse	5′-CACTTTTAGTCTCTTCTTCCCCTTCTT-3′	300
CD40
Forward	5′-AGCTCACTGGAACAGGGAGATC-3′	300
Reverse	5′-CTTCTTAACCTGAAGCCCTTCAT-3′	300
TLR-4
Forward	5′-CACAACTTCAGTGGCTGGATTTATC-3′	300
Reverse	5′-TTGTCTTCAATTGTCTCAATTTCACA-3′	300
CCR7
Forward	5′-AACTGCCCAGAGAGCATCATG-3′	300
Reverse	5′-CTGTGACCTCATCTTGCCAGAA-3′	300
GAPDH
Forward	5′TCATCCCTGCATCCACTGGT-3′	300
Reverse	5′CTGGGATGACCTTGCCCAC-3′	300

IL: interleukin; TLR: Toll-like receptor; GAPDH: glyceraldehyde 3-phosphate dehydrogenase.

Statistics

Results are expressed as means ± standard error (s.e.). Statistical analysis was performed by one-way analysis of variance (anova). Statistical level of significance was defined as P < 0·05.

Results

Phenotype of EPO-treated BM-derived DCs

We first evaluated the effect of EPO (Darbepoetin alfa, Aranesp) on DCs emerging from their precursors in a rat BM culture, testing three different EPO doses: 0·05, 5 and 500 ng/ml. As documented by FACS results, EPO (given to BM from day 7 of culture) did not influence the number of emerged CD11c⁺ DCs at doses of 0·05 and 5 ng/ml (65 ± 12% and 68 ± 14%, respectively, versus untreated DCs: 65 ± 12%, n = 6). Five hundred ng/ml EPO showed a trend to increase the formation of CD11c⁺ DCs from BM cultures (72 ± 10%, n = 6), even though a statistical significance was not reached. We then performed FACS analysis of MHC-II and CD86, both molecules involved in DC-mediated antigen presentation and T cell activation. At a dose of 500 ng/ml (Fig. 2a,b), EPO induced a slight increase in the percentage of CD86⁺ cells and a significant up-regulation of CD86 expression levels in DCs emerged from BM culture, without affecting MHC-II expression. Messenger RNA levels for CD40, another important co-stimulatory molecule, were doubled in DCs which emerged in the presence of 500 ng/ml EPO compared to untreated DCs, whereas the chemokine receptor CCR7 did not increase significantly (Fig. 2c). The cytokine profile (by real-time PCR) of EPO-treated BM cell cultures revealed that expression of IL-6 and IL-23 mRNA were comparable to those observed in untreated DCs, whereas IL-12 mRNA expression was doubled (Fig. 2c). At protein level, ELISA assay showed that EPO-DCs produce a significantly higher amount of IL-12 than untreated DCs (75 ± 12 versus 38 ± 7 pg/10⁵ cells, respectively, P < 0·05, n = 4 independent experiments). Viability in EPO-treated (at any dose used) and untreated emerged DCs was similar (around 95%, trypan blue dye exclusion).

Fig. 2.

Phenotype of erythropoietin-dendritic cells (EPO-DCs), lipopolysaccharide (LPS)-DCs and EPO/LPS-DCs. (a) Fluorescence activated cell sorter (FACS) analysis of CD86 and major histocompatibility complex II (MHC-II) expression in untreated-DCs, EPO-DCs, LPS-DCs and EPO/LPS-DCs (obtained as described in Fig. 1). Data are mean ± standard error (s.e.), n = 3 independent experiments. *P < 0·05 versus untreated DCs. MFI: median fluorescence intensity. (b) Dot-plots showing FACS analysis of CD86 and MHC-II expression in untreated DCs, EPO-DCs, LPS-DCs and EPO/LPS-DCs. Positive events are displayed in the region gate on the right. Dot-plots are representative of three independent and similar experiments. (c) mRNA expression of CD40, CCR7 and cytokines [interleukin (IL)-6, IL-12, IL-23] by real-time polymerase chain reaction (PCR) in untreated DCs, EPO-DCs, LPS-DCs and EPO/LPS-DCs harvested at the end of bone marrow culture (obtained as described in Fig. 1). The cDNA content in each sample was calculated by the ΔΔCT technique, using as calibrator the cDNA expression in untreated DCs. Data are mean ± s.e., n = 3 independent experiments. *P < 0·05 versus untreated DCs. #P < 0·05 versus EPO-DCs. $P < 0·05 versus LPS-DCs.

EPO enhances allostimulatory capacity of BM-derived DCs

The allostimulatory functions of BM-derived DCs were tested in MLR experiments. As shown in Fig. 3a, the allogeneic T cell proliferation towards DCs emerged in the presence of EPO at the dose of 0·05 ng/ml (Fig. 1, setting 1, EPO-DCs) was similar to that observed in response to untreated cultures. At variance, increasing EPO concentration in BM cultures (5 and 500 ng/ml) led to the emergence of DCs with a higher allostimulatory capacity than that of untreated DCs. Indeed, as shown in Fig. 3a, allogeneic T cell proliferation towards DCs treated with EPO (500 ng/ml) was significantly higher than that towards untreated DCs. Similarly, the frequency of IFN-γ producing T cells was significantly higher than that observed with untreated DCs (113 ± 24 versus 62 ± 8 spots/10⁵ cells, respectively, P < 0·05, n = 3 independent experiments). Similar findings were observed when rat recombinant EPO was added to BM cultures (Fig. 3b,c). The activation status of T cells at the end of MLR was evaluated by FACS analysis of CD25 and CD134 expression. Consistent with the proliferation data, MLR performed with EPO-treated BM-derived DCs (at the dose of 500 ng/ml) displayed an increased proportion of both CD25⁺ (43 ± 0·5%, n = 4, P < 0·05 versus untreated) and CD134^high+ (26 ± 3·5%, n = 4, P = 0·06 versus untreated) T cells, compared to MLR performed with untreated DCs (CD25⁺ cells 32 ± 2·5%, CD134^high+ 16 ± 2%, n = 4).

Fig. 3.

Allostimulatory capacity of erythropoietin-dendritic cells (EPO-DCs). (a) DCs treated with EPO (darbepoetin) at different doses (0·05 ng/ml, 5 ng/ml and 500 ng/ml) and untreated DCs were used as stimulators in a 4-day mixed leucocyte reaction (MLR) with allogeneic (from Lewis rat) lymph node cells. Proliferative response was evaluated by [³H]-thymidine incorporation and expressed as counts per minute (cpm). Data are mean ± standard error (s.e.), n = 6 independent experiments, *P < 0·05 versus untreated DCs. Syngeneic control: MLR with untreated DCs and syngeneic (from Brown Norway rat) lymph node cells. (b) DCs treated with EPO (rat recombinant EPO) at different doses (50 ng/ml and 500 ng/ml) and vehicle-treated DCs were used as stimulators in a 3-day MLR with allogeneic (from Lewis rat) lymph node cells. Proliferative response was evaluated by [³H]-thymidine incorporation and expressed as cpm. Data are mean ± s.e., n = 3 independent experiments, *P < 0·05 versus vehicle-treated DCs. (c) DCs treated with EPO (rat recombinant EPO) at different doses (50 ng/ml and 500 ng/ml) and vehicle-treated DCs were used as stimulators in a 4-day MLR. Frequency of interferon (IFN)-γ-producing T cells was assessed by enzyme-linked immunospot assay (ELISPOT) assay. Data are mean ± s.e., n = 3 independent experiments, *P < 0·05 versus vehicle-treated DCs. Syngeneic control: MLR with vehicle-treated DCs and syngeneic (from Brown Norway rat) lymph node cells. (d) MLR were performed with either untreated DCs or EPO-DCs in the presence or not of EPO (500 ng/ml). Data are mean ± s.e., n = 3 independent experiments.

When EPO at 500 ng/ml was added during MLR with either EPO-treated or untreated BM-derived DCs, T cell proliferation was not modified compared to that observed in control MLR (Fig. 3d).

Taken together, the above results indicate that EPO, given during BM culture, enhanced the immunostimulatory capacity of emerged DCs. No supplementary effect was obtained by adding EPO during MLR when DCs were fully differentiated.

EPO does not exert any effect on allostimulatory capacity of splenic DCs

To investigate further the effect of EPO on fully differentiated DCs, splenic DCs were treated with EPO. No differences were observed in T cell proliferation towards either untreated or EPO-treated splenic DCs (14 026 ± 3294 and 15 649 ± 3765 cpm, respectively, n = 6 independent experiments, 1286 ± 363 and 1315 ± 399 syngeneic controls with either untreated or EPO-treated splenic DCs, respectively). In the search for a possible explanation of this divergent EPO effect on different DC subsets, both BM-derived and splenic DCs were FACS-analysed for EPO receptor (Epo-R) expression using a specific antibody. As shown in Fig. 4a, on average 40% of BM DCs express Epo-R at day 7 of culture, while only a small percentage of splenic DCs showed Epo-R expression. Consistent with FACS results, real-time PCR experiments showed that splenic DCs express lower levels of Epo-R mRNA (on average threefold lower 0·33 ± 0·01 n = 3, compared to Epo-R expression in DCs at day 7 of BM culture taken as calibrator = 1, n = 3). Consistent with the FACS results, a 59 kDa band corresponding to Epo-R was detectable in the protein extracts of days 7 and 8 BM-differentiating DCs (Fig. 4b). At variance, no 59 kDa band was detected in the protein extracts of splenic DCs. According to Elliott et al. [11] the M-20 antibody, which we used for FACS analysis, clearly identified the Epo-R protein in the lysates of freshly isolated BM cells taken as positive control (Fig. 4b, lane 1).

Fig. 4.

Erythropoietin (EPO)-receptor expression on dendritic cells (DCs). (a) Fluorescence activated cell sorter (FACS) analysis of double staining for CD11c [fluorescein isothiocyanate (FITC)] and Epo-R [allophycocynin-cyanin 7 (APC-Cy7)] on bone marrow (BM)-DCs at day 7 of culture and splenic DCs. Percentage of CD11c⁺Epo-R⁺ cells and CD11c⁺Epo-R^- cells are reported inside the quadrant. Negative controls are performed with control isotype immunoglobulin (Ig)G. FACS dot-plots are representative of two independent and similar experiments. (b) Western blot with anti-Epo-R antibody (clone M-20) on protein extracts (20 µg for each lane) of freshly isolated BM cells (lane 1), BM-DCs at day 7 (lane 2) and 8 (lane 3) of BM culture and splenic DCs (lane 4). The migration of molecular mass markers in kDa is indicated on the left. The blot is representative of two independent and similar experiments.

EPO modulates LPS effect on DC allostimulatory capacity

The engagement of TLR-4 initiates a cascade of events in DCs that leads to maturation and secretion of immunomodulatory factors, thus pushing DCs to be strongly immunostimulatory [12]. TLR-4 signalling is typically triggered by products of viruses and bacteria. However, TLR-4 also participates in the recognition of endogenous molecules that are released upon tissue damage, as may occur during ischaemia/reperfusion injury [13]. To test whether EPO synergizes with TLR-4 engagement in increasing the allostimulatory capacity of DCs, DCs emerged from BM culture were stimulated with LPS and 24 h later treated with EPO (Fig. 1, setting 2, LPS/EPO-DCs). Compared to untreated DCs, DCs stimulated by LPS (LPS-DCs), were characterized by a more activated phenotype, namely: an increased percentage of CD86⁺ cells (Fig. 2a,b) and up-regulation of mRNAs for IL-6, IL-12, IL-23, CD40 and CCR7 (Fig. 2c). As expected, LPS-treated DCs released high amounts of IL-12 (Fig. 5a) and showed a very strong allostimulatory capacity compared to untreated DCs (Fig. 5b,c). EPO, when added after LPS stimulation, did not modify the allostimulatory activity of LPS-treated DCs (Fig. 5b). Conversely, when EPO was added to the medium of DC culture 24 h before LPS treatment (Fig. 1, setting 3, EPO/LPS-DCs), a synergistic effect was observed, as EPO/LPS-DCs displayed a stronger allostimulatory activity than LPS-DCs. Indeed, T cell proliferation towards EPO/LPS-DCs and the frequency of IFN-γ-producing T cells were significantly higher than those observed with either EPO-treated or LPS-treated DCs (Fig. 5b,c). Additional experiments, performed with recombinant rat EPO (50 and 500 ng/ml), confirmed that EPO and LPS displayed a synergistic effect (Fig. 5d,e). Consistent with allostimulatory activity, FACS analysis showed the highest percentage of CD86⁺ and MHC-II⁺ cells when DCs were treated with EPO before LPS stimulation (EPO/LPS-DCs) compared to EPO-treated or LPS-treated DCs alone (Fig. 2a,b). Similarly, mRNA expression of CD40 and CCR7 was higher in EPO/LPS-DCs than in DCs treated with either EPO or LPS alone (Fig. 2c). Finally, EPO/LPS-DCs displayed significantly higher mRNA expression for IL-12 and IL-23 than that observed in either EPO-treated or LPS-treated DCs alone (Fig. 2c). Consistent with mRNA expression, ELISA assay showed that EPO/LPS-DCs produced a higher amount of IL-12 than that released by either EPO-DCs or LPS-DCs (Fig. 5a). At variance, IL-6 expression in EPO/LPS-DCs was comparable to that measured in LPS-stimulated DCs (Fig. 2c).

Fig. 5.

Erythropoietin (EPO) potentiates the effect of Toll-like receptor (TLR)-4 stimulation on dendritic cells (DCs). (a–c) Untreated DCs, lipopolysaccharide (LPS)-stimulated DCs, EPO-DCs and EPO/LPS-DCs [obtained by incubating bone marrow (BM) cultures with darbepoetin as described in Fig. 1] were harvested at day 11 of culture and used as stimulators in mixed leucocyte reaction (MLR) with allogeneic (from Lewis rat) lymph node cells. (a) Interleukin (IL)-12 production, evaluated in MLR culture medium by enzyme-linked immunosorbent assay (ELISA). Data are mean ± standard error (s.e.), n = 4 independent experiments, *P < 0·05 versus untreated DCs. °P < 0·05 versus all the other conditions. (b) T cell proliferative response evaluated by [³H]-thymidine incorporation and expressed as counts per minute (cpm). Data are mean ± s.e., n = 3 independent experiments, *P < 0·05 versus untreated DCs. °P < 0·05 versus all the other conditions. (c) Frequency of interferon (IFN)-γ-producing T cells assessed by enzyme-linked immunospot (ELISPOT) assay. Data are mean ± s.e., n = 3 independent experiments, *P < 0·05 versus untreated DCs. °P < 0·05 versus all the other conditions. Syngeneic control: MLR with untreated DCs and syngeneic (from Brown Norway rat) lymph node cells. (d,e) Untreated DCs, LPS-stimulated DCs, EPO-DCs and EPO/LPS-DCs (obtained by incubating BM cultures with rat recombinant EPO as described in Fig. 1) were harvested at day 11 of culture and were used as stimulators in MLR with allogeneic (from Lewis rat) lymph node cells. (d) T cell proliferative response evaluated by [³H]-thymidine incorporation and expressed as cpm. Data are mean ± s.e., n = 3 independent experiments, *P < 0·05 versus untreated DCs. °P < 0·05 versus all the other conditions; 50: rat EPO 50 ng/ml; 500: rat EPO 500 ng/ml. (e) Frequency of IFN-γ-producing T cells assessed by ELISPOT assay. Data are mean ± s.e., n = 3 independent experiments, *P < 0·05 versus untreated DCs. °P < 0·05 versus all the other conditions. 50: rat EPO 50 ng/ml; 500: rat EPO 500 ng/ml. Syngeneic control: MLR with untreated DCs and syngeneic (from Brown Norway rat) lymph node cells. (f) TLR-4 mRNA expression. TLR-4 mRNA analysis in untreated-DCs (harvested at days 7 and 8 of BM culture) and EPO-DCs (harvested at day 8 of bone marrow culture, EPO, darbepoetin, was used at 500 ng/ml) was performed by real-time polymerase chain reaction (PCR). The cDNA content in each sample was calculated by the ΔΔCT technique, using as calibrator the cDNA expression in untreated-DCs harvested at day 7. Data are mean ± s.e., n = 3. *P < 0·05 versus calibrator.

In the search for the mechanism by which EPO renders DCs more susceptible to LPS activation, we focused our attention on the LPS receptor TLR-4. Analysis of TLR-4 mRNA by real-time PCR in DCs emerged from BM showed no differences among cells picked up at days 7 and 8 of BM culture. At variance, BM cells collected at day 8 after 24 h treatment with 500 ng/ml EPO showed a doubled TLR-4 mRNA expression compared to untreated cells (Fig. 5f).

Discussion

DCs originate from BM and, at an immature status, they patrol peripheral tissues to sense danger signals and activate specific immune response [14,15]. The capacity of DCs to activate T cells is acquired during a maturation programme [16]. The current study shows that differentiating BM-derived DCs but not fully differentiated DCs are targets of EPO. Indeed, DCs emerging from rat BM, but not splenic DCs, express the EPO receptor (Epo-R) and respond to EPO stimulation displaying a more activated phenotype with increased CD86, CD40 and IL-12 expression levels. Our results are consistent with recent reports from Neumann et al., who documented that mouse BM-derived DCs express the mRNA for the EPO receptor and that EPO-R stimulation in DCs leads to the activation of mitogen-activated protein kinase (MAPK), protein kinase B (AKT) and nuclear factor kappa B (NF-κB) pathways [8]. Such EPO-stimulated DCs showed increased expression of MHC-II and co-stimulatory molecules and increased IL-12 secretion [7]. In addition, in this study we found that EPO-treated BM-derived DCs display higher allostimulatory capacity on T cells than untreated DCs.

In view of their potent immunostimulatory functions that activate antigen-specific T cell responses, DCs play an important role in the context of cancer immunosurveillance. However, tumour cells are able to escape immune system attacks. The factors contributing to tumour immune escape are still not completely defined; however, inadequate presentations of tumour-associated antigens by host DCs is recognized as a mechanism that allows tumour progression [17]. Results presented here suggest that exogenous EPO administration could potentiate the immunostimulating activity of DCs in cancer patients in vivo. That this may be the case is supported in a study by Prutchi-Sagiv et al. [18] on patients affected by multiple myeloma, who were characterized by low reactivity and proliferation potential of the CD4⁺ and CD8⁺ T cells. When patients were treated with EPO to correct anaemia, T cell functions improved, an effect that was attributed to a stimulating action of EPO on macrophages and DCs [18]. However, it is important to point out that the potential immune-mediated benefits of high EPO administration in cancer might be offset by the potential pro-proliferative effects of administering such a huge dose.

We also found that EPO up-regulates TLR-4 in differentiating DCs, rendering these cells more sensitive to stimulation by the TLR-4 ligand LPS. BM-derived DCs treated with EPO before LPS stimulation were strongly allostimulatory and expressed CCR7, CD86, CD40, IL-12 and IL-23 at higher levels than DCs stimulated with LPS alone. Such events in vivo would result in DCs very efficiently equipped to move towards lymph nodes and drive a T helper type 1 (Th1) and Th17 T cell response through IL-12 and IL-23, respectively. Thus, EPO may act as an additional danger signal in concert with TLRs linking the innate to the adaptive immune function of DCs.

Tissue injury following ischaemia/reperfusion causes the release of endogenous molecules such as heat shock proteins, hyaluronic acid and high mobility group box 1 (Hmgb1) proteins which engage TLRs [13], as well as of haematopoietic cytokines which can accelerate revascularization, and among them, EPO [19,20]. Activation of both tissue-resident and recruited circulating DCs is a key event in tissue injury following ischaemia/reperfusion [21]. The results presented here show that in vitro EPO up-regulates TLR-4 in DCs and renders these cells more sensitive to TLR-4 stimulation. Whether ischaemia/reperfusion injury is a possible scenario in which endogenous EPO potentiates TLR-induced DC activation is worth investigating. Such an effect may have a negative impact on allograft outcome, as ischaemia/reperfusion injury is an unavoidable consequence of transplant surgery and DCs are the main inducers of allograft rejection by presenting alloantigens to recipient T cells through both the direct (donor DCs) and the indirect (recipient DCs) pathways [22,23].

It is tempting to speculate that EPO could act as an additional danger signal in concert with TLR-4 engagement, according to the proposed terminology by Matzinger [24]. Thus, EPO, beyond its well-documented erythropoietic and cytoprotective effects, also emerges as an immune modulator.

Disclosure

The Authors declare no financial conflict of interests.