Inhibitory effects of rat bone marrow-derived dendritic cells on naïve and alloantigen-specific CD4+ T cells: a comparison between dendritic cells generated with GM-CSF plus IL-4 and dendritic cells generated with GM-CSF plus IL-10

Background Unlike mouse immature bone marrow (BM)-derived dendritic cells (DC), rat immature BMDC have not been thoroughly characterised in vitro for the mechanisms underlying their suppressive effect. To better characterise these mechanisms we therefore analysed the phenotypes and immune inhibitory properties of rat BMDC generated with GM-CSF plus IL-4 (= IL-4 DC) and with GM-CSF plus IL-10 (= IL-10 DC). Results Both IL-4 DC and IL-10 DC exhibited lower surface expression of MHC class II and costimulatory molecules than mature splenic DC. They had a strong inhibitory effect on responsive T cells in vitro and despite their weak function as antigen-presenting cells they induced anergic T cells. However, only anergic T cells induced by IL-4 DC had a suppressive effect on responsive T cells. Induction of suppressive/tolerogenic T cells by IL-4 DC required direct contact between antigen-specific T cells and IL-4 DC. In addition, IL-4 DC and IL-10 DC prolonged allograft survival in an antigen-specific manner. Conclusion A unique phenotype of immature BMDC was isolated from the cultures. The mechanisms underlying the suppressive effect may be caused by their inability to deliver adequate costimulatory signals for T-cell activation. In addition, IL-4 DC but not IL-10 DC induce anergic T cells with suppressive function. This indicates that IL-4 DC and IL-10 DC may differ in the quality of their costimulation although no differences in the surface expression of costimulatory molecules were found.


Background
In recent years it has become clear that dendritic cells (DC) are not only potent inducers of adaptive immune responses, but also essential mediators in the induction and maintenance of T-cell tolerance [1]. The biological properties of DC depend on their phenotypically distinct states of development [2]. Their delaying effect on allograft rejection has been demonstrated in several rodent models [reviewed in [3]].
Mouse and human DC have both been studied thoroughly [reviewed in [4]]. Rat DC have been investigated particularly by groups interested in transplantation research [5][6][7]. They were not studied thoroughly, although established culture methods exist for the generation of bone marrow-derived rat DC (BMDC) [8,9]. The maturation of BMDC varies from species to species despite comparable culture conditions. In mice, for example, low doses of granulocyte macrophage colony stimulating factor (GM-CSF) combined with interleukin (IL)-4 induce the formation of mature BMDC [10], whereas in rats the same combination produces immature BMDC [11]. The effect of GM-CSF and IL-10 on the generation of rat BMDC is not clearly known.
In the present study we examined the ability of IL-4 DC and IL-10 DC to inhibit both the activation of naïve T cells and the restimulation of antigen-specific T cells in vitro. We also analysed their in vivo potential to prolong allograft survival.

Generation of rat BMDC
Femur and tibia bones of young (8-10 weeks) Lewis rats were extracted and disinfected in 70% ethanol. Both ends of the bones were cut and the bone marrow (BM) cells were flushed with 20 ml phosphate-buffered saline (PBS). The BM cells were cultured at a cell density of 5-8 × 10 5 cells/ml in culture dishes (Falcon, Becton Dickinson Biosciences). The RPMI 1640 culture medium [12] was supplemented with 5 ng/ml recombinant rat GM-CSF (R&D Systems, Heidelberg, Germany) and 5 ng/ml recombinant rat IL-4 (Miltenyi Biotech GmbH, Germany) or 5 ng/ml rat IL-10 (Miltenyi Biotech GmbH). On day 6, non-adherent cells and cells growing in clusters were collected.

Cell isolation
Splenic DC (S-DC), naïve and antigen-specific T cells were isolated from Lewis rats as described previously [12]. Allopeptide P1-specific T cells were induced by immunization with P1 [12].

Transwell experiments
Some of the assays were performed in 96-well transwell plates (Corning Life Sciences, The Netherlands). The upper compartment contained immature rat BMDC (1 × 10 4 ), the lower compartment antigen-specific T cells (1 × 10 5 ). After 3 days of culture, the transwells were removed and P1-loaded S-DC (10 4 /well) as well as antigen-specific T cells (1 × 10 5 /well) were added to the lower compartments. The cultures were then incubated for another 3 days and pulsed with 0.5 μCi/well [ 3 H]-thymidine for the last 6 h of culture. The incorporation of [ 3 H]-thymidine was measured as described [12].
Morphology and immunostaining of IL-4 DC and IL-10 DC

Heterotopic heart transplantation with and without antigen-loaded BMDC
The animal experiments were conducted in accordance with European, national and institutional animal care policies. Ten million BMDC were incubated with 20 μg P1 for 30 min in 500 μl PBS, washed 2 times with PBS and transferred intravenously via the penile vein into Lewis rats under full anesthesia with isoflurane 1 day before transplantation. Fully vascularised, heterotopic heart transplantation was performed, and graft survival was monitored as described [12].

IL-4 DC and IL-10 DC exhibited an identical phenotype
On approximately day 3 of culture low adherent cell clusters were obvious and both the number and size of the clusters increased during culture (Fig. 1A). The BMDC isolated from these clusters on day 6 were positive for Ox62 (a marker for rat DC), Ox6 (anti-MHC class II), and the macrophage marker ED1 (CD68) (Fig. 1B-D). More than 77% of the cells isolated from the clusters were double positive for Ox62 and Ox6 without further purification.
The flow cytometric analysis revealed that IL-4 DC and IL-10 DC expressed MHC II and the costimulatory molecules CD40, CD80 and CD86 at nearly identical levels on their surface (Fig. 2).

IL-4 DC and IL-10 DC neither activated naïve T cells nor restimulated antigen-specific T cells
Naïve T cells did not proliferate in the presence of IL-4 DC and IL-10 DC, whereas mature S-DC induced a strong Tcell proliferation (Fig. 3A). IL-4 DC and IL-10 DC loaded with the allogeneic peptide P1 [12] were not able to restimulate P1-specific T cells (Fig. 3B). Different numbers (10 3 , 10 4 and 10 5 ) of P1-loaded IL-4 DC or IL-10 DC suppressed the proliferation of antigen-specific T cells induced by P1-pulsed S-DC (Fig. 3C, D). For both BMDC types, IL-12-specific mRNA was not detectable by RT-PCR (Additional File 1, Table 1). This may indicate that they revealed properties of immature DC.
In the BMDC-uninfluenced T-cell proliferation assay, the strongest increase in proliferation of antigen-specific T cells occurred between days 2 and 3 of the 3-day culture. The addition of P1-pulsed IL-4 DC or IL-10 DC to the cultures on day 2 halted this strong increase in T-cell proliferation within 24 h (Fig. 3E, F).

IL-4 DC-T and IL-10 DC-T showed anergic properties
P1-specific T cells were named IL-4 DC-T when incubated with IL-4 DC and IL-10 DC-T when incubated with IL-10 DC. Following purification, the DC-T were transferred to 10 4 P1-pulsed mature S-DC and their proliferation rate after 72 h was low (Fig. 4). The anergic-like effect of IL-4 DC-T and IL-10 DC-T could be negated by adding exogenous IL-2 (27 ng/ml) to the cultures (Fig. 4).

IL-4 DC-T had an inhibitory effect on antigen-specific T cells
Different numbers (10 3 , 10 4 and 10 5 ) of IL-4 DC-T were added to assays containing P1-loaded S-DC and freshly isolated antigen-specific T cells. Measurement of proliferation revealed that IL-4 DC-T had a dose-dependent inhibitory effect on T-cell proliferation (Fig. 5A). Antigenspecific T cells cultured with IL-4 DC-T in transwell plates did not reduce the control T-cell proliferation (Fig. 5B). Antigen-specific T cells must have direct contact with IL-4 DC in precultures to have an inhibitory effect on T-cell proliferation in second cultures. The IL-10 DC-T, in con-  trast, had no influence on the proliferation of antigen-specific T cells (not shown).

P1-loaded IL-4 DC and IL-10 DC prolonged antigenspecific cardiac allograft survival
Ten million of P1-loaded BMDC administered intravenously to each Lewis rat 1 day before they received heart allografts let them survive 2 days longer than those animals transfused with unpulsed BMDC, which had no effect on allograft survival (Table 2 and Fig. 6). The survival of allografts from third party donors (Brown-Norway rats) was not affected by the transfer of P1-pulsed IL-4 DC and IL-10 DC (Table 2).

Discussion
In the present study, we compared the effects of 2 types of rat BMDC on the proliferation of naïve and antigen-specific T cells in vitro and on the survival of allogeneic heart allografts. Both DC types displayed lower surface expression of MHC class II and costimulatory molecules compared to mature splenic DC. IL-4 DC and IL-10 DC had a strong inhibitory effect on responsive T cells. This suppressive effect was detectable within 24 h after the BMDC were added to cultures with antigen-specific T cells and mature splenic DC (Fig. 3E, F). To our knowledge this is the first description of the time course BMDC needed to suppress T-cell proliferation. We found that antigen-specific T cells became anergic after incubation with IL-4 DC and IL-10 DC. The same effect has been described for human IL-10 modified DC on CD4+ T cells [13]. Anergic T cells isolated from cultures with P1-loaded IL-4 DC suppressed the DC-mediated activation of responsive T cells in a cell count-dependent manner. They share this sup-pressive effect with regulatory T cells [14]. We hypothesise that IL-4 DC and IL-10 DC may differ in the quality of their costimulation, with IL-4 DC inducing suppressive IL-4 DC-T whereas IL-10 DC do not. It should be emphasised, however, that we found no differences in the surface IL-4 DC-T and IL-10 DC-T do not proliferate and demon-strate anergy  Primers specific for rat IL-12p35 was purchased from Invitrogen/Biosource. Total RNA was isolated from 10 6 cells with 1 ml of the ready-to-use reagent Trizol Reagent (Invitrogen GmbH, Germany) according to the manufacturer's recommendations. Reverse transcription of 5 μl RNA was performed using the GeneAmp RNA-PCR-Kit (Applied Biosystems GmbH, Germany). Five μl cDNA was amplified using Gold AmpliTaq DNA-Polymerase (0.05 U/μl) and the specific primers (5 μmol/L each) mixed in nuclease free water (Promega GmbH, Germany) to an end volume of 50 μl per sample.
expression of costimulatory molecules between the two BMDC types.
The results of transwell experiments showed that the contact between IL-4 DC and the antigen-specific T cells is a prerequisite for inducing suppressive IL-4 DC-T. However, we made no attempt in this study to determine whether the anergic IL-4 DC-T mediate their suppressive effect via cell-cell contact or by soluble factors. Vendetti et al., for example, reported that inhibition mediated by anergic murine T cells is dependent on cell-cell contact [15]. They also described an inhibitory effect of anergic T cells on the antigen-presenting function of mature DC.
IL-4 DC and IL-10 DC, loaded with allopeptide P1, prolonged allograft survival (Fig. 6). In addition, survival time can be improved by increasing the number of transferred cells. The application of 30 million P1-pulsed IL-10 DC, for example, prolonged survival time to a median of 10.6 ± 0.8 days (not shown) from the 9.5 ± 0.8 days achieved with 10 million P1-pulsed IL-10 DC. Allograft survival was prolonged only when the BMDC were pulsed with the immunodominant allogeneic peptide P1 involved in allograft rejection [5]. Our results accord with those of Chowdhury et al. [16], who showed that the presentation of allogeneic peptides by tolerogenic thymic rat DC greatly prolongs allograft survival. Compared to the results of Chowdhury et al., our 2-3 day prolongation of allograft survival may seem meager, but considering the strength of the allogeneic immune response induced by alloreactive T cells and the fact that no immunosuppressive drugs were used, our findings appear very promising.

Conclusion
The data suggest that rat IL-4 DC and IL-10 DC have suppressive/regulatory properties comparable to those described for immature mice BMDC. They demonstrate a strong inhibitory effect on responsive T cells, probably the consequence of a reduced surface expression of costimulatory molecules paired with the inability to deliver adequate costimulatory signals to T cells. IL-4 DC and IL-10 DC are identical in phenotype and in some of their effects, but they are different in their capacity to induce suppressive T cells. IL-4 DC induce T cells with suppressive/regulatory function whereas IL-10 DC do not. This may indicate that IL-4 DC and IL-10 DC differ in the quality of