Autologous apoptotic T cells interact with dendritic cells, but do not affect their surface phenotype or their ability to induce recall immune responses

Abstract

Dendritic cells (DCs) play an important role in determining immunogenicity and the subsequent immune response. They may also have a role in maintaining peripheral tolerance to self-antigens by initiating an immune response only in the context of danger signals released from cells during stress, damage or death. These signals may originate from surrounding T cells as well as from other cells. Therefore, in this study the effect of autologous T cell injury on DC morphology and function has been investigated. Co-incubation of apoptotic or necrotic T cells with immature DCs altered their morphology towards a more mature appearance, with more cells showing activation as judged by spreading and formation of arborizing long processes. The apoptotic autologous T cells were rarely phagocytosed by immature DCs, compared to macrophages. The DC surface phenotype was not affected by the co-incubation with autologous injured T cells. The ability of DCs to elicit a secondary immune response was not altered by exposure to autologous injured T cells. These findings suggest that DC can continue to function in T cell activation, rather than in tolerogenic mode, even in the presence of large numbers of dying autologous T cells, such as may be present in the aftermath of an acute antiviral response.

INTRODUCTION

Dendritic cells (DCs) play an important role in both innate and adaptive immune responses. Immature DCs in peripheral tissues act as sentinels, sampling their microenvironment for the presence of antigen and acquiring antigen by phagocytosis, macropinocytosis and adsorptive endocytosis. Then they migrate to lymph nodes to initiate immune responses. Concomitant with migration DCs mature, as evidenced by increase in surface expression of adhesion and co-stimulatory molecules and by reduced capacity for antigen uptake [1]. A wide variety of factors, both exogenous and endogenous, can induce this DC maturation [2,3]. In addition signals from T cells, e.g. mediated via CD40–CD154 interaction, can have similar effects. The resultant mature interdigitating DCs are the potent inducers of adaptive antigen-specific immune responses, either primary or secondary, depending on the nature of the previous T cell repertoire.

Although this DC role in immune induction is widely accepted, some aspects of how they determine patterns of outcome remain controversial. The observation that immature DCs can phagocytose apoptotic cells [4–6] has led researchers to propose that this might be one way that peripheral tolerance to self-antigen is maintained. In this context engulfment of autologous apoptotic cells by DCs would occur in the absence of an additional DC maturation signal, hence avoiding the consequent induction of an active autoimmune response [7]. In contrast, the ‘danger hypothesis’ suggested that endogenous mediators released from cells during any form of stress − including apoptosis or necrosis − could result in DC activation and initiation of an immune reaction [8]. This can be either a CD4 [8,9] or a CD8 [cytotoxic T lymphocyte (CTL)] T cell response [9–11] and DCs have been shown recently to acquire antigens from both living and dead cells, and cross-present them to CTL [9,12,13].

In vivo, at the periphery, and even more in lymphoid tissues, DC are surrounded by T cells. Thus far the question as to whether or not apoptosis and/or necrosis of the T cells themselves can affect the DC maturation and antigen presentation pathways has not been addressed. Nevertheless, the presence of large numbers of apoptotic T cells is a common feature in the period immediately following acute antiviral responses (e.g. HIV-1, EBV) as the T cell population reverts to a homeostatic level [14]. T cells are also sensitive targets of both immunosuppresssive drugs and radiation.

Therefore, in this study we have examined the effect of preincubation of autologous injured (apoptotic or necrotic) T cells on DCs, examining changes in DC morphology after exposure to injured autologous T cells, uptake of these T cells by the DCs compared to control macrophages and DC phenotype. Furthermore, the ability of DCs to induce secondary immune responses to recall antigens has been investigated. The results provide evidence which suggest that there is preservation of DC antigen presenting function despite apparent ‘sampling’ contact with the T cells and surrounding T cell turmoil and death.

SUBJECTS AND METHODS

Healthy volunteers

Local ethical committee approval was obtained to perform this study and all healthy volunteers (n = 12) provided informed consent prior to venesection.

Antibodies

Mouse antihuman monoclonal antibodies (MoAbs) were prepared and diluted in AIM-V medium (Gibco, Paisley, UK): CD2 (RPA-2·10, IgG1, purified) (Insight Biotechnology Limited, Wembley, UK), CD3 (UCH-T1, IgG1, culture supernatant), CD14 (HB246, IgG2b, culture supernatant), HLA-DR (L243, IgG2a, culture supernatant) (all kindly provided by Professor P.C.L. Beverley, Edward Jenner Institute for Vaccine Research, Compton, UK), CD19 (BU12, IgG1, culture supernatant) (kind gift from D. Hardie, Department of Immunology, University of Birmingham Medical School, UK) and CD36 (CB38, IgM k, purified) (Pharmingen, San Diego, USA). Phycoerythrin (PE)-conjugated mouse antihuman MoAbs used were: IgG1 (679·1Mc7), HLA-DR (b8·12·2, IgG2b k), CD40 (mab 89, IgG1), CD80 (mab 104, IgG1), CD86 (HA5·2B7, IgG2b k), CD25 (B1·49·9, IgG2a k) and CD83 (HB15A, IgG2b) (all from Insight Biotechnology Limited).

Generation of DCs

Monocyte-derived human dendritic cells were generated from peripheral blood mononuclear cells (PBMCs) [15], with the following modifications. The non-adherent cells were retrieved on day 4, instead of day 3, and resuspended in 2 ml AIM-V medium supplemented with 0·075% sodium bicarbonate (Gibco), 0·05 mm 2-mercaptoethanol, 100 U penicillin–streptomycin/ml and 2·4 mm l-glutamine (AIM-V CM), instead of RPMI-1640 medium supplemented with 10% fetal calf serum (FCS). AIM-V CM was used for all subsequent studies. The DCs were incubated with CD3, CD2 and CD19 MoAbs (4°C, 30 min), and negative immunomagnetic depletion performed as described previously [15]. The resultant DC population was>95% pure as determined by flow cytometry. The DCs were recultured at 10⁶ DCs per well for functional and phenotypic analysis, or at 2 × 10⁵ DCs per well for confocal microscopy in fresh AIM-V CM supplemented with fresh cytokines until day 6. Day 6 DCs, generated in AIM-V CM, had the characteristic phenotype of immature DCs. They were CD1a^pos, CD14^low, HLA-DR^pos, CD40^pos, CD80^inter, CD86^inter, CD25^neg and CD83^neg, as described previously [16].

Generation of macrophages

Macrophages were prepared using a similar protocol except that from day 1 onwards adherent cells were cultured in the presence of granulocyte-macrophage colony-stimulating factor (GM-CSF) only. On day 4, the non-adherent cells were removed and 0·05% trypsin in PBS was added for 5 min. The reaction was stopped with excess AIM-V CM. After two washes the cells were resuspended in AIM-V CM, and the macrophages purified by negative immunomagnetic depletion (as above) followed by reculture at 2 × 10⁵/well in AIM-V CM with GM-CSF until day 6. The surface phenotype of the macrophages was CD1a^neg, CD14^high, HLA-DR^pos, CD40^pos, CD80^inter, CD86^inter, CD25^neg and CD83^neg.

Preparation of autologous and allogeneic T cells

On day 4, two-thirds of the autologous non-adherent cells (which had been frozen slowly to –70°C in cold FCS/10% dimethyl sulphoxide on day 1) were retrieved and purified by negative immunodepletion using CD14, CD19 and anti-HLA-DR MoAbs. T cells were cultured in RPMI CM (containing 10% FCS) supplemented with 5–10 IU/ml of interleukin (IL)-2 (Peprotech, Rocky Hill, NJ, USA). On day 5, the T cells were washed twice and resuspended in AIM-V CM (without 0·05 mm 2-mercaptoethanol). These cells were the viable T cells used on day 6 (see below). On day 7, the remaining frozen autologous T cells were retrieved and purified for use in the functional assays. Allogeneic T cells, used in mixed leucocyte reactions (MLR), were also retrieved and purified on day 7.

Induction of T cell apoptosis and necrosis

Apoptosis was induced by the incubation of 100 µm hydrogen peroxide (Sigma Chemical Co, St Louis, MO, USA) with T cells for either 18 h from day 5 (late apoptotic) or 1·5 h on day 6 (early apoptotic). Over 95% of the T cells incubated with hydrogen peroxide on day 5 were late apoptotic by 18 h, as judged by Annexin V and propidium iodide staining (TACS™ Annexin V-FITC apoptosis detection kit, R&D Systems, Minneapolis, USA). In contrast, only 25–40% of the early apoptotic T cell sample were Annexin V-positive at the time of co-culture with DCs. However, monitoring showed that during the next 18 h over 95% of T cells would become late apoptotic. The viable T cell population (T cells cultured in IL-2 until the co-culture period) contained less than 25% of T cells in apoptosis. T cell necrosis was induced in 0·2 ml AIM-V CM volume on day 6 by four repeated freeze–thaw cycles (from −70°C to 37°C). Over 95% of these cells were trypan blue positive.

Cell line culture conditions and induction of apoptosi

U937 monoblastoid leukaemia cells were obtained from the American Tissue Culture Collection (ATCC, Manassas, VA, USA) and cultured in RPMI CM. Cell cultures were tested for the presence of mycoplasma infection and found to be negative (Mycoplasma Detection Kit, ATCC). On day 5, prior to the induction of apoptosis, the cells were washed twice and resuspended in AIM-V CM (without 0·05 mm 2-mercaptoethanol). Apoptosis was induced by the incubation of 200 µm hydrogen peroxide with U937 cells for 18 h. Over 95% of these cells were trypan blue positive.

DC co-culture with viable, apoptotic or necrotic T cells

DCs were co-cultured with viable, apoptotic or necrotic T cells for 3 h (phagocytosis and functional assay; DC : T cell ratio 1 : 10), 18 h (DC morphology, phagocytosis and functional assay; DC : T cell ratios of 1 : 1 and 1 : 10) or 48 h (for phenotypic analysis; DC : T cell ratios of 1 : 1 and 1 : 10). AIM-V CM alone was added to one well to act as a control and in some experiments apoptotic U937 cells were added at a DC : U937 cell ratio of 1 : 4 to provide an additional control. In the DC phenotyping and morphology experiments 100 ng/ml of lipopolysaccharide (LPS) (Salmonella Minnesota, Sigma) was added to one well to act as a positive control. For phagocytosis studies, late apoptotic T cells were stained with 10 µg/ml of 7-AAD/10⁶ cells in AIM-V CM (30 min, 37°C) in the dark. The cells were washed three times prior to incubating with the DCs or the macrophages at either 4°C or 37°C for up to 18 h. Initial experiments, using flow cytometry, revealed that there was no significant transfer of dye at 4 or 37°C between the stained apoptotic T cells and unstained immature DCs (as detected using the FL-3 channel of the FACS machine).

Confocal microscopy

Confocal microscopy was performed on a Bio-Rad Confocal Microscope (Hercules, CA, USA). Images of DCs, DCs and T cells, and macrophage and T cells were taken using a ×20 objective at several time-points over the 18-h period, and analysed using Confocal Assistant software.

DC phenotypic analysis

DC phenotypic analysis was performed on day 8. DCs were incubated on ice for 15 min in Hanks's buffered saline solution (HBSS) (Gibco) with 0·1% sodium azide and 0·5% FCS prior to the addition of the PE-conjugated MoAbs. Control samples included cells alone and an isotype control. After 30 min DCs were washed and fixed with 0·5% formaldehyde in HBSS. The samples were acquired by flow cytometry (FACScan; Becton Dickinson, Mountain View, CA, USA) and analysed by WinMDI software (http://facs.scripps.edu/). For each sample at least 5000 events gated on DCs were analysed.

DC functional studies

Following the 3 or 18 h co-culture period DCs were recovered, centrifuged over lymphoprep and repurified by immunomagnetic depletion, as outlined above. DCs incubated with U937 cells were treated in an identical fashion apart from the addition of culture medium instead of monoclonal antibodies during the negative depletion step. For recall antigen studies assays were performed in triplicate in 96-well plates. DCs were used at 10⁴, 2·5 × 10³ and 1·25 × 10³/well. Responders were 10⁵ autologous T cells per well; 500 U/ml of purified protein derivative (PPD; Evans Medical Limited, Leatherhead, UK) or 10 µg/ml of tetanus toxoid (TT; Pasteur Merieux Connaught, France) was added. Controls included DCs alone, autologous T cells alone and DCs with autologous T cells in the absence of PPD or TT. Assays were for 6 days at 37°C in 5% CO₂. Thymidine incorporation was measured by adding 0·037 Mbq/well of methyl-[³H]thymidine (ICN Biomedical, High Wycombe, UK) for the final 16 h of culture, as previously described [15]. MLR assays were performed in a similar fashion, using the same DC concentrations and 10⁵ allogeneic T cells per well. Controls included DCs alone and allogeneic T cells alone.

Statistical analysis

The functional assays were analysed using the Mann–Whitney test. A P-value of <0·05 was taken as the level of significance.

RESULTS

Morphological features of DC co-culture with autologous apoptotic and necrotic T cells

Initial experiments documented T cell apoptosis (Fig. 1a,b,c) and then examined the interaction of day 6 DCs with viable, apoptotic (early and late) or necrotic autologous T cells by light microscopy. On day 7 control DCs had the typical morphological appearance of immature DCs: they were predominantly non-adherent and had few dendritic cell processes (Fig. 1d). DCs co-cultured with autologous viable T cells were more adherent and activated in appearance. These T cells were distributed randomly throughout the culture (Fig. 1e). In contrast, DCs incubated with apoptotic T cells were more adherent and activated than control DCs (both DCs incubated with culture medium alone or with viable T cells); they formed more dendritic processes per cell and had longer dendrites (see Fig. 1f,g). The apoptotic T cells clustered around DCs, both around the cell body and the dendritic processes (Fig. 1h,i). Few free apoptotic T cells were observed distributed between DCs. A CD36 (a putative receptor involved in DC apoptotic cell phagocytosis) MoAb did not inhibit this clustering of apoptotic T cells around the DC (data not shown). In contrast, DCs incubated with necrotic T cells were only slightly altered in appearance (Fig. 1j) and were similar in morphology to DCs co-incubated with viable T cells (Fig. 1e). LPS-induced DC maturation was associated with features of mature DCs: the DCs were adherent and extended multiple, long dendritic processes (Fig. 1k). Both increased dendrite extension and cluster formation (neither reported previously for autologous DC/apoptotic T cell interactions) provide further evidence that DC might respond directly to ‘apoptosis-specific’ molecular changes, and the functional consequences of this interaction were investigated further below.

Fig. 1.

Morphological features of DCs following 18 h co-culture with autologous apoptotic and necrotic T cells. Autologous T cells were retrieved on day 4 and following negative selection were cultured in the presence of IL-2 until day 5 or day 6. T cell apoptosis was induced by 100 µm of hydrogen peroxide for 18 h on day 5 (late apoptotic T cells) or for 1·5 h on day 6 (early apoptotic T cells). (a) Less than 25% of the retrieved autologous T cells cultured in the presence of IL-2 until the co-culture period (viable T cells) were apoptotic, as judged by annexin-V and propidium iodide staining. (b,c) The induction of early and late apoptotic T cells by hydrogen peroxide, respectively. Immature day 6 DCs were incubated for 18 h with either AIM-V CM alone (d), viable T cells (e), early apoptotic T cells (f), late apoptotic T cells (g,h,i), necrotic T cells (j) or 100 ng/ml LPS (k). The DC : T cell ratio used was 1 : 10. Images (d–k) were obtained by confocal microscopy (scale bar = 5 µm). Arrow in (j) indicates necrotic T cell debris. Representative experiment, n = 7. Similar results were observed in the presence of human serum, n = 3.

Phagocytosis of autologous apoptotic T cells by macrophages but not by DC

To evaluate whether interaction with apoptotic cells was a prelude to phagocytosis, autologous apoptotic T cells were stained with 7-amino actinomycin D (7-AAD) dye that incorporates predominantly into apoptotic DNA and can be detected by flow cytometry [17]. Labelled T cells were incubated with immature day 6 DCs for up to 18 h at 4°C and 37°C. There was little evidence by flow cytometry over a 3-h time-course that the DCs had phagocytosed the apoptotic T cells (Fig. 2). The percentage of DCs incubated with labelled apoptotic T cells that had increased their fluorescent intensity above the 98 percentile of the distribution of control DCs was <3%. The maximal increase in median fluorescent intensity was <20. By confocal microscopy, only very occasional DC − less than 1% − incubated at 37°C with apoptotic T cells stained with 7-AAD. In contrast, macrophages incubated with the same T cells at 37°C did phagocytose apoptotic T cells, as determined by the presence of macrophage staining with 7-AAD (Fig. 3). No significant dye transfer was observed at 4°C (data not shown). These results suggest therefore that under these conditions immature DC rarely phagocytose autologous apoptotic T cells.

Fig. 2.

Autologous apoptotic T cells were phagocytosed rarely by DCs. Day 6 DCs were co-cultured with late apoptotic T cells (ratio 1 : 10) stained with 7-AAD at 4°C and 37°C. (a) Minimal uptake of 7-AAD by DCs was observed by flow cytometry over a 3-h time-course, represented as overlay histograms (open histograms = DCs with AIM-V CM alone; solid histograms represent DCs incubated with labelled apoptotic T cells). Marker R1 was set at the 98th percentile of the distribution of control DCs. (b,c) Quantitative illustration of these data by demonstrating either the percentage increase in R1 or the change in the median fluorescence intensity, respectively (closed diamond = DCs with AIM-V CM alone; closed triangle = DCs incubated at 4°C and closed square = DCs incubated at 37°C in the presence of labelled apoptotic T cells). Representative experiment, n = 5. By confocal microscopy, <1% of DC, incubated at 37°C, stained with 7-AAD. Similar results were observed for DCs when cultured in human serum (data not shown).

Fig. 3.

Autologous apoptotic T cells were phagocytosed by macrophages. Day 6 macrophages were co-cultured for 3 h with late apoptotic T cells (ratio 1 : 10) stained with 7-AAD at 4°C and 37°C. There was phagocytosis as determined by 7-AAD uptake by macrophages. (a,b) Two representative fields of view from one experiment of five. Scale bar = 5 µm.

The phenotypic characteristics of day 6 DC incubated with autologous apoptotic and necrotic T cells for 48 h

Following the observation that DCs co-cultured with apoptotic autologous T cells had a more activated morphology than control DCs, surface phenotyping of the various DC groups was performed at two DC : T cell ratios [1 : 1 (three independent experiments) and 1 : 10 (four independent experiments)]. The results were extremely comparable. On two occasions the DCs co-cultured with early apoptotic autologous T cells had a small increase in both CD25 and CD83 expression, indicating slight DC activation; in these experiments an increase in DC HLA-DR expression was observed on exposure to necrotic T cells (Fig. 4). No changes in any of the surface molecules measured were observed with late apoptotic cells. None of these changes were statistically significant. In contrast, DCs matured with LPS up-regulated surface expression of HLA-DR, CD40, CD80, CD86, CD25 and CD83 (Fig. 4). Thus, in summary, exposure of DCs to the various T cell populations did not induce pronounced changes in DC surface phenotype despite the observed changes in DC morphology.

Fig. 4.

Representative flow cytometry profiles of day 6 DCs co-cultured with autologous apoptotic and necrotic T cells for 48 h. Immature day 6 DCs were incubated for 48 h with either AIM-V CM alone, viable T cells, early apoptotic T cells, late apoptotic T cells, necrotic T cells or 100 ng/ml of LPS. The DC : T cell ratio used was 1 : 10. The data are represented as overlay histograms. Open histograms show DCs only or the appropriate isotype matched control. Solid histograms show staining of gated DCs for the respective MoAb tested. Representative experiment, n = 4.

Pre-incubation of autologous apoptotic or necrotic T cells with immature DC does not impair the ability to elicit secondary immune responses

Following either a 3 h or 18 h incubation of day 6 DCs with either viable, apoptotic or necrotic autologous T cells, any residual T cells were depleted and the functional assays performed using fresh viable purified autologous T cells. The ability to induce secondary immune responses was unchanged irrespective of prior T cell exposure. Figure 5a,b shows representative experiments eliciting recall responses to tuberculin (PPD) and tetanus toxoid (TT) following either a 3 h (Fig. 5a; representative experiment, n = 3) or 18 h incubation (Fig. 5b; representative experiment, n = 6). There was no statistical difference in the proliferative responses observed between any of the preincubated T cell groups and the control after the 18 h co-culture period (Mann–Whitney test − all P-values> 0·05). These results were confirmed further in three independent experiments performed using complete medium containing 10% FCS. A similar result was observed in the MLR reaction, performed in AIM-V CM (three independent experiments). However, DCs incubated for 3 h in the presence of apoptotic leukaemic cells (DC : U937 cell ratio 1 : 4) and then centrifuged over lymphoprep, to remove the apoptotic cells, showed an impaired ability to elicit recall antigen responses [PPD (three independent experiments]; TT [three independent experiments)] and a MLR response (three independent experiments) at the lowest DC cell concentration (data not shown).

Fig. 5.

DCs elicit secondary immune responses following preincubation with autologous apoptotic and necrotic T cells. Day 6 DCs were incubated with AIM-V CM alone (), viable T cells (), early apoptotic T cells (), late apoptotic T cells () or necrotic T cells (▪) for either 3 h [ (a), representative experiment, n = 3) or 18 h (Fig. 5B); representative experiment, n = 6]. The DC : T cell ratio used during the co-culture period was 1 : 10. The DCs were repurified prior to performing the functional assays and titrated with 10⁵ fresh autologous T cells and either 500 U/ml of PPD or 10 mg/ml TT. After 6 days T cell proliferation was determined by methyl-[³H]thymidine incorporation and values represent the total thymidine counts obtained minus the background counted counts ± standard error.

DISCUSSION

Considerable controversy still surrounds the consequences of interaction between DC and dying cells. The demonstration that autologous apoptotic T cells cluster around DCs is consistent with previous reports of receptors for apoptotic cells on the surface of DCs [5]. The increase in dendrite length, following co-culture with apoptotic cells, may reflect an increased opportunity for sampling of the DC antigen microenvironment. Only a small proportion of autologous apoptotic T cells were phagocytosed by these immature DC; these cells may rely on other mechanisms of antigen uptake [18].

The incubation of autologous apoptotic and necrotic T cells with DCs in this culture system did not alter DC surface phenotype. These results are in agreement with those reported by Urban et al. [19], who demonstrated that co-incubation of autologous necrotic cells with DC did not alter DC surface phenotype. In contrast, in mice Gallucci et al. [8] showed that murine bone marrow-derived DCs (MBMDC) co-cultured with necrotic syngeneic fibroblasts had a more mature phenotype than control DCs and DCs exposed to syngeneic apoptotic fibroblasts; necrotic T cells were not examined.

DC maturation, similar to that induced by LPS and tumour necrosis factor (TNF)-α, has been observed in an immature DC cell line (D1) co-cultured with allogeneic apoptotic cells [11]. It has been suggested that in such co-cultures the presence of mycoplasma is responsible for induction of DC maturation [20]. However, Sauter et al. [9] used both primary cells and secondary cell lines that were not mycoplasma-contaminated and demonstrated differences in DC phenotype following incubation with allogeneic necrotic cells. As in this study, primary cells had no effect. Supernatants from the necrotic cells induced the same effects as did the cells themselves, but the nature of the maturation factor(s) was not established, although TNF-α and IL-1 were excluded. Xenogeneic apoptotic cells have also been used [1], and DCs that had phagocytosed apoptotic cells were unable to up-regulate CD86 surface expression compared to DCs that had not ingested the apoptotic cells. The results of experiments performed using in vitro culture models where the DC and dying cells are allogeneic or xenogeneic may be confounded by the activity of natural killer cells, and by other similar receptor mechanisms.

The current study demonstrates clearly that DCs induction of secondary immune responses is not impaired following preincubation with autologous apoptotic or necrotic T cells. These findings are in agreement with several studies where apoptotic T cells have been used as a source of antigen for DC antigen presentation. Zhao et al. [21] showed that immature DCs exposed to autologous T cells infected in vitro with laboratory strains of HIV-1 (which was later inactivated), and matured with CD154, induced gamma interferon synthesis by both autologous CD4⁺ and CD8⁺-specific T cells. In mice a single injection of MBMDCs pulsed with a syngeneic apoptotic Rauscher virus-induced T cell lymphoma cell line (RMA) expressing a leaderless non-secreted ovalbumin (OVA) protein induced an OVA-specific CTL response in vivo[13]. Spetz et al. [22] demonstrated that mice immunized with irradiated apoptotic HIV-1/murine leukaemia virus (MuLV) infected syngeneic splenocytes developed strong nef-specific CTL responses in addition to CD4 and CD8 T cell responses to p24 antigen. The mice were protected against infection following challenge with live HIV-1/MuLV infected cells. Thus apoptotic cells were potentially immunogenic, and co-incubation with DCs did not impair DC function.

In contrast, other reports have suggested impaired DC function following incubation with either autologous or syngeneic apoptotic cells, but not necrotic cells. Urban et al. [19] documented that exposed DCs were unable to stimulate a MLR, with or without addition of LPS. The response observed was below that seen with immature DCs. Also, DCs exposed to apoptotic cells and pulsed with antigen failed to induce T cell proliferation of an antigen-specific T cell clone. Gallucci et al. [8] showed that immature male murine DCs, co-cultured for 40 h with syngeneic male apoptotic fibroblasts, did not induce proliferation of the IA^b-restricted, CD4⁺ T cell clone specific for the male Y antigen. Male DCs co-cultured with necrotic female fibroblasts induced less male antigen-specific T cell proliferation than DCs incubated with necrotic male fibroblasts. Interestingly, female DCs co-cultured with apoptotic male fibroblasts were able to induce only minimal T cell proliferation while those incubated with necrotic male fibroblasts stimulated strong T cell responses. The discrepancy between our results and these two studies may reflect the different methodologies involved. For example, neither of the previous studies reported removal of apoptotic cells from the functional assays. The presence of large numbers of apoptotic cells which, as shown in Fig. 1, form clusters around DC, may interfere directly with interactions between DC and responder T cells, rather than signifying an altered state in DC functional activity. A similar explanation may underlie the profound block of DC function (of both primary and secondary responses) seen after exposure to apoptotic cells, observed by Sauter et al. [9]. The use of allogeneic or even xenogeneic cell lines in many studies is again a confounding factor in interpretation.

It is perhaps noteworthy that the present study of the association between DC and dying cells used apoptotic or necrotic cells, which resembled those that the DC might encounter in vivo as closely as possible, particularly in that the T cells were autologous and not transformed. In this model, exposure of DC to apoptotic or necrotic T cells neither induced maturation nor functional impairment. Significantly, DC retained their potent antigen-presenting ability, inducing secondary immune responses to recall antigens. There was no evidence that DC–T cell functional assays generated either anergic or hyporesponsive cells to these antigens. This robust DC function in the presence of large numbers of dead and dying lymphocytes may represent a necessary specialization to life in an immune system where high cell turnover is a necessary and commonplace aspect of the normal immune response.