SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

T-cell sensitization to indirectly presented alloantigens (indirect pathway of allorecognition) plays a critical role in chronic rejection. The usual very efficient priming of such self-restricted, T helper type 1 (Th1)-deviated CD4+ T cells obviously conflicts with the fact that allogeneic MHC molecules are poorly immunogenic per se. The aim of the present study is to elucidate whether direct allosensitization induces production of inflammatory mediators that may affect recruitment and activation of immature bystander (host) dendritic cells (DC). These potential mechanisms were studied in vitro by conducting primary allogeneic mixed leucocyte reactions (MLR), mimicking the priming phase in secondary lymphoid organs, and secondary MLR, mimicking the effector phase within the graft. Primary, and particularly secondary, MLR supernatants were found to contain high levels of monocyte/immature DC-recruiting CC chemokines and pro-inflammatory cytokines. Exposure of immature DC to primary or secondary MLR supernatants was found to upregulate CD40 expression and further enhanced lipopolysaccharide-induced interleukin-12 (IL-12) p70 production. Secondary MLR supernatants additionally induced upregulation of CD86 and deviated allogeneic T-cell responses towards Th1 (enhanced interferon-γ production without concomitant induction of detectable IL-4 or IL-10 production). These findings indicate that direct allorecognition may act as a Th1-deviating adjuvant for indirect allosensitization.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

There are two distinct, although not necessarily mutually exclusive, pathways in allorecognition. First, naïve as well as memory CD4+ and CD8+ T cells from the recipient can become activated upon direct recognition of intact MHC/peptide complexes present on graft-derived antigen-presenting cells (APC) [1]. This vigorous response, which appears to violate the rule of self-MHC restriction, is driven primarily by antigenic mimicry [2]. Naïve recipient T cells can also become activated by recognizing processed alloantigens in the form of peptides presented by the recipient's APC which is analogous to that of physiological antigen recognition [3–5].

It is now generally assumed that T cells activated via the direct allorecognition pathway are of main importance for initiation of early acute rejection. Over time, the importance of directly activated T cells seems to diminish which is due to the successive depletion of passenger leucocytes [6]. Furthermore, activated T cells that directly recognize donor MHC antigens on parenchymal cells, lacking proper co-stimulatory elements, may subsequently become anergized [7]. Several lines of evidence have indicated that the indirect pathway of allorecognition also contributes to the pathogenesis of allograft rejection [3, 5], and convincing experimental evidence linking indirect allorecognition and chronic rejection was recently presented [8]. Indirectly primed CD4+ cells usually exhibit a Th1-deviated cytokine profile and have been shown to initiate effector mechanisms of chronic rejection including delayed type of hypersensitivity responses [9, 10] and cell-mediated cytotoxicity [11] and provide B-cell help for antibody production [12]. The prevailing view as to direct allosensitization is that passenger dendritic cells (DC), activated by the act of transplantation itself (surgical trauma, ischaemia and reperfusion), migrate to secondary lymphoid organs (SLO) where they induce the initial priming of T cells with direct alloreactivity [2]. Transplantation of vascularized solid organs induces a blood-borne migration of donor DC to the host spleen [13–15] and hepatic lymph nodes [15], while skin transplantation mainly induces a local migration of graft DC to host-draining lymph nodes [16]. Small numbers of donor DC are usually found in the T-cell areas of SLO within 1–2 days after transplantation, and immunohistological studies have documented that these migrated donor DC form clusters with proliferating host T cells, indicating primary sites of direct allosensitization [15]. Once activated, these alloreactive CD4+ and CD8+ T cells are thought to recirculate to the graft and cause the acute rejection process after being restimulated by allogeneic MHC class II and/or class I molecules expressed on graft cells.

As to indirect allosensitization, there are two main sites where host DC potentially may capture donor-derived MHC molecules after transplantation of vascularized allogeneic organs. First, donor DC may function as vehicles for transporting donor antigens directly to resident host DC within SLO [17]. This mechanism was convincingly documented in experiments using the Y-Ae monoclonal antibody to monitor the processing and presentation of H2-Eαin vivo. When DC from fully allogeneic H2-Eα donors were injected subcutaneously into H2-Ab recipients, short-lived migrating DC were shown to be processed by most of the recipient DC in the lymph node [18]. Second, host-derived DC or DC precursors, including circulating monocytes, may be recruited into the graft [19, 20]. At face value, the usual very efficient priming of the indirect allorecognition pathway that occurs after transplantation conflicts with the fact that MHC-derived molecules are poorly or nonimmunogenic per se within a species [21]. Immunization with allogeneic peptides thus requires co-administration of strong adjuvants, such as Freud's complete adjuvant [8, 21, 22]. The ability of viable passenger DC from the graft to directly deliver allogeneic molecules to resident host DC within SLO has been suggested to explain this immunogenic discrepancy between ‘dead’ and viable allogeneic sources in indirect alloimmunization [17]. However, as resident DC within SLO are immature and probably tolerizing at steady state [23–25], experiments have yet to elucidate whether indirect presentation of alloantigens by resident DC in SLO really promotes indirect alloimmunization [2].

The aim of the present study is therefore to elucidate whether the different phases of direct allorecognition that occurs at anatomical sites where host DC most likely capture donor-derived MHC molecules induce production of inflammatory mediators that affects recruitment and activation state of immature DC in bystander position (host DC) and their subsequent Th1/Th2 polarization profile. These potential mechanisms were studied in vitro by conducting primary mixed leucocyte reactions (MLR), mimicking the priming phase in SLO, and secondary MLR, mimicking the effector phase within the graft, and the impact of these supernatants on phenotype and function of bystander immature DC was subsequently investigated.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Culture medium and reagents.  RPMI-1640, l-glutamine, penicillin, streptomycin, recombinant human granulocyte/macrophage colony-stimulating factor (GM-CSF), interleukin-4 (IL-4), tumour necrosis factor-α (TNF-α), IL-1β, IL-6, prostaglandin E2 (PGE2) and lipopolysaccharide (LPS) from Escherichia coli were all purchased from Sigma-Aldrich (Stenheim, Germany). Serum-free X-VIVO 15 medium was purchased from Bio Whittaker (Verviers, Belgium).

Primary and secondary MLR.  Peripheral blood mononuclear cells (PBMC) from healthy blood donors were isolated after centrifugation over a density gradient (Lymphoprep; Nycomed, Oslo, Norway) and used immediately or cryopreserved for restimulation assays. The study was approved by the Human Research Ethics Committee at the Sahlgrenska Academy, Göteborg University, Göteborg, Sweden. A standard one-way MLR protocol with gamma-irradiated (30 Gy) PBMC as stimulator cells and PBMC from random allogeneic donors as responder cells was used. This protocol is a well-accepted in vitro model for measurement of the direct allorecognition activity, as the frequency of cells directly recognizing alloantigens in such MLR cultures is 100-fold higher than that for indirect recognition [4]. Irradiated stimulator PBMC (1 × 105 cells/well) and responder PBMC (1 × 105 cells/well) were co-cultured in U-bottom 96-well plates in a total volume of 200 µl of serum-free X-VIVO 15 per well. Bovine or human serum was not added in order to exclude any influence of xenogeneic proteins or allogeneic proteins from third-party donors. Supernatants were collected and pooled at day 6 and subsequently kept at −20 °C until analysed. Responding cells from the primary MLR were harvested at day 6, washed twice and plated in X-VIVO 15 (1 × 105 cells/well in 96-well U-bottom plates) with irradiated PBMC (1 × 105 cells/well) taken from the same allogeneic donor as used for primary stimulation. Supernatants were collected and pooled after 48 h and subsequently kept at −20 °C until analysed.

Generation of early immature DC.  Isolated PBMC in complete medium consisting of RPMI-1640 supplemented with 2 mm of l-glutamine, 100 U/ml of penicillin and 100 µg/ml of streptomycin and 2% human heat-inactivated AB serum were plated in 24-well plastic culture plates at 2.5 × 106 cells per well and allowed to adhere for 2 h. Nonadherent cells were removed by two washes. The remaining adherent cells, mainly consisting of CD14+ monocytes (data not shown), were cultured in complete medium with addition of GM-CSF (1000 U/ml) and IL-4 (1000 U/ml) for 24 h in order to obtain early immature DC as recently described [26]. These cells were still firmly adherent and exhibited low CD14 expression and high MHC class II expression consistent with early immature DC development (data not shown).

Maturation of DC.  After culture of adherent monocytes for 24 h in RPMI-1640 supplemented with GM-CSF and IL-4, half the volume (0.5 ml) of culture media in each well was replaced by primary or secondary MLR supernatants or X-VIVO 15 with or without addition of a maturation cocktail consisting of 5 ng of TNF-α, 5 ng of IL-1β, 150 ng of IL-6 and 1 µg of PGE2 [27, 28]. The cells were cultured for another 48 h when phenotypic characterization was performed.

LPS-induced IL-12 production.  After culture of adherent monocytes for 24 h in RPMI-1640 supplemented with GM-CSF and IL-4, half the volume (0.5 ml) of culture media in each well was replaced by primary or secondary MLR supernatants or X-VIVO 15 with or without addition of a maturation cocktail followed by addition of 1 µg of LPS to all wells. The cells were cultured for another 24 h when the supernatants were collected and stored at −20 °C until analysed.

MLR using DC treated with MLR supernatants as stimulator cells.  After culture of adherent monocytes for 24 h in RPMI-1640 supplemented with GM-CSF and IL-4, half of the volume (0.5 ml) of culture media from each well was replaced by X-VIVO 15 or MLR supernatants. The cells were cultured for another 24 h and subsequently washed. Approximately 2.5 × 105 DC and 1 × 106 responder PBMC were then co-cultured in 24-well plates in a total volume of l ml of serum-free X-VIVO 15 per well. Supernatants were collected and pooled at day 6 and subsequently kept at −20 °C until analysed.

Phenotyping of DC.  Allophycocyanin-conjugated anti-CD40, anti-CD86, anti-CD83, anti-chemokine receptor-7 (CCR7) antibodies, phycoerythrin (PE)-conjugated CD40 and CD86 antibodies, fluorescein isothiocyanate (FITC)-conjugated anti-CD86 and isotype-matched controls were purchased from BD Pharmingen (San Diego, CA, USA). FITC-conjugated anti-CCR7 was purchased from R&D Systems (Abingdon, UK). Fluorochrome-conjugated antibodies were added to DC cultured for 2 days in different maturation conditions, incubated at 4 °C for 20 min and washed. Flow cytometry was performed on a FACScan (BD Biosciences, Erembodegem, Belgium). Ten thousand gated events were acquired for each maturation condition and antibody. Data were analysed using cell quest software (BD Biosciences).

Enzyme-linked immunosorbent assay (ELISA) and multiplex bead-array assay for cytokine determination.  The human cytokines IL-6, IL-10 and IL-12, CC chemokines macrophage inflammatory protein 1α (MIP-1α) and MIP-1β, regulated upon activation of T cell expressed and secreted (RANTES), and CXC chemokines IL-8 and growth-related oncogene α (GRO-α) were determined by a multiplex bead-array assay using the Luminex 100 system (Luminex Corporation, Austin, TX, USA). Multiplex bead kits were purchased from LINCO Research, Inc. (St Charels, MO, USA). Interferon (IFN)-γ, IL-4, TNF-α and IL-1β were determined by ELISA from R&D Systems according to the manufacturer's instructions.

Statistical analysis.  The statistical significance of differences between experimental samples was determined using Student's t-test.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The direct pathway of allorecognition induces substantial production of monocyte/immature DC-recruiting CC chemokines

The primary one-way MLR using irradiated PBMC as stimulator cells and allogeneic PBMC as responder cells was chosen as a model with potential relevance to the in vivo situation when passenger DC reach the T-cell areas within secondary lymphoid organs. The bulk PBMC stimulator population thus contains a low but substantial number of DC with a high allostimulatory capacity in vitro[29]. These DC have further been shown to form discrete aggregates or clusters with alloantigen-specific T cells early in the MLR which is similar to findings in situ[15]. By using bulk PBMC in the responder cell population, most of the cell types we found in SLO are represented, including naïve and central memory T cells [30], NK cells [31] and DC [32, 33]. Even if the priming phase of the direct pathway more strictly concerns the bi-directional interaction between alloreactive T cells and passenger DC, other cell types in the responder population such as alloreactive NK cells may potentially affect the final outcome of this direct allorecognition pathway [34]. A second challenge (secondary MLR) of culture cells from primary MLR with irradiated PBMC from the primary allogeneic stimulator was used to mimic the effector phase in the graft when primed T cells become reactivated by directly presented MHC antigens on graft cells. In the first set of experiments, different chemokines, including CC chemokines with well-known impact on monocyte/immature DC recruitment, were analysed. As shown in Table 1, a significant increase above background levels (autologous MLR) was seen for the CC chemokines MIP-1α, MIP-1β and RANTES in primary MLR supernatants. CC chemokine production in secondary MLR supernatants was further increased and reached five to 10-fold higher levels than those seen in primary MLR in each individual experiment. Such enhancement, from primary to secondary MLR, was not seen for the granulocyte-recruiting CXC chemokines IL-8 and GRO-α.

Table 1.  Chemokine concentration (pg/ml) in MLR supernatants
 MIP-1αMIP-1βRANTESIL-8GRO-α
  1. GRO-α, growth-related oncogene α; IL, interleukin; MIP-1α, macrophage inflammatory protein 1α; MLR, mixed leucocyte reactions; PBMC, peripheral blood mononuclear cells.

  2. Irradiated stimulator PBMC (1 × 105 cells/well) and responder PBMC (1 × 105 cells/well) were co-cultured in 96-well U-bottom plates in a total volume of 200 μl of serum-free X-VIVO 15 per well. Supernatants were collected and pooled at day 6 and subsequently kept at −20 °C until analysed. Responding cells from the primary (allogeneic) MLR were harvested at day 6, washed twice and plated in X-VIVO 15 (1 × 105 cells/well in 96-well U-bottom plates) with irradiated PBMC (1 × 105 cells/well) taken from the same allogeneic donor as used for primary stimulation. Supernatants were collected and pooled after 48 h and subsequently kept at −20 °C until analysed.

Auto-MLR
 Experiment A<8130587.150147
 Experiment B<83340018.425802
Primary MLR
 Experiment 12841.57632134.1862.926
 Experiment 25201.25822745.2435.946
 Experiment 31451.34341926.7732.869
 Experiment 44672.11858828.1653.300
Secondary MLR
 Experiment 18.84334.1133.04426.1501.538
 Experiment 26.88236.8281.09041.7841.698
 Experiment 312.79933.4644.06467.9266.447
 Experiment 411.89038.2443.87652.4486.320

The direct pathway of allorecognition induces substantial production of pro-inflammatory cytokines

As shown in Table 2, a significant increase above background levels (autologous MLR) was seen in primary MLR supernatants for the pro-inflammatory cytokines IL-6, IFN-γ and TNF-α. The secondary MLR was characterized by a rapid (within 48 h) and strong production of IL-1β, IFN-γ and TNF-α at levels that usually were five to 10-fold higher than those seen in primary cultures. No detectable (<16 pg/ml) IL-4, IL-10 or IL-12 was found in primary or secondary MLR cultures.

Table 2.  Cytokine concentration (pg/ml) in MLR supernatants
 IL-1βIL-6TNF-αIFN-γ
  1. IFN, interferon; IL, interleukin; MLR, mixed leucocyte reactions; PBMC, peripheral blood mononuclear cells; TNF, tumour necrosis factor.

  2. Irradiated stimulator PBMC (1 × 105 cells/well) and responder PBMC (1 × 105 cells/well) were co-cultured in 96-well U-bottom plates in a total volume of 200 μl of serum-free X-VIVO 15 per well. Supernatants were collected and pooled at day 6 and subsequently kept at −20 °C until analysed. Responding cells from the primary (allogeneic) MLR were harvested at day 6, washed twice and plated in X-VIVO 15 (1 × 105 cells/well in 96-well U-bottom plates) with irradiated PBMC (1 × 105 cells/well) taken from the same allogeneic donor as used for primary stimulation. Supernatants were collected and pooled after 48 h and subsequently kept at −20 °C until analysed.

Auto-MLR
 Experiment A<4<5<16<16
 Experiment B<4<5<16<16
Primary MLR
 Experiment 1<42.831231476
 Experiment 2<415.241282261
 Experiment 3<4929104417
 Experiment 4<41.606215312
Secondary MLR
 Experiment 1643.7573.8515.621
 Experiment 2514.3902.2521.700
 Experiment 312013.3902.7239.968
 Experiment 415211.7321.9158.670

Culture supernatants from primary and secondary MLR promote a significant phenotypic maturation of bystander immature DC

In order to study the impact of MLR supernatants on bystander DC as to phenotype and function, a recently described model for rapid propagation of monocyte-derived DC [26] was established. As shown in Fig. 1, ‘fast’ mature DC could be propagated by cultivating adherent monocytes with GM-CSF and IL-4 for 24 h followed by addition of a standard DC maturation cocktail consisting of IL-1β, TNF-α, IL-6 and PGE2 [27, 28] for another 48 h. Adherent monocytes that were cultivated for 1–3 days in GM-CSF and IL-4 without addition of maturating factors became negative for CD14 (data not shown) and showed no neoexpression of CD83 and CCR7 (Fig. 1), thus consistent with immature DC. Addition of supernatants from primary as well as secondary MLR after 24 h was found to induce a significant upregulation of CD40 expression in DC (Fig. 2). Primary MLR supernatants further induced a significant neoexpression of CCR7, while secondary MLR supernatants induced a significant upregulation of CD86. Some supernatants were found to enhance CD83 expression, but this was an inconsistent finding excluding any statistically significant increase (Fig. 2). In all experiments, CD86 and CD40 expression was higher using supernatants from secondary MLR when compared with primary MLR supernatants from the same recipient/donor combination (data not shown).

image

Figure 1. Maturation markers on early mature dendritic cells (DC). The histograms show expression of maturation-associated cell-surface antigens on human DC. Early mature DC were propagated by cultivating adherent monocytes in GM-CSF and IL-4-containing media for 24 h and subsequently exposed to an inflammatory cocktail consisting of IL-1β, IL-6, TNF-α and PGE2 for another 48 h (upper row). Thin lines represent staining with isotype-matched irrelevant monoclonal antibodies. GM-CSF, granulocyte/macrophage colony-stimulating factor; IL, interleukin; PGE2, prostaglandin E2; TNF, tumour necrosis factor.

Download figure to PowerPoint

image

Figure 2. Exposure of immature dendritic cells (DC) to MLR supernatants induces phenotypic maturation. Early immature DC were propagated by cultivating adherent monocytes in GM-CSF and IL-4-containing media for 24 h. These immature DC were subsequently exposed to primary or secondary MLR supernatants (MLR-SN) by replacing half the volume of culture medium with MLR-SN or X-VIVO 15 (control medium) for another 48 h. The mean fold increase (MFI) of mean fluorescence intensity as compared to control using MLR-SN from four to six different donor/responder MLR combinations is shown. Error bars are shown and represent one SD from the mean. *P < 0.05; **P < 0.01. GM-CSF, granulocyte/macrophage colony-stimulating factor; IL, interleukin; MLR, mixed leucocyte reactions.

Download figure to PowerPoint

Culture supernatants from primary and secondary MLR prime immature DC for LPS-induced IL-12 p70 production

T cells with indirect alloreactivity that develops after transplantation are usually skewed towards Th1 with high IFN-γ production compared to IL-4 production [9, 10]. Maturation of DC is a prerequisite to induce activation of T cells, but their polarization of T cells towards Th1 is closely linked to their capacity to produce bioactive IL-12 (IL-12 p70). To investigate the impact of MLR supernatants on IL-12 production, immature DC were stimulated with LPS with or without addition of MLR supernatants for 24 h. Mean values for IL-12 production for control DC and DC stimulated with primary or secondary MLR supernatants were 232 pg/ml (range 50–600), 372 pg/ml (range 60–650) and 616 pg/ml (range 68–960), respectively. As illustrated in Fig. 3, DC stimulated with primary and secondary MLR supernatants significantly enhanced LPS-induced bioactive IL-12 production as compared with control DC.

image

Figure 3. MLR supernatants prime dendritic cells (DC) for increased LPS-induced IL-12 production. Early immature DC were propagated by cultivating adherent monocytes in GM-CSF and IL-4-containing media for 24 h. These immature DC were subsequently exposed to LPS (1 µg/ml) and primary or secondary MLR supernatants (MLR-SN) or medium by replacing half the volume of culture medium with MLR-SN or X-VIVO 15 (control) for another 48 h. * P < 0.05; **P < 0.01. GM-CSF, granulocyte/macrophage colony-stimulating factor; IL, interleukin; LPS, lipopolysaccharide; MLR, mixed leucocyte reactions.

Download figure to PowerPoint

Pretreatment of immature DC with secondary MLR supernatants induces Th1-polarized allogeneic T-cell responses

In order to test whether MLR supernatants have any impact on functional Th1/Th2-polarizing activity by DC, primary allostimulatory experiments were performed. Mean values for IFN-γ production in MLR cultures, using control DC or DC pretreated with primary or secondary MLR supernatants as stimulator cells, were 30 pg/ml (range 8–51), 32 pg/ml (range 7–65) and 103 pg/ml (range 24–179), respectively. As illustrated in Fig. 4, DC pretreated with secondary MLR supernatants significantly enhanced subsequent IFN-γ production by allogeneic T cells as compared with control DC. IL-4 and IL-10 production remained undetectable in all experiments (<16 pg/ml).

image

Figure 4. Pretreatment of immature dendritic cells (DC) with secondary MLR supernatants induces Th1-polarized allogeneic T-cell responses. Early immature DC were propagated by cultivating adherent monocytes in GM-CSF and IL-4-containing media for 24 h. These immature DC were subsequently exposed to MLR supernatants (MLR-SN) or medium by replacing half the volume of culture medium with MLR-SN or X-VIVO 15 (control) for another 48 h. After washing, 2.5 × 105 DC and 1 × 106 responder PBMC were co-cultured in 24-well plates in a total volume of l ml of serum-free X-VIVO 15 per well. Supernatants were collected and pooled at day 6 and subsequently kept at −20 °C until analysed. * P < 0.05. GM-CSF, granulocyte/macrophage colony-stimulating factor; IL, interleukin; MLR, mixed leucocyte reactions.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The present study was undertaken in order to elucidate whether the inflammatory reaction that is elicited by the direct pathway of allorecognition, and that occurs at anatomical sites where host DC most likely capture donor-derived MHC molecules, may affect recruitment and/or activation of bystander DC, including their subsequent Th1/Th2 polarization profile. Such potential mechanisms were studied in vitro by conducting primary MLR, reflecting the priming phase in SLO, and secondary MLR, reflecting the effector phase within the graft. Interestingly, a strong production of well-known monocyte/immature DC-recruiting chemokines such as MIP-1α and RANTES [35] was seen. This novel finding indicates that recipient monocytes/immature DC may be actively recruited to the anatomical sites where the priming and particularly the effector phase of direct allorecognition takes place (SLO and graft, respectively) and is thus in accordance with in situ findings [15, 18, 20]. The pro-inflammatory cytokines TNF-α, IL-6 and IFN-γ were found to be produced most abundantly in primary and secondary MLR which confirm data from earlier reports [36]. TNF-α and IL-6 are both well-known DC-maturating factors [37] that are found in supernatants from monocyte cultures activated by a IgG-coated plastic surface (monocyte-conditioned medium) and frequently included in recombinant cytokine cocktails used for phenotypical and functional maturation of monocyte-derived DC in vitro[27, 28]. IFN-γ is a potent inducer of maturation in myeloid DC [38] and also implicated as a co-factor for LPS-induced maturation of monocyte-derived DC [39]. A recent study has also implicated IFN-γ in DC maturation that is induced by supernatants taken from anti-CD3-activated T-cell cultures [40]. Our findings as to cytokine production in primary and secondary MLR are further in line with in vivo data, demonstrating that the priming process in SLO and the acute rejection process within the graft are characterized by a prominent expression of Th1 cytokines including IFN-γ and TNF-α[41–45].

In order to study the functional relevance of the inflammatory mediators that were produced during the different phases of direct allorecognition, we next investigated the impact of MLR supernatants on phenotypic DC activation. Addition of primary MLR supernatants to immature DC was found to induce a modest but significant increase of CD40 on early immature DC, while supernatants taken from secondary MLR induced a highly elevated CD40 expression in all experiments performed. In some experiments, the strong CD40 expression induced by secondary MLR was even higher than that induced by a well-established CD40-inducing inflammatory cocktail (data not shown). Supernatants from primary MLR further induced a modest but significant increase in CCR7 expression, while secondary MLR induced a significant increase in CD86 expression. These data are only partially in line with a recent publication using a similar primary MLR approach [46]. In this particular study, supernatants from primary MLR were found to induce a significant neoexpression of both CD83 and CCR7 and a prominent increase in CD86, while a lower increase in CD40 expression was noted. The partial discrepancy between our results and those presented by Laurin et al. may be explained by the fact that we used early immature DC, while Laurin et al. used immature DC cultured for 6 days in GM-CSF/IL-4-containing media. Moreover, in the study performed by Laurin et al., 10% of human serum was added to the MLR supernatants that were produced in culture media already containing 10% human serum. Such high serum concentration (approximately 20%) could potentially affect DC maturation by substantial adherence of serum IgG to culture plates with subsequent Fc receptor-mediated activation of DC [47]. The reason for the rather selective increase in CD40 expression in DC that was induced by our MLR supernatants is unclear, but the concomitant presence of inflammatory CC chemokines may possibly affect such selective impact on DC phenotype. Indeed, in a recent study, the inflammatory chemokines MIP-1α, MIP-1β and RANTES were shown to act in concert with relative low amounts of IFN-γ (100 pg/ml) to induce a selective CD40 upregulation (without any concomitant upregulation of CD80 and CD86) on bone marrow-derived macrophages infected by Listeria monocytogenes[47].

As the responder PBMC population used in our primary MLR contains blood-derived DC that may be affected by inflammatory mediators produced during direct allorecognition (and these bystander DC likely capture and process released alloantigens from stimulator PBMC), it could be predicted that some T cells with indirect alloreactivity would be immunized during the course of a one-way MLR. Indeed, in a study performed by Liu and colleagues [4], the frequency of T cells with indirect alloreactivity, as assessed by limiting dilution analysis, was found to increase from 1:250,000 in unprimed T-cell populations to 1:40,000 after primary stimulation in a one-way MLR using irradiated PBMC as stimulators and allogeneic PBMC as responders. The potential immunological mechanism that could explain such in vitro immunization was never addressed by the authors but may obviously be explained by our present observations.

Our data further represent the first demonstration that the direct pathway of allorecognition in vitro induces the release of soluble inflammatory mediators that prime bystander DC for enhanced LPS-induced IL-12 production and Th1 polarization of allogeneic T-cell responses. These novel findings in vitro are in accordance with in vivo data where cross-primed alloreactive T cells usually exhibit a Th1-deviated cytokine profile [9, 10]. The most likely candidate to explain these effects by MLR supernatants is IFN-γ which is known to prime DC for enhanced IL-12 production by stimulation with LPS [48] or TNF-α[48, 49]. Moreover, IFN-γ has been shown to be an obligate co-factor for CD40L-induced IL-12 p70 production [50]. Inflammatory chemokines may also affect IL-12 production, as shown in macrophages, either by their own action or by acting in a synergistic fashion with IFN-γ[51].

To summarize, the present study provides evidence that the inflammatory response that is induced during the different phases of direct allorecognition in vitro and that occurs at anatomical sites where immature host DC capture donor-derived MHC products in vivo, promotes activation of immature bystander (host) DC and prime these DC for subsequent Th1 polarization. Such mechanism could thus finally explain why the presence of passenger DC not only promotes direct allosensitization and subsequent acute rejection but also leads to a local release of inflammatory mediators that may promote indirect alloimmunization.

Finally, our data would further predict that repeated injections of allogeneic APC that are loaded with nonself antigens other than MHC antigens not only would induce alloimmunization but also would promote a self-MHC-restricted Th1 immunization against any other nonself antigen that is present within the injected allogeneic APC. Interestingly, a recent study convincingly demonstrated that repeated injections of ovalbumin (OVA)-loaded allogeneic APC (allogeneic splenocytes or adherent peritoneal exudates cells), but not autologous APC, induced a strong and long-lasting humoral and cellular Th1 response against OVA [52].

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

We thank Ulla Karlsson for technical assistance (multiplex bead-array assay and FACS). This study was supported by grants from TUA-SAM and LUA-SAM, Göteborg University, Wilhelm and Martina Lundgrens Scientific Fund, Swedish Dental Society and the Royal Society of Arts and Sciences in Göteborg.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References