Introduction

The most potent initiator of the immune response in vivo and in vitro is a heterogeneous group of professional APC termed 'dendritic cells' (DC).^{1, 2, 3} Dendritic cells have been shown to carry infectious organisms^{4, 5, 6} or may be associated with defence against tumours.^{7, 8} Therefore, DC are believed to be key cells in the maintenance of the integrity of the body against both the attack from outside by infectious agents and from within by tumour cells.

The life history of DC has been separated into two main functional stages. After leaving the bone marrow, the first function of the newly emigrated myeloid-derived DC is to pick up Ag.^{9, 10, 11} After appropriate signals, DC migrate from the periphery of the body to the T cell area in lymphoid organs.¹² During migration, DC acquire full Ag-presenting capacities by altering biosynthesis and relocalization of MHC class II molecules.^{9, 10, 11} Therefore, DC appear to have important survey functions throughout the body, as well as a central role in the initiation of a T cell immune response.

Interferon appears to regulate haematopoiesis, cell growth and differentiation,^{13, 14} control the host defence against infections,¹⁵ and inhibit tumour development by appropriately interfering with individual cells, or support the immune response.¹⁶ Two types of IFN families have been described. The type-I IFN family includes IFN-, - and -, the type-II IFN is represented by IFN- as the only member. Interferon molecules bind to species-specific receptors present on the cell surface. It is thought that the binding of IFN to its specific receptor induces the plethora of functions associated with IFN.

Mice (AG129Sv/Ev) lacking both type-I and type-II IFN receptors (IFN-r) have been generated and the role of IFN in anti-viral immune response was analysed.^{17, 18, 19} The animals were unable to cope with very low doses of virus infections such as influenza, vesicular stomatitis virus and lymphoicytic choriomeningitis virus (LCMV). It has been shown that IFN-/ has the capacity of directly depressing haematopoietic cells which allow LCMV replication during the early phase of virus infection.¹⁵ Despite the absence of IFN-r, initiation of the CD8-associated T cell and humoral immune response against virus was possible in the AG129Sv/Ev mice. Even though a cellular immune response was present initially, this response was found to be exhausted in a later phase of the infection due to the overwhelming virus replication. Although the role of the IFN system to control initial virus replication was clearly shown, the functional capacity of DC with defective IFN-r was blurred by the uncontrolled virus replication.

Therefore, the aim of the present study was to evaluate both the physiology and immunological capabilities of DC lines lacking functional IFN-r I and II and to compare the cells to cutaneous wild-type (wt) DC.²⁰

We show evidence that the immortalized DC without functional IFN-r I and II differ in their growth rates when compared to immortalized wtDC, are competent in priming naive T cells in vitro, possibly aided by IL-6, which is secreted constitutively, and IL-12, which is strongly up- regulated by a cognate CD–T cell interaction. Furthermore, virus-infected DC are able to induce antibodies in vivo against viral Ag present in limited amounts. Therefore, the DC lines with non-functional IFN-r appear to unify some functions of immature and mature DC.

Materials and methods

Mice

Inbred 129Sv/Ev (H-2^b) or congenic 129Sv/Ev mice with deleted IFN-r I and II (IFN-r o/o) termed AG129Sv/Ev, or G129Sv/Ev (IFN-r II o/o),^{17, 18, 19} were kindly provided by Rolf Zinkernagel, University of Zürich, Switzerland. The presence or absence of the IFN-r in mice or selected cell lines was confirmed by PCR.¹⁷

Media and antigens

Iscove's modified Dulbecco's culture medium (Gibco BRL, Life Technologies, Basel, Switzerland) was used supplemented with 10% FCS and a mixture of antibiotic-antimycotic (Gibco BRL) containing penicillin 100 U/mL, streptomycin 100 g/mL and amphotericin B 0.25 g/mL (complete medium). The Ag ovalbumin (OVA) and hen egg lysozyme (HEL) were from Sigma Chemical Co. (St Louis, MO, USA). The OVA was analysed by PAGE under reducing conditions and shown to contain a molecule of the expected size without contamination of degraded products (data not shown).

Cell lines

The fibroblast cell line, termed 'VN 11',^{9, 21} was obtained from Paula Ricciardi-Castagnoli, CNR Center of Cytopharmacology, University of Milan, Italy. The T cell lines Hd-1AC5 (specific for HEL and I-E^d restricted), E3 (OVA-specific and I-E^k-restricted), E8 (OVA-specific and I-E^k-restricted) were obtained from Stephen I. Katz, Department of Dermatology, NIH, Bethesda, MD, USA. The cell line BO97.10.5 (OVA-specific and I-A^b-restricted), was a gift from Philippa Marrack, Jewish Medical Center, Denver, CO, USA.

Antibodies and FACS staining

Cell-surface-expressed molecules were determined with mAb and analysed by FACS using standard procedures.²² For this, cultured adherent cells were detached by a gentle trypsin-EDTA (Gibco) treatment, the cells counted and the viability determined by trypan blue exclusion. Only cells with more than 95% viability were used. In order to block the Fc receptors, cell suspensions (10⁶ cells/mL) were incubated on ice with mAb 2.4G2.²² After one wash in FACS buffer (PBS containing 5% FCS and 0.01% sodium azide), cells were resuspended in an equal volume of FACS buffer, and aliquots of 100 L of the cell suspensions (10⁶ cells/mL) were seeded in individual wells of a 96-well plate. Appropriate concentrations of different mAb were added to the seeded cells and the mixture was incubated on ice for 30 min. Rat mAb to CD11b, MHC, CD4, B-220, CD45, CD8, appropriate Ig isotype controls and streptavidin-labelled PE were from PharMingen (San Diego, CA, USA) and used according to the manufacturer's recommendation. The FACS analysis was performed at the FACS core facility (ETH, Zürich, Switzerland) with FACS equipment (Coulter Electronics, Hialiah, FL, USA). At least 5000 cells were analysed. Analysis of the data was performed with the PC lysis software (Becton Dickinson, San Jose, CA, USA).

Immortalization of cells

Newborn and adult mice were killed under anaesthesia, submerged in cold ethanol for 5 min and air dried for ~ 10 min. Tissues were removed and processed by gentle teasing until cell suspensions were obtained. The cells were counted and the viability was determined. Skin of newborn mice was removed carefully from the body, cut into small pieces and incubated with 1% PBS-buffered trypsin-EDTA for 30 min at 37°C. Epidermal cell suspensions were then obtained by carefully pipetting the cells through an 18-G needle several times. Cell suspensions seeded in 25-mL tissue culture flasks (Corning Glass Works, Corning, NY, USA) were infected with a 1:1 volume of 0.2 mol/L filtered cell-free supernatants from the cell line containing the MIBY2-N11 recombinant defective retroviral vector that carries the v-myc MH2 gene.^{21, 22} After incubation for 1 h at 37°C, half the volume was replaced by fresh medium and the culture was further incubated at 37°C, in 5% CO₂. Twice a week half of the medium was changed with fresh complete medium. Approximately 4–6 weeks after infections the first cell clusters were observed. The clusters were transferred to a new flask after a gentle trypsin-EDTA treatment. Thereafter, biweekly passing with fresh complete medium was performed. The cell lines termed AG101, derived from skin tissue, and AG116, derived from brain, were cloned by limiting dilution without feeder cells (Table 1).

Table 1 - Organs of AG129 (IFNr I und II 0/0) used to develop virus immortalized cell lines.

Full table (6K)

Isolation of murine migratory cutaneous DC and macrophages

Migratory cutaneous DC were obtained from wt 129Sv/Ev mice using the method of Ortner et al.,²⁰ with minor modifications. In brief, small pieces from tissue taken from both ears were incubated for 1 h with 1% PBS-buffered trypsin-EDTA at 37°C. Thereafter, the tissue was removed from epidermal sheets and incubated for 24 h in complete medium and Ag. Migrating wtDC were collected, counted and up to 90% were identified as DC by their typical features of 'veiled or hairy' morphology. The morphology was confirmed by electron microscopy where up to 10% were found to have Birbeck granules indicating immature DC. The isolated DC were used for MLR and Ag presentation without further purification.²⁰

Macrophages were isolated from wild-type 129Sv/Ev or AG129Sv/Ev mice by peritoneal lavage with complete medium as described.²³

Antigen uptake by the immortalized cells as determined by FACS and confocal microscopy

Determination of Ag uptake by the immortalized cells was assayed using 2 10⁵ cells pulsed with OVA conjugated with fluorescein (OVA-FITC, Molecular Probes, Eugene, OR, USA) as Ag. Graded doses (0.001–1 mg/mL) of Ag were added to cells kept on ice for 10 min. Thereafter, the cell Ag mixture was incubated at 37°C for various amounts of time. Control cells incubated with the same amount of Ag were kept on ice for the same period of time. Ag uptake was stopped by adding ice-cold FACS buffer (PBS containing 5% FCS and 0.01% sodium acide). Subsequently, the cells were washed three times with FACS buffer, fixed with FACS buffer containing 2% formaldehyde, followed by FACS analysis. Data on Ag binding of control cells that were kept on ice throughout the entire experiment were used as the control.

Determination of Ag uptake by cells analysed by confocal microscopy was performed by seeding and culturing 10⁴ cells/well in Lab-tek chamber slide systems (Nunc Inc., Naperville, IL, USA). After 24 h in culture, the medium was removed from the adherent cells, and replaced at various intervals by fresh medium containing OVA-FITC. After three gentle washes with cold FACS buffer, the cells were fixed with 1% formaldehyde in FACS buffer and analysed the same day using a Meridian confocal microscope⁹ from the FACS core facility, ETH, Zürich.

Antigen processing and stimulation of T cell hybridomas

In order to evaluate the Ag-processing activity of the AG129Sv/Ev derived cell lines, T cell hybridoma BO97.10.5²⁴ with specificity to OVA and c0ontrol T cell hybridomas were used.²³ Graded numbers (100–2000 cells/well) of APC were co-cultured in triplicates with 5 10⁴–5 10⁵ hybridoma cells in round-bottom 96-well plates with or without OVA as Ag (500 g/mL). Low input numbers of APC seeded were verified by direct counting of the cells using an inverted microscope. The activation of the hybridoma cells was determined by the production of IL-2, measured in the supernatants using a commercial ELISA (R&D Systems, Minneapolis, MN, USA). Supernatants were harvested after 48 h of co-culture. Data (reported in pg) represent the mean value of triplicate wells. The standard deviation of the triplicate values was between 10 and 20%. A total of 5.3 pg of IL-2 is equivalent to 1 unit of standard NIBSC/WHO 93/566 mouse IL-2 preparation, as indicated by the supplier.

Mixed lymphocyte reaction

The MLR was performed by co-cultivating variable numbers of AG101 or AG116 cells as stimulating APC with 5 10⁴–2 10⁵ responder splenic T cells of Balb/c mice (H-2^d). Spleen cells were depleted of adherent cells by incubation in polystyrene flasks at 37°C for 2 h. Thereafter the floating cells were harvested, and purified T cells were obtained by affinity isolation (Cellect, Biotex Laboratories Inc., Edmonton, Canada). The eluted T cells were seeded in round-bottom 96-well plates and kept in culture for 7 days. The activation of the responder T cells was analysed by determining tritiated thymidine incorporation and the production of IL-2 by ELISA (R&D Systems). For the ELISA, supernatant was taken after 7 days of culture and replaced by tritiated thymidine (185 MBq = 5 mCi, Dupont-New England Nuclear, Boston, MA, USA) at a final concentration of 37 mBq/well and left for another 18 h. The cells were harvested onto filter paper (Wallac, Turku, Finland) with a LKB Wallac cell harvester (Wallac), and thymidine incorporation was measured in a Betaplate-liquid scintillation counter (Wallac). Data are listed as counts per minute (c.p.m.) and represent the means of triplicate wells.

The function of cloned and immortalized DC lines AG101 and AG116 were compared to the Ag-presenting capacity of freshly isolated migratory cutaneous wtDC obtained from the ears of wild type 129Sv/Ev mice²⁰ or freshly isolated macrophages (1 10² –1 10⁵) obtained from wild type 129Sv/Ev mice after peritoneal lavage with complete medium as described earlier. In all cases where thymidine incorporation of the responding T cells was measured, the APC were irradiated using 3000 rad.

Antigen presentation and stimulation of naive T cells

The AG129Sv/Ev-derived cell lines AG101 and AG116 were tested for their capability to prime naive syngeneic splenic T cells prepared from wt 129Sv/Ev mice as described earlier (MLR). Graded numbers (20–2000 cells/well) of AG101 or AG116 cells were used as APC and incubated in triplicates in round-bottom 96-well plates with or without varying concentrations of Ag for at least 2 h. The capa-city of the AG101 and AG116 cells to stimulate naive T cells was compared using similar culture conditions with freshly isolated migratory cutaneous DC, or freshly isolated macrophages obtained from wt 129Sv/Ev or AG129Sv/Ev mice (1 10² –1 10⁵) prepared as described earlier (MLR). Thereafter, 2 10⁴ –1 10⁵ purified naive T cells were added to the APC and co-cultured for 7 days.

Levels of cytokines IL-1, IFN-, IL-2, IL-4, IL-6, IL-10, and IL-12-p40 secreted by the activated cells were determined with a commercial ELISA kit (R&D Systems) and compared to the standards provided by the supplier. IL-12p70 was determined using mAb 5C3²⁵ bound to anti-rat Ab pre-coated plates (Anawa Laboratories, Zürich, Switzerland). The IL-12p70 bound by mAb 5C3 was determined using a polyclonal Ab (R&D Systems). mAb 5C3 and IL-12p70 standards were obtained from Hoffmann-La Roche, Switzerland (courtesy of Dr Alber).

Infection of DC line AG116 with Borna disease virus

A monolayer of ~ 10⁶ AG116 in T25 flasks was overlaid with 10⁶ C6 cells²⁶ permanently infected with Borna disease virus (BDV): a non-cytopathic virus. After co-cultivating the two cell lines for 1 day, the AG116 cells were selected against C6 cells using 500 g/mL of G418¹⁷ pre-determined to be lethal for C6 cells. The presence of BDV in AG116 cells was determined by immunofluorescence using mAb²⁷ and reverse transcriptase polymerase chain reaction (RT-PCR) using appropriate primers.²⁸ The selected and re-cloned cell line termed AG116+BDV was re-analysed by FACS and shown to express the same markers as described for the parent AG116 cell line (data not shown). Cell line AG116+BDV kept in culture for more than 1 year remained permanently infected with BDV.

Injection of 129Sv/Ev mice with AG116+BDV cells

Mice of both sexes aged 6–10 weeks were inoculated with 10⁷–10³ infectious units AG116+BDV, i.p. Controls included C6+BDV, AG116+BDV (which were disrupted by four cycles of thawing and freezing), or the parent AG116 cells. Twenty-seven days after the cell inoculation, mice were bled retrobulbarly and the serum was analysed by ELISA.

Determination of Ab specific for BDV p40 or p24 by ELISA

The genes encoding BDVp40 or BDVp24²⁹ were expressed in Escherichia coli as fusion proteins with the albumin binding domain³⁰ and a six-histidine affinity tail for purification.³¹ To ensure correct folding and disulphide bond formation, appropriate leader peptides³² were used to export the fusion protein to the periplasmic space of E. coli. The expressed fusion protein was purified by affinity chromatography using Ni⁺ affinity purification.³¹ To perform the ELISA, saturating amounts of BDVp40 or BDVp24 fusion proteins or the albumin-binding domain as control was added to rat serum albumin pre-coated plates (Anawa, Laboratories, Dübendorf, Switzerland), exploiting the high affinity between rat serum albumin and the albumin-binding domain.³⁰ This Ag coating procedure avoids denaturation of proteins bound to solid phase.³³ Phosphate-buffered saline Tween 20 (0.05%) was used as buffer for Ag coating, for the dilution of all reagents used in ELISA and to wash the plates after each incubation step. Two-fold serum dilutions and the detection reagents labelled with alkaline phosphatase were incubated each for 1 h at 37°C. A polyclonal goat Ab specific against murine Ig directly labelled with phosphatase was from Anawa. Biotinylated rat mAb specific to the mouse Ig isotypes and streptavidin labelled with phosphatase were from Phar Mingen (Lugano, Switzerland). After the final wash, the substrate P-nitrophyl-phosphate (Sigma) was added and the optical density at 450 nm (OD₄₅₀) determined after 30 min. The titre was defined as the reciprocal of the highest serum dilution with the absorbency twice that of the background. Pre-immune sera were used to determine the background.

Results

Immortalization of DC-like cells from organs of AG129Sv/Ev, G129Sv/Ev, 129Sv/Ev

Cells prepared from various organs obtained from AG129Sv/Ev, G129Sv/Ev or wt129Sv/Ev mice were infected with the MIBY2-N11 retrovirus (Table 1). After 1 month in culture, several adherent cell lines with a variety of individual morphological features including dendritic cell extensions of different shape were obtained (Figure 1a,b). The DC-like morphology was typical for each cell line analysed and did not change in culture over time. All the immortalized cells attached to plastic. The cells grew in culture without externally added cytokines. Some of the cell lines have been kept in culture for more than 1 year without gross alteration of the characteristic phenotype of each cell line.

Figure 1.

Microscopical appearance of the cloned (a) AG101 (skin) and (b) AG116 (brain) cell lines. The irregular shaped cells with long dendrites and veils are typical for dendritic cells.

Full figure and legend (118K)

The cell lines AG101 and AG116 with non-functional type-I and -II IFN-r appeared to have different growth rates than G129#3 with defective type-II IFN-r or 129Sv/Ev thymus carrying wt IFN-r. To analyse the influence of the IFN-r on growth rates, ³H thymidine incorporation was determined for 10 000 cells from cell lines AG101, AG116, G129#3 and 129Sv/Ev thymus seeded in quadruplicates in 96-well plates. The incorporation of the label was determined 18 h later. The uptake of ³H thymidine resulted in 54 560 c.p.m. 2343 for the cell line AG101; 52 097 c.p.m. 8621 for AG116; 4337 c.p.m. 789 for the cell line G129#3; and 1110 c.p.m. 98 for the cell line 129Sv/Ev thymus. Therefore, the absence of functional IFN-r type II resulted in a four-fold growth advantage over wt cells, but the absence of functional IFN-r I and II was associated with a 50-fold growth advantage over the wt cells. A detailed morphological and functional analysis of the DC-like cells was mainly performed with the lines AG101 and AG116 (Table 2). For this, cells from lines AG101 and AG116 were cloned by limiting dilution without the addition of feeder cells.

Table 2 - Determination of the Ab titre against BDV p40 and p24.

Full table (7K)

Surface analysis of the DC lines

To further characterize the phenotype of cell lines with dendritic morphology, the cloned cell lines AG101 and AG116 and some non-cloned cell lines derived from skin, brain and thymus were analysed by FACS. Forward and side scatter clearly showed that the cells were several times the size of normal mouse lymphocytes and showed extensive granularity. The cells were positive for the myeloid-dendritic cell-associated molecules F4/80, N418 and NLDC145, expressed the costimulatory molecules B7.2 and the adhesion and homing molecules intercellular adhesion molecule (ICAM)-1 and CD11b. Low amounts of MHC class II molecules were detected. Representative FACS profiles for the cell line AG101 is shown in Figure 2.

Figure 2.

Surface markers of AG101 as analysed by FACS analysis. The fluorescence intensity is graphed as histograms against the cell number. Open histograms represent staining of cells with isotype-matched Ig; filled histograms were obtained using mAb against the following markers resulting in different numbers of positive cells (⁺): (a) MHC class II (46%⁺); (b) B7.2 (43%⁺); (c) F4/80 (83%⁺); (d) MAC I (50%⁺); (e) NLDC 145 (43%⁺); (f) N418 (36%⁺). Staining was performed as described in Materials and Methods. Before the labelling experiments, Fc receptor (FcR) blocking was performed by incubating cells with anti FcR (2.4G2) mAb.

Full figure and legend (16K)

The skin-, brain- and thymus-derived cell lines also expressed significant amounts of the Thy-1 marker, but expressed either low amounts of CD4 molecules (AG116) or no CD4. All the cell lines stained negative for both CD8 and B220. Thus, representative markers for B and cytotoxic T cells were absent from the cell lines analysed. Further functional characterization of the DC-like cells was restricted to the cloned cell lines AG101 and AG116. All functional assays were performed without the addition of externally added stimulatory molecules.

Antigen uptake and processing by DC lines

Dendritic cells in stage 1 have the capacity to take up and process Ag. In a second stage in the development of DC, the Ag-presenting and T cell stimulatory capabilities are acquired.^{9, 10, 11} We thus analysed Ag uptake by AG101 and AG116 using both FACS and confocal microscopy. The FITC-labelled OVA was taken up and was degraded within 2 h, as visualized by confocal microscopy (Figure 3). The capacity of AG101 and AG116 cells to take up and process Ag was thus consistent with the stage 1 of DC. To evaluate the potential of AG101 and AG116 cells to function as stage 2 DC, it was important to analyse the capacity of these cells to present Ag and to stimulate T cells.

Figure 3.

Uptake and processing of ovalbumin (OVA)-FITC by AG101 cells as analysed by confocal laser scanning microscopy. (a) Two min after the addition of OVA-FITC, some protein has been taken up by the cells and can be detected at a Z position of ~ 42 m. (b) At 30 min, the cells have expanded in size and the intensity of the fluorescence has increased in specific cellular compartments. (c) At 120 min, no fluorescence signal can be detected in the cells, which appear normal in size.

Full figure and legend (127K)

Antigen processing and stimulation of T cell hybridomas

The capability of the cell lines AG101 and AG116 to process native soluble proteins like OVA and to present Ag-derived peptides to MHC II-restricted T cells was investigated next. For this, the T cell hybridoma BO97.10.5 specific for OVA peptides bound to H-2^b was used. Ag-specific activation of the T cell hybridoma was measured by their IL-2 production. As few as 50–100 Ag-pulsed AG101 or AG116 cells acting as APC reproducibly triggered the production of 20–70 pg/mL of IL-2 by this T cell hybridoma. The gradual increase in the number of APC to 2000 cells while keeping the number of T cells constant led to an increased production of IL-2 up to 300 pg/mL. The T cell hybridoma BO97.10.5 cultured in the presence of 2000 APC of either AG101 or AG116 cells, but in the absence of Ag it did not result in detectable amounts of IL-2 production (Figure 4). The T cell hybridomas E3 or E8 specific to OVA but restricted to I-E^b did not result in IL-2 production when used with either AG101 or AG116 (data not shown).

Figure 4.

AG101 and AG116 present MHC restricted antigen to the T cell hybridoma BO97.10.5. A total of 10⁵ BO97.10.5 T cells were incubated with ovalbumin (OVA; 500 g/mL) alone or with different numbers of () AG101 (skin) or () AG116 (brain) cells and kept in culture for 2 days. Thereafter, IL-2 released by the T cells was determined by ELISA.

Full figure and legend (7K)

Induction of mixed leucocyte reaction

Antigen presentation and priming of naive T cells are some of the most important functions of DC performed in vivo. Part of this function can be mimicked in vitro using a MLR. Therefore, the cell lines AG101 or AG116 were irradiated and used as stimulators for purified splenic T cells from Balb/c mice (H-2^d) as responder cells. Increasing the number of DC added to the culture was paralleled by an increased production of IL-2 by the responding T cells. The data were compared to freshly isolated and cultured cutaneous DC²⁰ from wt 129Sv/Ev mice, as shown in Figure 5, for a representative experiment. The IL-2 production was very similar in all experiments, irrespective of whether wtDC, AG101 or AG116 cells were used as APC. In some experiments, peritoneal macrophages from wt129Sv/Ev mice were also used in MLR. At least 100 times more macrophages than DC had to be used as stimulator cells to induce similar amounts of IL-2, indicating that AG101 or AG116 had T cell-stimulating functions similar to DC rather than macrophages (data not shown).

Figure 5.

AG101 and AG116 cells are able to provoke a mixed lymphocyte reaction. A total of 10⁵ purified splenic T cells from MHC mismatched Balb/c mice were co-cultivated alone or in increasing numbers of freshly isolated cutaneous () wild-type dendritic cells (wtDC); () AG101 (skin); and () AG116 (brain) cells and kept in culture for 7 days. Thereafter, IL-2 released by the T cells was determined by ELISA.

Full figure and legend (7K)

Priming of naive T cells in vitro using soluble antigen

The AG129Sv/Ev-derived cell lines AG101 and AG116 were tested for their capability to prime naive purified syngeneic splenic T cells prepared from wt 129Sv/Ev mice, and were compared to freshly isolated cutaneous wtDC. The irradiated cell lines AG101 or AG116 and wtDC ranging from 20 to 2000 cells/well were each incubated with OVA as Ag and co-cultivated with 10⁵ purified T cells. After 7 days in culture, IL-2 was measured and thymidine incorporation was determined (Figure 6,Figure 7). Increasing the number of APC resulted in an increased IL-2 production and thymidine incorporation. The IL-2 production increased from seven- (wtDC) to 13-fold (AG101), whereas the proliferation of T cells increased from 24- (wtDC) to 30-fold (AG101). Therefore, based on the criteria of IL-2 production and thymidine incorporation, AG101 or AG116 were at least as efficient in priming naive T cells as wtDC.

Figure 6.

Naive T cells proliferate in the presence of antigen, AG101 and AG116 cells. A total of 10⁵ purified splenic T cells from 129Sv/Ev mice were cultivated with ovalbumin (OVA; 500 g/mL) alone or with increasing numbers of freshly isolated cutaneous () wild-type dendritic cells (wtDC); () AG101 (skin); and () AG116 (brain) cells and kept in culture for 7 days. Thereafter, the cultures were pulsed with ³H thymidine and the incorporation of the label was determined 18 h later.

Full figure and legend (9K)

Figure 7.

Naive T cells produce IL-2 in the presence of antigen, AG101 and AG116. A total of 10⁵ purified splenic T cells from 129Sv/Ev mice were cultivated with ovalbumin (OVA; 500 g/mL) alone or with increasing numbers of freshly isolated cutaneous () wild-type dendritic cells (wtDC); () AG101 (skin); and () AG116 (brain) cells and kept in culture for 7 days. Thereafter, IL-2 released by the T cells was determined by ELISA.

Full figure and legend (9K)

IL-6 and IL-12p40 are up-regulated in AG101 or AG116 during antigen-specific cognate interaction with T cells

To investigate which pro-inflammatory cytokine(s) might facilitate priming of T cells, the constitutive and induced expression of IL-1, TNF-, IFN-, IL-4, IL-6, IL-12p40, and IL-12p70 was determined. Of all the cytokines tested, IL-6 was present at 120 19 ng/mL per 7 days per 10⁶ irradiated AG101 cells; 3.8 0.27 ng/mL per 7 days per 10⁶ irradiated AG116 cells; and IL-12p40 was present at 10 3 pg/mL per 72 h in cultures of either 10⁶ Ag101 or Ag116.

After cognate interaction with T cells and Ag, IL-6 was found to be up-regulated 20–100-fold and IL-12p40 was found to be up-regulated more than 1000-fold using either of the DC lines AG101 or AG116. Interleukin-6 and IL-12p40 were up-regulated with increasing wtDC used (Figure 8,Figure 9). IL-12p70 was not detected under the culture conditions used.

Figure 8.

IL-6 is up-regulated in the presence of antigen, AG101 and AG116 cells. A total of 10⁵ purified splenic T cells from 129Sv/Ev mice were cultivated with ovalbumin (OVA; 500 g/mL) alone or with increasing numbers of freshly isolated cutaneous () wild-type dendritic cells (wtDC); () AG101 (skin); and () AG116 (brain) cells and kept in culture for 7 days. Thereafter, IL-6 released by DC was determined by ELISA. ^*Wild type not determined.

Full figure and legend (8K)

Figure 9.

IL-12p40 is up-regulated by AG101 and AG116 cells in the presence of antigen and naive T cells. A total of 10⁵ purified splenic T cells from 129Sv/Ev mice were cultivated with ovalbumin (OVA; 500 g/mL) alone or with increasing numbers of freshly isolated cutaneous () wild-type dendritic cells (wtDC); () AG101 (skin); and () AG116 (brain) cells and kept in culture for 7 days. Thereafter, IL-12p40 released by the DC was determined by ELISA. ^*Wild type not determined.

Full figure and legend (7K)

Induction of Ab in vivo against BDVp40 and BDVp24 using virus-infected DC for immunization

AG116 cells permanently infected with BDV were used to determine the power of the newly developed cells to induce a humoral immune response against BDV proteins. Of special interest was the Ab response to BDVp24, which is present in low amounts.³⁴ Adult 129Sv/Ev mice were injected with different numbers of AG116+BDV or control cells (Table 2). Ab against BDVp40 were induced using intact or disrupted AG116+BDV or C6 infected with BDV. In general, the Ab titre was highest using intact AG116+BDV when comparable cell numbers were used for immunization (Table 2). Cells not infected with BDV did not induce Ab against the BD viral proteins. However, Ab against BDVp24 expressed at low levels by the virus were only induced when AG116+BDV cells were used for immunization (Table 2). As expected with virus used for immunization, most of the Ab response was of the IgG2a isotype irrespective of the immunization schedule used.

Discussion

Studies mimicking the maturation process of DC in vitro^{2, 3, 9, 10, 11} and recently ex vivo¹¹ provided evidence for at least two functional stages of DC recognized by their transitory phenotypes and functions. In a first stage, DC show low mobility, are adherent and their MHC class II molecules are mainly contained in the cytoplasm. At this stage the cells are very potent in Ag uptake.^{9, 11} Stimulation by infectious particles, inflammatory signals, or particular cytokines are required to induce the progression from stage 1 towards stage 2. Upon progression to this second stage, DC lose their ability to take up Ag. Instead, they stably translocate MHC class II molecules to the cell surface, up-regulate adhesion molecules and may acquire the ability to produce pro- inflammatory cytokines.^{9, 10, 11} However, it is not clear how the different signals guide the DC through the different functional stages and what the role of the IFN system on the various functions of DC is. Therefore, the aim of the experiments was to directly compare phenotype and functions of clonally derived DC devoid of the IFN system with those of cutaneous wtDC.

Dendritic cell lines AG101 or AG116 obtained from IFN-r type-I- and type-II-deficient mice¹⁸ attached to plastic, displayed low levels of surface-expressed MHC class II molecules and were efficient in Ag uptake. Thus, AG101 or AG116 have cell surface and functional characteristics consistent with immature DC of stage 1. Surprisingly, the same DC cultured and used in all experiments without the addition of externally added cytokines or non-specific stimuli were very efficient in promoting cognate interaction with wt T cells. This was shown by Ag-specific stimulation of MHC-restricted T cell hybridomas, MLR, or priming of naive T cells both in vitro and in vivo. The resulting proliferation of T cells and production of cytokines during the cognate interaction with AG101 or AG116 were at least as efficient as with wtDC isolated from skin. Moreover, the cognate interaction was at least 100 times more efficient than with wt peritoneal macrophages used in the same set of experiments.

Paglia et al. generated wtDC lines from various mouse strains using the same protocol for immortalization of DC as used in the present experiments.³⁵ The wtDC lines appeared to be frozen in the immature stage 1 with low functional capacity, unless different cytokines or other non-specific stimuli were added to enhance function, or progression to the second stage. Only then were the cells able to take up Ag and present Ag to MHC restricted to T cells.³⁶ Similar results were obtained using a cell line kept in culture with a cocktail of cytokines but without immortalization of the cells.⁹ In contrast, AG101 or AG116 appeared to combine several in vitro characteristics of DC, normally assigned to individual stages. Therefore, we propose that the missing IFN-r might explain the particular functions of AG101 or AG116.

Interferons have very broad biological activities. On one hand, they include the well-known immune-enhancing functions such as up-regulation of MHC class II molecules on the cell surface and activation of NK cells.³⁷ On the other hand, IFN-/ induce anti-proliferative effects,^{13, 38} inhibition of differentiation³⁹ or depression of haematopoiesis after particular viral infection.¹⁵ Indeed, immortalization of DC from various organs of AG129Sv/Ev mice by the retrovirus MIBY2-N11^{21, 35} was very efficient compared to similar attempts using cells from G129Sv/Ev or wt129Sv/Ev mice (Table 1). Furthermore, the cell lines derived from organs of AG129Sv/Ev mice grew vigorously in culture as indicated by a 10-times higher thymidine incorporation, compared to G129#3 with an intact IFN-r I, or 50 times higher than wt129Sv/Ev thymus cells. AG101 and AG116 appear to have lost inhibition of contact and are tumourigenic in nude mice but not in wt129Sv/Ev mice (data not shown).

Since AG101 or AG116 are devoid of the immune-enhancing functions exerted by the IFN system, we expected that the cell lines would be poor stimulators of T cells. We observed quite the opposite because the AG129 DC are at least as effective as dermal wtDC matured in vitro.

The power to induce cognate interactions by AG101 or AG116 with T cells may be the result of the pro-inflammatory cytokines IL-6 and IL-12, or possibly the other cytokines induced such as IL-18.⁴⁰ After T cell interaction, IL-6 (20–100-fold) and IL-12p40 (100–1000-fold) was similarly up-regulated in either AG101, AG116 or mature wtDC. This argues for the maintained regulation of both IL-12 and IL-6 during cognate interaction in the absence of IFN-r.^{41, 42}

It has been proposed that high amounts of IL-6 could induce IL-4 production in neighbouring cells such as T cells, thus providing an environment favourable for the induction of a Th2 immune response.⁴³ It is noteworthy that in all our assays IL-4 was never detected, even though appreciable amounts of IL-6 were present. It is possible that the secretion of IL-12 by APC after cognate interaction with T cells is not efficiently controlled by IFN,⁴² and IL-12 may therefore dominate over IL-6. Alternatively, IL-6 may not be present in large enough amounts to provide an environment favourable for a Th2-like immune response.⁴³ Other Th2-inducing molecules such as IL-10 were not detected in our assays.

AG116 stably infected with BDV as Ag was used to induce a specific humoral anti-viral immune response in wt mice for the following reasons: (i) BDV neither replicates in adult mice nor in AG129Sv/Ev when culture-adapted virus is used for infection, or in a brain tumour induced by AG116+BDV (Suter et al., unpubl. data 1997); (ii) BDV is not cytopathic for DC or other cells such as C6,²⁶ and only very few copies (0.01–0.05) of infective virus per individual cell are present at the time;³⁴ and (iii) in permanently infected cell lines BDV nucleoprotein p40 is expressed in higher amounts than the BDV phosphoprotein p24.²⁹ We reasoned that the limited amount of Ag, in particular that of BDV phosphoprotein p24 present in cells used for immunization, would be a measure of the Ab-inducing system. All immunization regimens containing cell-associated BDV led to the production of Ab against BDV nucleoprotein p40 (Table 2). In contrast, only AG116 infected with BDV induced detectable amounts of Ab against the BDV phosphoprotein p24, which is present in limited amounts in infected cells.³⁴ The data indicate that intact AG116+BDV induce a stronger Ab response against the BDV nucleoprotein p40 when compared to an Ab response against the phosphoprotein p24. Repeated injections of destructed BDV-containing cells or at least 100 times higher cell numbers of BDV-infected C6 were needed to achieve similar levels of Ab responses. Most of the Ab response was of the IgG2a isotype, which appears to be typical for viral infections⁴⁴ but emphasizes the need for T cell help.¹⁸ The Ab response induced in vivo substantiated the results obtained in vitro using T cell stimulation as the main readout. The fact that specific IgG2a anti-BDVp24 Ab were exclusively induced using intact AG116+BDV as immunogen, raises the possibility that AG116 not only primes naive T cells but may also prime naive B cells.^{45, 46} However, it is not clear whether AG116+BDV simply survive longer in the animals than C6+BDV cells, and thus more Ag is potentialy available for a more powerful induction of an immune response, or whether the AG116+BDV cells are indeed able to migrate to the lymphoid organs. We presently test the hypothesis that AG116+BDV can migrate to lymphoid organs and present Ag directly to T cells.¹² In lymphoid organs, some Ag from the T cell area might be transported to the B cell area.⁴⁷

Dendritic cells represent a broad collection of cells grouped in at least three subgroups.³ The heterogeneity among the three groups is unknown but can be appreciated by comparing skin-derived AG101 and brain-derived AG116, which have similar but distinct potency to stimulate T cells as analysed in in vitro assays. Some of the data might be explained by the constitutive expression of IL-6, which was 30 times higher in AG101 than AG116. On the other hand, infecting AG116 with BDV was easy but several attempts to infect AG101 with BDV were unsuccessful, indicating maintained tissue tropism of AG116 for BDV, a virus present normally in the brain.²⁹

At the molecular level we start to better appreciate the complexity of cognate interaction between APC and T cells at the cell surface receptor level.⁴⁸ This cognate interaction has to be analysed further, starting from the level of receptor to the signal transduction machinery influencing the transcription of cytokines or cell surface receptors in the presence or absence of IFN and their receptors.

The immortalized cell lines AG101 or AG116 will now enable research to be carried out that addresses molecular cell–cell interactions at the clonal level and in the absence of functional IFN-r. The same cells will promote the performance of such analysis under the influence of graded (re)expression of the deleted genes encoding for the appropriate IFN-r type I or II using inducible gene expression systems.⁴⁹ The re-expressed genes encoding for the IFN-r might further clarify the role that the IFN system plays in cell division, cytokine regulation or cognate interaction. AG101 or AG116 cells could further be used to better understand the parasite–host relationship in the presence or absence of the IFN system.

Finally, AG101 or AG116 loaded with Ag could be used to study cell migration and the initiation of an immune response in vivo. Such data could ultimately be used for improved immunotherapy.

References

1. Katz SI, Cooper KD, Iijima M, Tsuchida T. The role of Langerhans cells in antigen presentation. J. Invest. Dermatol. 1985; 85 2: S96–8. | Article |
2. Steinman RM. The dendritic cell system and its role in immunogenicity. Annu. Rev. Immunol. 1991; 9: 271–96. | Article | PubMed | ISI | ChemPort |
3. Cella M, Sallusto F, Lanzavecchia A. Origin, maturation and antigen presenting function of dendritic cells. Curr. Opin. Immunol. 1997; 9: 10–16. | Article | PubMed | ISI | ChemPort |
4. Guzman CA, Rohde M, Chakraborty T et al. Interaction of Listeria monocytogenes with mouse dendritic cells. Infect. Immun. 1995; 63: 3665–73. | PubMed | ChemPort |
5. Moll H, Flohe S, Rollinghoff M. Dendritic cells in Leishmania major-immune mice harbor persistent parasites and mediate an antigen-specific T cell immune response. Eur. J. Immunol. 1995; 25: 693–9. | PubMed | ISI | ChemPort |
6. Cameron P, Pope M, Granelli-Piperno A, Steinman RM. Dendritic cells and the replication of HIV-1. J. Leukoc. Biol. 1996; 59: 158–71. | PubMed | ISI | ChemPort |
7. Grabbe S, Beissert S, Schwarz T, Granstein RD. Dendritic cells as initiators of tumor immune responses: A possible strategy for tumor immunotherapy? Immunol. Today 1995; 16: 117–21. | Article | PubMed | ISI | ChemPort |
8. Mayordomo JI, Zorina T, Storkus WJ et al. Bone marrow-derived dendritic cells pulsed with synthetic tumour peptides elicit protective and therapeutic antitumour immunity. Nature Med. 1995; 1: 1297–302. | Article |
9. Winzler C, Rovere P, Rescigno M et al. Maturation stages of mouse dendritic cells in growth factor-dependent long-term cultures. J. Exp. Med. 1997; 185: 317–28. | Article | PubMed | ISI | ChemPort |
10. Cella M, Engering A, Pinet V, Pieters J, Lanzavecchia A. Inflammatory stimuli induce accumulation of MHC class II complexes on dendritic cells. Nature 1997; 388: 782–7. | Article | PubMed | ISI | ChemPort |
11. Pierre P, Turley SJ, Gatti E et al. Developmental regulation of mouse class II transport in mouse dendritic cells. Nature 1997; 388: 787–92. | Article | PubMed | ISI | ChemPort |
12. Steinman RM, Pack M, Inaba K. Dendritic cells in the T-cell areas of lymphoid organs. Immunol. Rev. 1997; 156: 25–37. | Article | PubMed | ISI | ChemPort |
13. Sen GC, Lengyel P. The interferon system. A bird's eye view of its biochemistry. J. Biol. Chem. 1992; 267: 5017–20. | PubMed | ISI | ChemPort |
14. Balkwill F, Pitha PM. Interferons and cytokines on a grand scale. Cytokine Growth Factor Rev. 1997; 8: 91–5. | Article | PubMed | ChemPort |
15. Binder D, Fehr J, Hengartner H, Zinkernagel RM. Virus-induced transient bone marrow aplasia: Major role of interferon-alpha/beta during acute infection with the noncytopathic lymphocytic choriomeningitis virus. J. Exp. Med. 1997; 185: 517–30. | Article | PubMed | ISI | ChemPort |
16. Moritz T, Kloke O, Nagel-Hiemke M et al. Tumor necrosis factor alpha modifies resistance to interferon alpha in vivo: First clinical data. Cancer Immunol. Immunother. 1992; 35: 342–6. | Article | PubMed | ChemPort |
17. Huang S, Hendriks W, Althage A et al. Immune response in mice that lack the interferon-gamma receptor. Science 1993; 259: 1742–5. | Article | PubMed | ISI | ChemPort |
18. Muller U, Steinhoff U, Reis LF et al. Functional role of type I and type II interferons in antiviral defense. Science 1994; 264: 1918–21. | Article | PubMed | ISI | ChemPort |
19. Steinhoff U, Muller U, Schertler A, Hengartner H, Aguet M, Zinkernagel RM. Antiviral protection by vesicular stomatitis virus-specific antibodies in alpha/beta interferon receptor-deficient mice. J. Virol. 1995; 69: 2153–8. | PubMed | ISI | ChemPort |
20. Ortner U, Inaba K, Koch F et al. An improved isolation method for murine migratory cutaneous dendritic cells. J. Immunol. Methods 1996; 193: 71–9. | Article | PubMed | ISI | ChemPort |
21. Sassano M, Granucci F, Seveso M, Marconi G, Foti M, Ricciardi-Castagnoli P. Molecular cloning of a recombinant retrovirus carrying a mutated envAKR-mycMH2 fusion gene immortalizing cells of the monocytic-macrophage lineage. Oncogene 1994; 9: 1473–7. | PubMed | ChemPort |
22. Lutz MB, Granucci F, Winzler C et al. Retroviral immortalization of phagocytic and dendritic cell clones as a tool to investigate functional heterogeneity. J. Immunol. Methods 1994; 174: 269–79. | Article | PubMed | ChemPort |
23. Koch F, Trockenbacher B, Kampgen E et al. Antigen processing in populations of mature murine dendritic cells is caused by subsets of incompletely matured cells. J. Immunol. 1995; 155: 93–100. | PubMed | ISI | ChemPort |
24. Roehm NW, Marrack P, Kappler JW. Antigen-specific, H-2-restricted helper T cell hybridomas. J. Exp. Med. 1982; 156: 191–204. | Article | PubMed | ISI | ChemPort |
25. Guery JC, Ria F, Galbiati F, Adorini L. Normal B cells fail to secrete interleukin-12. Eur. J. Immunol. 1997; 27: 1632–9. | PubMed | ChemPort |
26. Carbone KM, Rubin SA, Sierra-Honigmann AM, Lederman HM. Characterization of a glial cell line persistently infected with borna disease virus (BDV): Influence of neurotrophic factors on BDV protein and RNA expression. J. Virol. 1993; 67: 1453–60. | PubMed | ChemPort |
27. Thiedemann N, Presek P, Rott R, Stitz L. Antigenic relationship and further characterization of two major Borna disease virus-specific proteins. J. Gen. Virol. 1992; 73: 1057–64. | PubMed | ChemPort |
28. Sauder C, Muller A, Cubitt B et al. Detection of Borna disease virus (BDV) antibodies and BDV RNA in psychiatric patients: Evidence for high sequence conservation of human blood-derived BDV RNA. J. Gen. Virol. 1996; 70: 7713–24. | ChemPort |
29. Schneemann A, Schneider PA, Lamb RA, Lipkin WI. The remarkable coding strategy of borna disease virus: A new member of the nonsegmented negative strand RNA viruses. Virology 1995; 210: 1–8. | Article | PubMed | ChemPort |
30. Nygren PA, Eliasson M, Abrahmsen L, Uhlen M, Palmcrantz E. Analysis and use of the serum albumin binding domains of streptococcal protein G. J. Mol. Recogn. 1988; 1: 69–74. | ChemPort |
31. Dudler T, Chen WQ, Wang S et al. High-level expression in Escherichia coli and rapid purification of enzymatically active honey bee venom phospholipase A2. Biochem. Biophys. Acta 1992; 1165: 201–10. | PubMed | ChemPort |
32. Crameri R, Suter M. Display of biologically active proteins on the surface of filamentous phages: A cDNA cloning system for the selection of functional gene products linked to the genetic information responsible for their production. Gene 1993; 137: 69–75. | Article | PubMed | ISI | ChemPort |
33. Suter M, Butler J. The immunochemistry of sandwich ELISAs. II. A novel system prevents the denaturation of capture antibodies. Immunol. Lett. 1986; 13: 313–16. | Article | PubMed | ChemPort |
34. Pauli G, Ludwig H. Increase of virus yields and releases of Borna disease virus from persistently infected cells. Virus Res. 1985; 2: 29–33. | Article | PubMed | ChemPort |
35. Paglia P, Girolomoni G, Robbiati F, Granucci F, Ricciardi-Castagnoli P. Immortalized dendritic cell line fully competent in antigen presentation initiates primary T cell responses in vivo. J. Exp. Med. 1993; 178: 1893–901. | Article | PubMed | ISI | ChemPort |
36. Lutz MB, Assmann CU, Girolomoni G, Ricciardi-Castagnoli P. Different cytokines regulate antigen uptake and presentation of a precursor dendritic cell line. Eur. J. Immunol. 1996; 26: 586–94. | PubMed | ISI | ChemPort |
37. DeMaeyer E, DeMaeyer-Guignard J. Interferons and Other Regulatory Cytokines. New York: Wiley, 1988.
38. Billiau A. Interferon-gamma: Biology and role in pathogenesis. Adv. Immunol. 1996; 62: 61–130. | PubMed | ISI | ChemPort |
39. Pestka S, Langer JA, Zoon KC, Samuel CE. Interferons and their actions. Ann. Rev. Biochem. 1987; 56: 727–77. | Article | PubMed | ChemPort |
40. Okamura HH, Tsutsi T, Komatsu M et al. Cloning of a new cytokine that induces IFN-gamma production by T cells. Nature 1995; 378: 88–91. | Article | PubMed | ISI | ChemPort |
41. Heufler CF, Koch U, Stanzl G et al. Interleukin-12 is produced by dendritic cells and mediates T helper 1 development as well as interferon-gamma production by T helper 1 cells. Eur. J. Immunol. 1996; 26: 659–68. | PubMed | ISI | ChemPort |
42. Cousens LP, Orange JS, Su HC, Biron CA. Interferon-alpha/beta inhibition of interleukin 12 and interferon-gamma production in vitro and endogenously during viral infection. Proc. Natl Acad. Sci. USA 1997; 94: 634–9. | Article | PubMed | ChemPort |
43. Rincon M, Anguita J, Nakamura T, Fikrig E, Flavell RA. Interleukin (IL)-6 directs the differentiation of IL-4-producing CD4+ T cells. J. Exp. Med. 1997; 185: 461–9. | Article | PubMed | ISI | ChemPort |
44. Coutelier JP, van der Logt JT, Heessen FW, Vink A, van Snick J. Virally induced modulation of murine IgG antibody subclasses. J. Exp. Med. 1988; 168: 2373–8. | Article | PubMed | ChemPort |
45. Bjorck P, Flores-Romo L, Liu YJ. Human interdigitating dendritic cells directly stimulate CD40-activated naive B cells. Eur. J. Immunol. 1997; 27: 1266–74. | PubMed | ISI | ChemPort |
46. Dubois B, Vanbervliet B, Fayette J et al. Dendritic cells enhance growth and differentiation of CD40-activated B lymphocytes. J. Exp. Med. 1997; 185: 941–51. | Article | PubMed | ISI | ChemPort |
47. Tew GJ, Wu J, Qin D, Helm S, Burton GF, Szakal AK. Follicular dendritic cells and presentation of antigen and costimulatory signals to B cells. Immunol. Rev. 1997; 156: 39–52. | Article | PubMed | ISI | ChemPort |
48. Lanzavecchia A. Understanding the mechanism of sustained signaling and T cell activation. J. Exp. Med. 1997; 185: 1717–19. | Article | PubMed | ISI | ChemPort |
49. Gossen M, Freundlieb S, Bender G, Muller G, Hillen W, Bujard H. Transcriptional activation by tetracyclines in mammalian cells. Science 1995; 268: 1766–9. | Article | PubMed | ISI | ChemPort |

Acknowledgements

Eva Niederer, Luis Filgueira and Peter Wild were very helpful in FACS analysis, confocal microscopy and microscopy. Cell lines were provided by Paola Ricciardi-Castagnoli, CNR, Center of Cytopharmacology, University of Milan, Italy; Stephen I Katz, Department of Dermatology, NIH, Bethesda, MD; Philippa Marrack, Jewish Medical Center, Denver, CO; Luciano Adorini, Roche Milano Richerca, Milano, Italy; Gerold Schuler, University of Erlangen, Erlangen, Germany and Nikolaus Romani, University of Innsbruck, Austria. We thank Paola Ricciardi-Castagnoli and her team for intellectual input in this work; Ken Shortman, WEHI, Melbourne, Australia; Antonio Lanzavecchia, Basel Institute for Immunology, Basel; and Rolf Zinkernagel, University of Zürich for discussing the data. Jeanne Straub from our Institute carefully edited the manuscript.

This investigation was funded by grants from the Kanton Zürich, Swiss National Science Foundation and Bundesamt für Bildung und Wissenschaft, Switzerland.