© 2003 Wiley-Liss, Inc.
Cytometry Part A 54A:19 –26 (2003)
Sensitive Visualization of Peptide Presentation In Vitro and Ex Vivo De´sire´e Kunkel,1 Dennis Kirchhoff,1 Rudolf Volkmer-Engert,2 Andreas Radbruch,1 and Alexander Scheffold1* 1
Deutsches Rheuma-Forschungszentrum Berlin, Germany Charite´, Humboldt University-Charite´, Berlin, Germany
2
Received 8 August 2002; Revision Received 14 February 2003; Accepted 7 March 2003
Background: The presentation of antigenic peptides to specific T cells is one of the key events for the induction of a T-cell-dependent immune response. The nature of the antigen-presenting cells which present distinct peptides has been difficult to analyze so far due to the low number of peptides presented in vivo by a single antigen-presenting cell. Methods: We have used magnetofluorescent liposomes to identify and characterize antigen-presenting cells according to presentation of a hapten-labeled antigenic peptide in vitro and ex vivo. Results: Magnetofluorescent liposomes allowed the identification and isolation of antigen-presenting cells according to the presentation of less than 100 peptides per cell, the physiological threshold for activation of specific T cells. Ex vivo, we could demonstrate peptide presentation
The presentation of major histocompatibility complex (MHC) class II/peptide-complexes on the surface of professional antigen-presenting cells (APC), i.e., dendritic 兾 ), to cells (DC), B lymphocytes, and macrophages (MO naive CD4⫹ T cells is a crucial step for the induction of an adaptive immune response as well as for the generation of antigen-specific tolerance. The type of APC or its activation status, in particular its capacity to provide costimulatory signals, are considered as the major factors for the induction of T-cell differentiation to a reactive versus a tolerant or suppressive state (1,2). Despite the decisive role of the signals provided by the antigen-presenting cell during the activation of a naı¨ve T cell, relatively little is known about the phenotype and frequency of antigenpresenting cells, which are responsible for the induction of an immune reaction in vivo following tolerogenic or immunogenic ways of antigen application. Monoclonal antibodies recognizing specific MHC class II/peptide complexes have been used to analyze peptide presentation in vivo and in vitro (3–7); however, cross-reactivity and lack of sensitivity have limited their usefulness (4,7). Under physiological conditions, professional APC present a multitude of different peptides derived from exogenous and
by B lymphocytes and dendritic cells already 1 h after intravenous peptide injection; this rapidly declined to background level after 12–24 h. Conclusions: The sensitive visualization of peptide presentation allows the phenotypical and functional characterization of those antigen-presenting cells which present specific peptides at physiological relevant quantities. This technology will help to characterize the antigen-presenting cells (APC) which are responsible for the induction of distinct immune reactions in vivo, e.g., the generation of tolerance or immunity. Cytometry Part A 54A:19 –26, 2003. © 2003 Wiley-Liss, Inc. Key terms: magnetofluorescent liposomes; flow-cytometry; high-sensitivity immunofluorescence; antigen-presentation; T-cell activation
endogenous proteins bound to the MHC class II molecules. Therefore, the copy number of a particular peptide which is presented per APC is low (8). Consequently, T cells can sense as few as 100 –200 specific MHC/peptide complexes on the APC surface (8,9), which is approximately 10 times less than the detection limit of classical immunofluorescence (10,11). Therefore, the identification of distinct APC populations according to peptidepresentation at physiological doses of antigen by immunofluorescent techniques is still limited. We have previously described magnetofluorescent liposomes for the 100- to 1,000-fold enhancement of immunofluorescent signal intensity as compared to conventional techniques. This technology allows the detection of cells according to the expression of less than 100 molecules per cell (11,12). Here, we have used magnetofluorescent liposomes for the identification of APC according Contract grant sponsor: EC; Contract grant number: BIO-CT98-0458. *Correspondence to: Alexander Scheffold, Deutsches Rheuma-Forschungszentrum, Schumannstr. 21/22, 10117 Berlin, Germany. E-mail:
[email protected] Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/cyto.a.10055
20
KUNKEL ET AL.
to the presentation of a particular peptide. We show that the cytometric detection limit correlates well with the physiological threshold for T-cell activation and allows the identification and isolation of peptide-presenting APC in vitro and ex vivo. The high signal intensity allows clear-cut discrimination between positive and negative cells, a prerequisite for the further characterization of these cells. MATERIALS AND METHODS Mice Mice homozygously transgenic for the DO11.10 alpha/ beta-T-cell receptor (OVA-TCRtg/tg) (13) on BALB/c background (generous gift of Dennis Y. Loh, Washington University School of Medicine, St. Louis, MO) and BALB/c mice were bred and maintained under specific pathogenfree (SPF) conditions in our animal facility (BgVV, Berlin, Germany). Six- to eight-week-old male or female mice were used in all experiments. Antigens and Immunization A modified Ova-peptide323–339 (iSQAVHAAHAEINEAGRc) was synthesized and purified by J. Schneider-Mergener, Charite´, Berlin, Germany. Labeling of the peptide was performed at pH 9.5 with a 10-fold molar excess of fluoresceineisothiocyanate (FITC) (Sigma Chemical Co., Taufkirchen, Germany). Unconjugated FITC was removed from FITC-labeled Ova (FOva) by gel filtration on a Sephadex G10 column (Amersham Pharmacia, Freiburg, Germany) using phosphate-buffered saline (PBS) as the eluent. The final concentration of FOva was calculated as the initial amount of Ova-peptide minus an average loss of 30% due to the purification procedure. There was variability regarding physiological and staining activity between different FOva batches, which resulted in slight differences among the different experiments. However, for comparative analyses, identical batches of FOva were used. For in vivo experiments, FOva was given intravenously into the tail vein. Antibodies and Liposomes The following antibodies were used for flow-cytometric analysis: AN18 (anti-IFN␥)-FITC, N418 (anti-CD11c)-PE and -Cy5, Mac-1 (anti-CD11b)-PE, B220 (anti-CD45R)-PE (RA3-6B2), KJ1-26.1 (anti-OvaTCR)-Cy5, anti-CD4 (GK1.5)PE, M5/114 (anti-I-A)-PE, anti-FcR (2.4G2), and monoclonal anti-FITC-digoxigenin (Dig). In costimulation experiments, anti-CD28 (37.51) (BD Biosciences, Pharmingen, Heidelberg, Germany) was used. Rat immunoglobulin G was purchased from Biotrend, Ko ¨ ln, Germany. Magnetofluorescent liposomes (Cy5 or FITC) were generated and conjugated to sheep-anti-Digoxigenin (fab fragments) (Roche, Mannheim, Germany) as described earlier (12). Briefly, liposomes were prepared by the ‘extrusion method’, using 45% phophatidylcholine, 10% phosphatidylglycerol, 5% phosphatidylethanolamine, and 45% cholesterol (a gift from Lipoid KG, Ludwigshafen, Germany). The liposomes contained 10-mM carboxyfluorescein or palmitoylated-CY5 (lipid-to-dye ratio of 500:1) (Molecular Probes, Leiden, The Netherlands) and superparamagnetic
microparticles (diameter, 50 –100 nm; absorbance at 450 nm of 1,000-fold dilution ⫽ 0.5) (a gift from Miltenyi Biotec, Bergisch Gladbach, Germany). Small, nonmagnetic liposomes were excluded by filtration on magnetic cell separation columns (MACS), resulting in uniformly sized magnetofluorescent liposomes approximately 200 –300 nm in size. Antibodies were conjugated through their sulfhydryl groups, as obtained by reduction with 5-mM dithiothreitol, to the amino groups of the liposome’s phosphatidyl-ethanolamine, using SMCC (succinimidyl-4-(Nmaleimidomethyl)cyclohexane-1-carboxylate) (Pierce, Bonn, Germany). Propidium iodide and scatter gating were used to exclude dead cells. Cytometric analysis was performed on a FACSCalibur using CELLQuest research software ( BD Biosciences, Heidelberg, Germany ). Analysis of Peptide Presentation Cells were either incubated in vitro with Fova, or FOva was given i.v. in PBS; control cells and mice only received PBS. The in vitro-labeled cells were washed three times with PBS/bovine serum albumin (BSA). For ex vivo analysis, mice were sacrificed and cells from spleen and peripheral lymphnodes (LN) (inguinal, brachial, axilar) were isolated and washed twice with PBS/BSA. Unspecific binding of antibodies was blocked by preincubation with antiFcR (2.4G2) and rat IgG (both at 50 g/ml). Peptide presentation was analyzed by staining with anti-FITC-Dig followed by staining with anti-Dig-liposomes as previously described (12). Specificity of the staining was verified by staining of cells that were incubated with RPMI medium only, or by staining of cells from mice that were injected with PBS alone. APC subsets were identified by staining for B220 (B lymphocytes), CD11c (DC), and CD11bhigh CD11c– (macrophages), respectively. SORTING OF FOva PRESENTING CELLS Mouse spleen cells were incubated in vitro with FOva for 6 – 8 h. Cells were then washed twice and labeled with anti-FITC-Dig and anti-Dig-liposomes as described. For the separation of cells, the MidiMACS system (Miltenyi Biotech, Bergisch Gladbach, Germany) was used. The sorted cells were then either stained for MHC class II on the surface and analyzed by FACS or incubated with preactivated Ova-specific T cells for functional analysis. Analysis of T-cell Activation DC, B cells, and macrophages were purified by the MidiMACS system and subsequently incubated in vitro with FOva for 8 h. Cells were then washed twice. CD4⫹ Ova-TCR transgenic T cells were also sorted by MACS and labeled with CFDA (1 M in PBS, for 3 min followed by two washing steps with 10% FCS) and subsequently cultured with peptide loaded B-cells (ratio of 1:2), DC (ratio of 100:1), and macrophages (ratio of 10:1) for 72 h. For costimulation anti-CD28 (5g/ml) was added in some experiments. Proliferation of Ova-specific T cells was then analyzed by flow-cytometry. Alternatively, Ova-TCR transgenic spleen cells were preactivated with 1 g/ml Ovapeptide323–339 for six days and subsequently re-stimulated
CYTOMETRIC ANALYSIS OF PEPTIDE PRESENTATION
21
FIG. 1. Detection system for peptide presentation using magnetofluorescent liposomes. FOva-presenting cells are detected by a Dig-labeled anti-FITC antibody followed by anti-Dig-liposomes.
with FOva loaded APC for 6 h. To block cytokine secretion, brefeldin A was added for the last 3 h. Cells were then fixed with 2% para-formaldehyde and stained for Ova-specific TCR and intracellular IFN-␥. RESULTS Cytometric Detection of Peptide Presentation by Antigen-Presenting Cells T cells efficiently recognize less than 100 –200 of specific peptide/MHC class II complexes on the surface of an APC (8,9). To detect APC presenting such low numbers of a particular peptide, we used magnetofluorescent liposomes, which have a detection limit of less than 100 molecules per cell (11). The antigenic peptide derived from chicken ovalbumin was labeled with fluorescein using FITC. Fluorescein labeling did not alter the T-cell stimulatory capacity of the modified versus unmodified peptide (14, and data not shown). For detection of the fluorescein-labeled ovalbumin peptide (FOva) on the cell surface, an anti-FITC antibody, conjugated to digoxigenin (anti-FITC-Dig) was used followed by staining with antiDig-conjugated magnetofluorescent liposomes (anti-Dig liposomes) or PE-labeled anti-Dig antibodies (anti-Dig PE) (Fig. 1). To test the sensitivity of this detection system, murine spleen cells were incubated with various concentrations of FOva for 8 h in vitro and stained for FOva and MHC class II on the cell surface. A comparison of the stainings obtained with anti-Dig PE and anti-Dig liposomes is shown in Figure 2. At high doses of FOva (100 g/ml), conventional anti-Dig-PE staining resulted in an increase of the mean fluorescence intensity (MFI) of all MHC class II positive cells from 3 to 52.6. This increase of the MFI is not sufficient for optical separation of positive cells from the negative control. Only about 47% of the MHC class II⫹ cells reached the statistical threshold value defined according to the fluorescence intensity of control cells. In contrast, anti-Dig liposome staining brightly labeled all cells within the MHC class II⫹ population at a five-fold lower peptide dose of 20 g/ml, resulting in a MFI of 521.5 versus a MFI of 40.2 of unstained or MHC class II– control cells. This signal intensity allows the clear-cut
FIG. 2. Flow-cytometric detection of FOva presentation by spleen cells. Balb/c splenocytes were incubated in vitro with the indicated concentrations of FOva for 6 – 8 h. Cells were then washed and stained on the surface for presentation of FOva with anti-Dig-PE (conventional staining; A–D) or anti-Dig-Liposomes (liposome staining; E–H). Cells were counterstained for MHC class II (FL4). The indicated frequencies are given as the percentage of FOva⫹ cells within the MHC class II⫹ or MHC class II– population, respectively. The mean fluorescence intensities (MFI) are shown for all MHCII⫹ cells or MHCII– cells (H), respectively. Uptake of the labeled peptide can also be observed in FL1. Data are representative for three similar experiments.
discrimination of positive and negative populations. At 2 and 0.2 g/ml of peptide, representing concentrations which are routinely used for in vitro stimulation of OvaTCR transgenic T cells, no significant anti-Dig-PE staining was detected (MFI 3.4 and 2.9). Staining with anti-Dig liposomes resulted in labeling of 55% and 7% of MHC class II⫹ cells. At peptide concentrations of 2 and 20 g/ml, a shift in FL1 MFI of all cells was observed which was independent of anti-Dig liposome staining (data not shown), indicating peptide uptake by MHC class II positive and negative cells. However, MHC class II– cells, which were used as an internal specificity control, did not show significant peptide staining on the surface. The Sensitivity of the Cytometric Detection of Peptide Presentation Correlates With the Threshold of T-Cell Activation To directly compare the sensitivity of the liposome staining with the threshold of T-cell activation, we analyzed in parallel the FOva presentation of purified APC populations following in vitro incubation at various concentrations of FOva by flow-cytometry as well as by stimulation of Ovaspecific T cells. DC, B cells, and macrophages were purified by magnetic cell sorting, incubated with FOva at concentrations ranging from 0.01–1 g/ml for 12 h at 37°C, and washed and co-cultured for 72 h with CFDA-labeled Ova-TCR transgenic T cells. For cytometric analysis, spleen cells were incubated with FOva as described above and stained for presented peptide using anti-Dig liposomes. For the analysis
22
KUNKEL ET AL.
cells, 1:10 for macrophages, and 1:100 for DC, reflecting the APC composition in the spleen. As shown in Figure 3, the same strong T-cell stimulatory capacity was observed when DC and B cells were used as APC, although the number of B cells is 200 times higher than the number of DC. The frequency of proliferating T cells decreased in a peptide-dosedependent fashion from 80%–90% at 1 g/ml of peptide to ⬍ 5% at 0.01 g/ml. The average number of cell divisions was also decreased in a dose-dependent manner. Macrophages induced only about 5%–20% of T cells to proliferate at peptide doses from 0.1–10.01 g/ml, and no proliferation was observed at concentrations below 0.1 g/ml. Since resting B cells do not provide costimulatory signals we added an antiCD28 costimulatory antibody to some of the cultures where B cells were used as APC. This costimulation increased the frequency of proliferating T cells, especially at lower peptide concentrations, and 12%–30% of proliferating T cells were still observed at 0.01 g/ml. In contrast, only a minimal 兾) effect of anti-CD28 was observed when macrophages (MO were used as APC (Table 1). Staining with liposomes resulted in a significant population of peptide presenting DC at concentrations ranging from 1 g/ml (60%) to 0.03 g/ml (4%). No significant staining was detected below peptide concentrations of 0.03 g/ml, when compared to control DC without peptide incubation. Peptide presentation by macrophages and B cells was considerably lower, ranging from about 40% for both populations at 1 g/ml to 13% for macrophages and 8% for B cells at 0.1 g/ml. At 0.03 g/ml, peptide presentation was only detectable on macrophages (4%) but not on B cells. No peptide presentation was detectable below 0.03 g/ml. Conventional staining with antiDig-PE was not different from background staining at all peptide concentrations (data not shown). These data indicate that the sensitivity of the cytometric detection of peptide presentation by liposome staining reaches the same order of magnitude as the threshold of T-cell activation. We conclude that liposome staining can detect peptide presentation at 100- to 1,000-fold lower levels (100 g/ml versus ⬍ 0.1 g/ml) when compared to conventional reagents.
FIG. 3. The detection limit of liposome staining reaches the threshold of T-cell activation. A) Balb/c splenocytes were incubated in vitro with the indicated amounts of FOva for 8 h. Cells were then washed and 兾 ), CD11c stained on the surface for B220 (B lymphocytes), CD11b (MO (DC), and presentation of FOva (anti-FITC-Dig, anti-Dig-liposomes), and analyzed by FACS. Numbers indicate the percentage of FOva-presenting cells within the various APC subsets. All percentages are corrected for background staining obtained on control cells. B) Splenic B cells, DC, and 兾 were purified by MACS separation and cultured for 8 h with FOva at MO the indicated concentrations, washed, and cultured with CFDA-labeled Ova-TCR transgenic T cells for 72 h. For the determination of proliferation, the cells were stained for the Ova-specific TCR and CD4 and analyzed by flow-cytometry. Numbers indicate the percentage of proliferating cells within the KJ1.26 population.
FOva Positive APC Stimulate Ova-specific T Cells In Vitro
of peptide presentation by B lymphocytes, DC, or macrophages, the cells were costained in addition with antibodies against B220 or CD11c and CD11b, respectively. For the co-culture experiment, the APC-T cell ratios were 2:1 for B
To confirm that the cells detected as FOva-presenting cells by the liposome technology are indeed capable of presenting peptide to Ova-specific T cells, Balb/c splenocytes were incubated with FOva at a concentration of 0.2
Table 1 Influence of Co-stimulation on T Cell Proliferation 0 g/ml ⫺CD28 ⫹CD28 B cells MA
1,2 1
1 1
0,01 g/ml ⫺CD28 ⫹CD28 3,5 1,2
32 1
0,03 g/ml ⫺CD28 ⫹CD28 15 1,4
55 2,5
0,1 g/ml ⫺CD28 ⫹CD28 58 3,8
85 4,8
0,3 g/ml ⫺CD28 ⫹CD28 84 8
90 12
1 g/ml ⫺CD28 ⫹CD28 91 16
92 28
Splenic B cells and macrophages were purified by MACS separation and cultured for 8 h with FOva at the indicated concentrations, washed and cultured with CFDA-labeled Ova-TCR transgenic T cells for 72 h ⫾ anti-CD28. For the determination of proliferation the cells were stained for the Ova-specific TCR and CD4 and analyzed by flow-cytometry. Numbers indicate the percentage of proliferating cells within the KJ 1.26 population.
CYTOMETRIC ANALYSIS OF PEPTIDE PRESENTATION
23
In contrast, the FOva-enriched population induced IFN-␥ production by 20% of the Ova-specific T cells, indicating that we have indeed identified and enriched a subpopulation of cells which present the FOva associated with MHC class II on the cell surface. All together, these results show that the detection method presented here is capable of detecting and isolating those cells within a heterogenous population of cells, which present MHC class IIassociated peptides at levels above the threshold for T-cell activation. In Vivo Presentation of Intravenously Applied FOva
FIG. 4. FOva sorted cells present peptide to specific T cells. Balb/c splenocytes were incubated in vitro with 0.2 g/ml FOva for 6 h. Cells were washed and labeled on the surface with anti-FITC-Dig and anti-DigLiposomes and separated by MACS. A) sorted fractions were stained for MHC class II and analyzed by FACS. B) sorted fractions were incubated with preactivated Ova-specific T cells for 6 h (addition of Brefeldin A after 3 h), fixed, and stained for intracellular IFN-␥ and Ova-specific TCR. Numbers indicate the percentage of FOva-presenting cells within the MHC class II⫹ population (A) or the percentage of IFN␥-producing Ovaspecific CD4⫹ T cells (B), respectively. The results are representative of two separate experiments.
g/ml for 6 h, and FOva-presenting cells were stained with anti-Dig liposomes as described. The few FOva-positive cells within the MHC class II⫹ population were magnetically enriched from 2.8% to 35% using MACS technology (Fig. 4A). The FOva depleted fraction still contained 0.6% weakly labeled peptide-positive cells. The original, enriched, and depleted spleen cell populations were cocultured for 6 h with Ova-TCR transgenic T cells which had been in vitro preactivated with Ova peptide for one week. Following restimulation, the frequency of IFN-␥ producing T cells was determined by intracellular staining (Fig. 4B). The original population activated about 3% of the transgenic T cells to produce IFN-␥, and the frequency of IFN-␥ producing cells was reduced to 1.7% when cells depleted for FOva⫹ cells were used for T cell stimulation.
To assess the potential of liposome staining for the identification of APC ex vivo, we injected FOva intravenously into Balb/c mice. At various timepoints after the injection of 0.5-mg FOva, the cells from the spleen were stained for FOva by liposomes and costained for B220 or CD11c to analyze the frequency of peptide-presenting cells within the B lymphocyte and DC population. As shown in Figure 5, intravenously applied FOva was rapidly presented by MHC class II expressing cells within 1 h, with 35.8 % of the B lymphocytes and 47.3% of DC being positive for FOva, followed by a rapid disappearance of peptide-presenting cells within 12–24 h. Conventional staining did not result in any significant staining compared to control cells (data not shown). To determine whether in vivo DC and B cells show the same differences observed in vitro regarding the capacity to present peptides, we intravenously applied decreasing doses of peptide (0.5– 0.05 mg) and stained for peptide presentation on DC, B cells as described above and, in addition, on macrophages (CD11c– CD11b⫹). As shown in Figure 6, at all antigen doses DC had a higher capacity to present peptides com兾 ), ranging from 15% pared to B cells and macrophages (MO 兾 ) at 0.5 mg to 2.5% (DC) versus 8.5% (B cells) and 5% (MO 兾 ) at 0.05 mg. These data (DC) versus 0% (B cells and MO show that magnetofluorescent liposomes allow APC identification ex vivo, which present specific peptides at least 10-fold below the level of detection of conventional immunofluorescence. Furthermore, this shows that independent of the antigen dose, DC have the highest capacity to present soluble peptides in vivo (Fig. 6). DISCUSSION Presentation of specific peptide/MHC complexes by APC is a decisive step for the generation of adaptive immune responses against pathogens as well as for the induction of tolerogenic mechanisms. We describe here a cytometric technique for the in vitro or ex vivo identification of those APC which present a certain antigenic peptide at physiologically relevant levels, i.e., at levels above the threshold required for the activation of specific T cells. We have used magnetofluorescent liposomes for the detection and isolation of APC on the single-cell level according to the presentation of a specific hapten-labeled peptide. It has previously been shown that fluorochromeor hapten-labeled peptides can be used for the immunofluorescent detection of specific peptide/MHC complexes
24
KUNKEL ET AL.
FIG. 5. Ex vivo analysis of FOva-presentation. Balb/c mice were injected with 500 g FOva i.v. into the tail vein. Presentation of FOva in the spleen was analyzed at the timepoints indicated on B220⫹ B lymphocytes (A), CD11c⫹ dendritic cells (B), or MHC class II– cells (C), respectively. Spleen cells were stained for B220, CD11c, or MHC class II, and for FOva on the cell surface using anti-FITC-Dig/anti-Dig-Liposomes. The indicated frequencies are given as the percentage of FOva⫹ cells within the APC subsets or the MHC class II– population, respectively. The data are representative of two independent experiments with three mice in each experimental group.
FIG. 6. Titration of the antigen dose in vivo. Balb/c mice were injected with the indicated amount of FOva i.v. into the tail vein. Presentation of FOva in the spleen was analyzed after 3 h on CD11c⫹ dendritic cells, B220⫹ B lymphocytes, or CD11bhigh/CD11c– macrophages. Spleen cells were stained for B220, CD11c, CD11b, or MHC class II, and for FOva on the cell surface using anti-FITC-Dig/anti-Dig-liposomes. The indicated frequencies are given as the percentage of FOva⫹ cells within the APC subsets.
on the APC surface, following in vitro antigen loading (14,15). However, the cytometric detection of APC according to the presentation of physiologically low numbers of MHC/peptide complexes was not possible due to the limited sensitivity of conventional immunofluorescent staining techniques. As demonstrated here, magnetofluorescent liposomes allowed the detection and magnetic enrichment of peptide-presenting APC at 100- to 1,000fold lower peptide concentrations compared to conventional PE-labeled antibodies. In the presence of costimulation via CD28 antibody, both the cytometric and the functional assay showed comparable sensitivity, although the threshold for T-cell activation (0.01 g/ml) was slightly below the detection level of the cytometric detection (0.03 g/ml). However, at such low peptide concentrations, T-cell stimulation was inefficient, since only 10%– 30% of T cells proliferated, and the number of cell cycles per cell was also reduced compared to higher peptide concentrations. Therefore, we conclude that the high sensitivity of cytometric detection using liposomes allows the detection of peptide presentation at all levels required for efficient T-cell activation. We have previously shown that magnetofluorescent liposomes can detect less than 100 molecules per cell (11), and this value correlates well with the estimated threshold for T-cell activation (8,9).
CYTOMETRIC ANALYSIS OF PEPTIDE PRESENTATION
Therefore, we conclude that the method presented here can detect less than 100 peptides per APC. Importantly, the high signal intensity enables the clear-cut discrimination of positive and negative cells for the identification and functional characterization of peptide-presenting APC subpopulations. Different batches of FOva occasionally resulted in different frequencies of peptide-presenting cells, probably due to variations in the peptide concentration and the efficiency of labeling with FITC. However, the described differences in peptide presentation between the three APC populations as well as the correlation of peptide staining with the functional activity of the APC were the same with all FOva preparations, although at different absolute levels. DC showed the highest capacity to present soluble peptides at all peptide concentrations, both in vitro and in vivo. This has also been reported earlier for protein antigens (16 –18). One reason for enhanced presentation by DC at low peptide levels could be the higher absolute number of MHC class II molecules on the surface. It has been shown previously that following i.v. injection of protein, B cells and DC do not differ in the amount of specific MHC class II/peptide complexes when normalized to the overall MHC class II expression (16,18). In contrast to the in vitro peptide incubation, macrophages show reduced frequencies of peptide-presenting cells compared to B cells following in vivo peptide application. This difference between the in vivo and in vitro presentation may be explained by the fact that splenic macrophages express low level of MHC class II in vivo but become activated during the isolation procedure and the in vitro culture and up-regulate MHC class II and antigen presentation. On a per-cell basis, DC were also the most potent inducers of T-cell proliferation, since they induced the same or a five-times-higher frequency of proliferating T cells compared to B cells or macrophages, though at a 200-fold (for B cells) or 20-fold (for macrophages) lower cell number. B cells and macrophages had comparable levels of peptide presentation following in vitro incubation. Taking into consideration the 20-fold lower number of macrophages compared to B cells used for the functional assays, their capacity to activate T cells is probably similar or even higher than that of B cells. This is also supported by the observation that the addition of costimulation only increased T-cell stimulatory activity of B cells but not macrophages, indicating that macrophages already provide sufficient costimulation to T cells. Intravenous injection of soluble antigen usually induces antigen-specific tolerance. Our data show that DC are most efficient in presenting soluble peptides in vitro and in vivo and most efficiently activate T cells in vitro. In contrast, the large number of B cells also presenting peptides have very low stimulatory capacity. Since it is known that the state of tolerance which is induced following soluble antigen application is preceded by a transient stage of activation and proliferation of specific T cells (19), our data suggest that DC are mainly responsible for
25
tolerance induction to intravenously applied antigens. In fact, it has been shown that DC under steady-state conditions can induce antigen-specific tolerance (20). Our method detects peptide-presentation independent of antigen processing. As an alternative to labeled peptides, antibodies have been developed which recognize a specific MHC class II/peptide complex (3– 6). These antibodies allow analysis of the presentation of processing dependent antigens, for example whole proteins, and they could be used in combination with magnetofluorescent liposomes for sensitive analysis. However, MHC/ peptide-specific antibodies tend to show intrinsic crossreactivity to MHC molecules of the same haplotype loaded with many different endogenous peptides (4,7). Therefore, to successfully combine high-sensitivity staining using magnetofluorescent liposomes with MHC/peptide-specific antibodies, the antibodies have to be preselected according to high specificity. In summary, we have shown that magnetofluorescent liposomes can be used to specifically stain and isolate APC according to the presentation of less than 100 peptides. This method for the first time allows the unequivocal identification and the characterization of APC subpopulations, which present antigenic peptides at levels above the threshold for T-cell activation in vitro and in vivo. The ex vivo identification of APC involved in immunogenic or tolerogenic immunization protocols will be a valuable tool in the identification of the cellular and molecular factors required for the induction of these different types of immune responses. ACKNOWLEDGMENTS We thank Andreas Thiel, Farah Hatam, and Osman So ¨ zeri for critical reading of the manuscript. LITERATURE CITED 1. Fuchs EJ, Matzinger P. B cells turn off virgin but not memory T cells. Science 1992;258:1156 –1159. 2. Steinman RM. The dendritic cell system and its role in immunogenicity. Annu Rev Immunol 1991;9:271–296. 3. Rudensky A, Rath S, Preston-Hurlburt P, Murphy DB, Janeway Jr CA. On the complexity of self. Nature 1991;353:660 – 662. 4. Zhong G, Reis e Sousa C, Germain RN. Production, specificity, and functionality of monoclonal antibodies to specific peptide-major histocompatibility complex class II complexes formed by processing of exogenous protein. Proc Natl Acad Sci USA 1997;94:13856 –13861. 5. Dadaglio G, Nelson CA, Deck MB, Petzold SJ, Unanue ER. Characterization and quantitation of peptide-MHC complexes produced from hen egg lysozyme using a monoclonal antibody. Immunity 1997;6: 727–738. 6. Krogsgaard M, Wucherpfennig KW, Canella B, Hansen BE, Svejgaard A, Pyrdol J, Ditzel H, Raine C, Engberg J, Frugger L. Visualization of myelin basic protein (MBP) T cell epitopes in multiple sclerosis lesions using a monoclonal antibody specific for the human histocompatibility leukocyte antigen (HLA)-DR2-MBP 85-99 complex. J Exp Med 2000;191:1395–1412. 7. Porgador A, Yewdell JW, Deng Y, Bennink JR, Germain RN. Localization, quantitation, and in situ detection of specific peptide-MHC class I complexes using a monoclonal antibody. Immunity 1997;6:715– 726. 8. Harding CV, Unanue ER. Quantitation of antigen-presenting cell MHC class II/peptide complexes necessary for T-cell stimulation. Nature 1990;346:574 –576. 9. Demotz S, Grey HM, Sette A. The minimal number of class II MHCantigen complexes needed for T cell activation. Science 1990;249: 1028 –1030.
26
KUNKEL ET AL.
10. Shapiro HM. Practical flow cytometry. New York: Alan Liss, Inc.; 1988. p. 84 – 85. 11. Scheffold A, Assenmacher M, Reiners-Schramm L, Lauster R, Radbruch A. High-sensitivity immunofluorescence for detection of the pro- and anti- inflammatory cytokines gamma interferon and interleukin-10 on the surface of cytokine-secreting cells. Nat Med 2000;6: 107–110. 12. Scheffold A, Miltenyi S, Radbruch A. Magnetofluorescent liposomes for increased sensitivity of immunofluorescence. Immunotechnology 1995;1:127–137. 13. Murphy KM, Heimberger AB, Loh DY. Induction by antigen of intrathymic apoptosis of CD4⫹CD8⫹TCRlo thymocytes in vivo. Science 1990;250:1720 –1723. 14. Gariepy J, Watts TH. Probing the accessibility of peptide in an MHC class II protein — foreign peptide complex. Int Immunol 1991;3: 279 –284. 15. Falk K, Lau JM, Santambrogio L, Esteban VM, Puentes F, Rotzschke O, Strominger JL. Ligand exchange of major histocompatibility complex class II proteins is triggered by H-bond donor groups of small molecules. J Biol Chem 2002;277:2709 –2715.
16. Zhong G, Sousa CR, Germain RN. Antigen-unspecific B cells and lymphoid dendritic cells both show extensive surface expression of processed antigen-major histocompatibility complex class II complexes after soluble protein exposure in vivo or in vitro. J Exp Med 1997;186:673– 682. 17. Reis e Sousa C, Germain RN. Analysis of adjuvant function by direct visualization of antigen presentation in vivo: endotoxin promotes accumulation of antigen-bearing dendritic cells in the T cell areas of lymphoid tissue. J Immunol 1999;162:6552– 6561. 18. Manickasingham S, Reis e Sousa C. Microbial and T cell-derived stimuli regulate antigen presentation by dendritic cells in vivo. J Immunol 2000;165:5027–5034. 19. Kearney ER, Pape KA, Loh DY, Jenkins MK. Visualization of peptidespecific T cell immunity and peripheral tolerance induction in vivo. Immunity 1994;1:327–339. 20. Hawiger D, Inaba K, Dorsett Y, Guo M, Mahnke K, Rivera M, Ravetch JV, Steinman RM, Nussenzweig MC. Dendritic cells induce peripheral tolerance under steady state conditions in vivo. J Exp Med 2001;194: 769 –779.
