Dendritic Cell Aggresome-Like-Induced Structure Formation and Delayed Antigen Presentation Coincide in Influenza Virus-Infected Dendritic Ce
http://www.100md.com
免疫学杂志 2005年第14期
Abstract
Influenza virus infection induces maturation of murine dendritic cells (DCs), which is most important for the initiation of an immune response. However, in contrast to EL-4 and MC57 cells, DCs present viral CTL epitopes with a delay of up to 10 h. This delay in Ag presentation coincides with the up-regulation of MHC class I molecules as well as costimulatory molecules on the cell surface and the accumulation of newly synthesized ubiquitinated proteins in large cytosolic structures, called DC aggresome-like-induced structures (DALIS). These structures were observed previously after LPS-induced maturation of DCs, and it was speculated that they play a role in the regulation of MHC class I Ag presentation. Our findings provide the first evidence for a connection between DC maturation, MHC class I-restricted Ag presentation, and DALIS formation, which is further supported by the observation that DALIS contain ubiquitinated influenza nucleoprotein.
Introduction
Dendritic cells (DCs)3 play a central role in the induction of primary adaptive immune responses by providing unspecific and specific stimuli for T cells (1). These abilities require the development from an immature into a mature state of differentiation and the presentation of a diverse array of Ags (2). The former is accompanied by the increase in the expression of surface MHC and costimulatory molecules and the secretion of proinflammatory cytokines (3). The latter requires efficient Ag uptake and processing that involves proteasomal activity for most Ags presented by MHC class I molecules (4). For the activation of naive T cells, these two independent processes should take place in a coordinated fashion. In nonprofessional APCs for example, MHC class I-restricted peptide presentation after influenza virus infection reaches its maximum already after 2–4 h (5). In professional APCs, MHC class I-restricted Ag presentation should be slowed down to allow it to take place together with maximum expression of costimulatory molecules and secretion of proinflammatory cytokines that require 12–24 h for the former (6, 7, 8) and 6–12 h for the latter (9). Recently, a new feature of DCs was discovered that could represent a mechanism to enable coordinated maturation and Ag presentation in this cell type. After LPS-induced maturation, DCs transiently accumulate newly synthesized ubiquitinated proteins in an aggregated form in the cytosol (10, 11). These protein aggregates are called DC aggresome-like-induced structures (DALIS) because they were observed initially in DCs and resemble aggresomes. Both DALIS and aggresomes contain ubiquitinated proteins, but in contrast to DALIS, aggresomes result from microtubule-dependent conglomeration of smaller aggregates and redistribution of the intermediate filament protein vimentin (12). Furthermore, DALIS formation is not due to impaired proteolysis and does not affect the ubiquitin-proteasome pathway. Ubiquitin is a conserved 8.5 kDa protein that is involved in many cellular functions, including gene expression, ribosome biosynthesis, receptor expression, and stress response. Ubiquitin is covalently attached to proteins that bind to and are degraded by 26S proteasomes (13, 14). This process results in the generation of peptides from which ligands presented by class I molecules are selected.
It has been shown that protein neosynthesis is necessary for the presentation of a T cell epitope from a long-lived viral protein that is a property anticipated by the defective ribosomal product (DRiP) hypothesis (15). It was speculated that most of the peptide ligands for MHC class I molecules are made from DRiPs (16), which are newly synthesized polypeptides that never attain native structure (17, 18, 19). Normally, DRiPs are degraded rapidly by proteasomes (20). However, in DCs, DRiPs are stored in DALIS, which delay their degradation and stabilize them for 8–16 h (11). Recently, it was shown that DALIS formation can occur not only in DCs but also in stimulated macrophages with a similar time course (21).
By delaying degradation of newly synthesized misfolded proteins, maturing DCs might have the ability to delay MHC class I loading and peptide presentation until the secretion of proinflammatory cytokines and the expression of costimulatory molecules has been initiated. This feature would prevent the presentation of T cell epitopes by immature DCs and therefore avoid the induction of tolerance.
In this study, we show that activation of bone marrow-derived DCs (BMDCs) by influenza virus infection results indeed in a delayed processing and presentation of viral Ags to CTLs compared with Ag presentation in nonprofessional APCs. This delay in Ag presentation coincides with DALIS formation during BMDC maturation. Furthermore, we show that DALIS contain ubiquitinated influenza nucleoprotein (NP). Thus, our results demonstrate that MHC class I-restricted Ag presentation and formation of DALIS appear to be tightly linked in DCs.
Materials and Methods
Generation of BMDCs
Mouse immature BMDCs were generated from bone marrow of C3H/HeN, C3H/HeJ/TLR2–/– (TLR2/4–/–), and C57BL/6 mice. For the generation of mouse BMDCs, IMDM (BioWhittaker) supplemented with 2 mM L-glutamine (Seromed), 100 U/ml penicillin/streptomycin (BioWhittaker), 5% FCS (Vitromex), and 200 U/ml GM-CSF (PeproTech) was used. Bone marrow cells were incubated in GM-CSF-containing medium for 6 or 7 days, with full medium replacement on day 3 (removal of nonadherent cells) and feeding on day 5. On day 6 or 7, nonadherent and semiadherent cells were harvested. These cells were typically CD14– and >80% CD11c+ as determined by FACS analysis. Furthermore, the CD11c+ cells were CD40low, CD80low, CD86low, and H-2 Kb medium.
Generation of NP366–374-specific T cell line
A C57BL/6 mouse (8 wk of age) was immunized i.p. with 500 hemagglutinating units of human influenza virus A (PR/8/34) (obtained from Dr. S. Pleschka, Institute for Virology, University of Giessen, Giessen, Germany). After 9 days, the mouse was killed, and the splenocytes were stimulated with the H2-Db-restricted immunodominant epitope from the influenza nucleoprotein (NP366–374). Cells were split and restimulated in weekly intervals using modified Eagle’s medium (-modification) supplemented with 10% (v/v) heat-inactivated FCS, 2 mM L-glutamine, antibiotics, and 50 μM 2-ME. Specificity and reactivity were tested routinely by 51Cr release assay.
IFN- and 51Cr release assay
BMDCs and EL-4 cells were stimulated at different time points with 1 PFU/cell human influenza virus A (PR/8/34). Half of the activated cells were incubated with 50 ng/ml NP366–374 peptide for 30 min before washing and fixation with 0.05% glutaraldehyde. Fifty thousand targets per well were incubated overnight with 100,000 T cells specific for the immunodominant viral NP366–374 epitope. After 20 h, supernatants were collected and IFN- was measured using sandwich ELISA (see below). For 51Cr release assays, BMDCs, EL-4 cells, and MC57 cells were infected with 1 PFU/cell PR/8/34 at 37°C. Thereafter, cells were counted and washed twice with medium before labeling with [51Cr]Na2CrO4 plus 5% FCS for 20 min. To control the general susceptibility to CTL lysis, half of the cells were incubated with 500 pmol of ASNENMETM during radioactive labeling. After intense washing, 20,000 targets were used in a 4-h release assay with 80,000 or 400,000 effector cells from a NP366–374-specific T cell line.
To determine the influence of protein neosynthesis on NP peptide presentation, BMDCs were treated with or without 40 μM cycloheximide (CHX; Sigma-Aldrich) during 6 h of PR/8/34 infection or cells were incubated with UV-treated influenza virus. For UV treatment, PR/8/34 was exposed to short-wave UV light (254 nm) for 5 min.
To test the viability of the CHX-treated cells, BMDCs were incubated with 500 pmol of ASNENMETM during radioactive labeling.
Cytokine ELISAs and FACS staining
BMDCs and EL-4 cells were activated with 1 PFU/cell PR/8/34, 50 μg/ml poly(I)poly(C) (poly(I:C); Amersham Biosciences), or 1 μg/ml LPS (Salmonella typhimurium; Sigma-Aldrich). After 18 h, supernatants were collected, and IL-6 was measured using standard sandwich ELISA protocol. Briefly, the capture Ab was coated on an ELISA plate (MaxiSorb; Nunc) and the biotinylated detection Ab was revealed by streptavidin-conjugated HRP and tetramethylbenzidine substrate (Sigma-Aldrich). The assay was read at 450 nm after adding 2 M H2SO4 to stop the reaction.
For FACS staining, BMDCs and EL-4 cells were activated at different time points by adding 1 PFU/cell PR/8/34. Cells were stained either with FITC-conjugated anti-CD80 and PE-conjugated anti-CD86 Abs or biotinylated anti-H-2 Kb/H-2 Db and streptavidin-FITC. All Abs and the cytokine standard were obtained from BD Biosciences. All analyses were performed using FACSCalibur and CellQuest software (BD Biosciences).
Immunoprecipitation
For influenza NP immunoprecipitation (IP), 3 x 106 BMDCs were treated with 1 PFU/cell PR/8/34 or with 1 μg/ml LPS for 11 h and washed two times with PBS before cell lysis was performed in PBS supplemented with Complete protease inhibitors (Roche Applied Science). After five freeze-thaw cycles, lysed cells were centrifuged at 12,000 rpm for 5 min before 8 μg of anti-influenza Ab (clone N3F3; gift from Dr. O. Planz, Institute for Immunology, Bundesforschungsanstalt für Viruserkrankungen der Tiere, Tübingen, Germany) was added to the supernatants. After a 2-h rotation, 100 μl of protein A-Sepharose were added for 1 h. Afterward, the Sepharose was washed three times with PBS before it was resuspended in SDS loading buffer and separated by SDS-PAGE. Then Western blots using anti-influenza Ab (N1A4; gift from Dr. O. Planz) or FK2 Ab as primary Abs were performed as described below.
For FK2 IP, 10 x 106 BMDCs were treated as described above. After 11 h of incubation, cells were washed two times with PBS before cell lysis with Triton lysis buffer (20 mM Tris-HCl (pH 7.4), 1% Triton X-100, 1.5 mM MgCl2, and 10 mM NaCl) at 4°C for 20 min was performed. After centrifugation at 12,000 x g for 10 min, pellets were resuspended in PBS with Complete protease inhibitors (Roche Applied Science) before adding 5 μg of biotinylated FK2 Ab (FK2 Ab was biotinylated with D-biotinoyl--aminocaprons?ure-N-hydroxysuccinimidester (Boehringer Mannheim) according to the manufacturer’s instructions). After 1.5 h at 4°C, 10 μl of MACS streptavidin microbeads (Miltenyi Biotec) and the IPs were performed following the manufacturer’s instructions. Afterward, pellets were resuspended in SDS loading buffer and separated by SDS-PAGE, followed by anti-influenza NP Western blotting as described below.
Western blotting
BMDCs, EL-4 cells, and MC57 cells were infected with 1 PFU/cell PR/8/34 at different time points or activated with LPS (1 μg/ml). Cells were harvested and washed three times with PBS before they were treated with lysis buffer (20 mM Tris-HCl (pH 7.4), 1% Triton X-100, 1.5 mM MgCl2, and 10 mM NaCl) at 4 °C for 20 min. After centrifugation at 12,000 x g for 10 min, supernatants and pellets were resuspended separately in SDS loading buffer and separated by SDS-PAGE. After electrophoresis, proteins were transferred to a Hybond ECL membrane (Pharmacia Biotech) by semidry Western blotting. The membrane was blocked for 1 h with 50 mM Tris, 150 mM NaCl, 0.1% Nonidet P-40, 3% BSA, and 5 mM EDTA and then incubated with rabbit anti-influenza NP Abs (gift from Dr. J. Yewdell, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD) or FK2 Ab (Affiniti/Biomol), followed by HRP anti-rabbit IgG (Amersham Biosciences) or HRP anti-mouse IgG (Jackson Immunoresearch/Dianova). For actin Western blotting, anti-actin (C2)-HRP (Santa Cruz Biotechnology) was used. To analyze the influence of CHX on influenza NP expression, cells were treated simultaneously with or without 100 μM CHX (Sigma) and 1 PFU/cell PR/8/34 for 6 h at 37°C before intense washing and counting. Afterward, cell pellets were resuspended in SDS loading buffer and separated by SDS-PAGE. The proteasome inhibitor epoxomicin (Biomol) was added 10 min before infection of BMDCs and EL-4 cells with the influenza virus at a final concentration of 1 μM. All Western blots were developed using ECL substrate (Pierce/Perbio) and analyzed by the ChemiDoc XRS System and Quantity One software (Bio-Rad).
Immunofluorescence staining and Abs
BMDCs, EL-4 cells, or MC57 cells were treated with 1 PFU/cell human influenza virus A PR/8/34, 50 μg/ml poly(I:C) (Amersham Biosciences), or 1 μg/ml LPS (Sigma-Aldrich) at different times. Activated cells were washed in PBS, 0.5% BSA (Roth), and 0.1% Saponin (Sigma-Aldrich) and fixed with Cytofix/Cytoperm (BD Pharmingen) for 20 min at 4°C. For DALIS staining, biotinylated FK2 Ab (Affiniti) and streptavidin-FITC (Sigma-Aldrich) were used as secondary reagents. Influenza NP staining was done with polyclonal anti-influenza NP serum (generously provided by Dr. J. Yewdell) and Alexa 546- or Alexa 594-conjugated anti-rabbit IgG (Molecular Probes). All Abs were diluted in washing buffer, and cells were stained 30 min at 4°C. In some cases, Ab incubation was followed by 4',6-diamidino-2-phenylindole dihydrochloride (Molecular Probes) staining for 10 min at room temperature. For colocalization experiments, BMDCs were harvested after 12 h of influenza infection and coated on 1% alcian blue (Sigma-Aldrich)-treated coverslips for 10 min at 37°C. Cells were permeabilized with 1% Triton X-100 in PBS for 5 min at 4°C and fixed with 3% paraformaldehyde in PBS for 10 min at room temperature before staining with FK2 and anti-influenza NP serum (provided by Dr. J. Yewdell) or FK2 and anti-20S proteasome Ab (BioTrend).
A Zeiss LSM 510 laser-scanning microscope or the Olympus Cell System was used.
Results
Influenza virus infection of BMDCs results in delayed Ag presentation
To analyze the kinetics of Ag presentation, we used BMDCs, EL-4 cells, and MC57 cells that were infected with human influenza virus A PR/8/34 at different time points as target cells for T cells specific for the immunodominant viral NP366–374 epitope. After 20 h of coincubation of T cells with fixed target cells, supernatants were analyzed by standard ELISA for IFN- (Fig. 1A). Whereas EL-4 cells induced IFN- production as early as 4 h after influenza infection, we found a delay in BMDC-induced IFN- secretion. In this study, IFN- production was first detectable after 12 h of infection. 51Cr release assays confirmed this observation. Influenza infection of the target cells for 4 h before adding them to the T cells was sufficient to induce maximum lysis of EL-4 cells. Similar kinetics were obtained for MC57 cells infected with the influenza virus. In contrast, specific killing of BMDCs increased until 10 h of virus infection (Fig. 1B). Target cells infected with PR/8/34 for >10 h showed spontaneous 51Cr release levels above 50% and were excluded from the experiments.
PR/8/34 induces BMDC maturation
Recently, it has been reported that during LPS-induced maturation, DCs accumulate newly synthesized ubiquitinated proteins in large cytosolic structures called DALIS (10). It was speculated that these structures play a role in the regulation of MHC class I Ag presentation. To test the ability of human influenza virus PR/8/34 to induce BMDC maturation, we infected BMDCs with 1 PFU/cell PR/8/34 or activated them with 1 μg/ml LPS or 50 μg/ml poly(I:C) and compared production of the proinflammatory cytokine IL-6 and up-regulation of the costimulatory molecules CD80 and CD86. All stimuli tested induced IL-6 secretion (Fig. 3A) and up-regulation of CD80 and CD86 (Fig. 3B). Poly(I:C)- and LPS-activated BMDCs showed similar and higher IL-6 production than PR/8/34-infected cells. In contrast, PR/8/34 infection induced a higher up-regulation of CD80 and CD86 molecules than LPS. Accumulation of ubiquitinated proteins during BMDC maturation was analyzed by confocal microscopy using the FK2 mAb, which is specific for mono- and poly-ubiquitinated proteins. The Ab does not react with free ubiquitin. FK2-positive accumulations of proteins could be observed in PR/8/34-infected BMDCs as well as in LPS-activated cells (Fig. 3C) and poly(I:C)-treated cells (data not shown).
In contrast, LPS treatment of TLR2/4–/–-derived BMDCs did not result in either cell activation as judged by IL-6 secretion and up-regulation of costimulatory molecules or formation of FK2-positive protein accumulations, whereas both could be observed in wild-type C3H/HeN-derived BMDCs. These results indicate that PR/8/34-induced activation of BMDCs is not due to LPS contamination and suggest a tight coupling between TLR-mediated cell activation and formation of FK2-positive protein accumulations in BMDCs. Poly(I:C) treatment of TLR2/4–/– BMDCs was performed as positive control for the cell ability to get activated.
PR/8/34-infected BMDCs show DALIS formation
A characteristic feature of DALIS is their transient nature. They can be observed 4 h after LPS stimulation and start to disappear after reaching their maximum around 8 h (10). We investigated whether formation of accumulations of ubiquitinated proteins after PR/8/34 infection shows similar kinetics. BMDCs were infected with 1 PFU/cell at different time points, stained with FK2 Ab and analyzed by confocal microscopy. Aggregates of ubiquitinated proteins appeared as early as 4 h after infection in the cytosol of BMDCs and reached their maximum in size between 8 and 15 h. Afterward, they start to become smaller and are undetectable 43 h after virus infection (Fig. 5A).
To provide additional evidence that FK2 staining of BMDCs visualizes cytosolic DALIS located in the cytosol and not promyelocytic leukemia oncogenic domains (PODs) or microtubule organizing center (MTOC) (22), simultaneous staining of ubiquitinated protein aggregates with FK2 (green) and of the nucleus with 4',6-diamidino-2-phenylindole dihydrochloride (blue) was performed (Fig. 5B). The detection of several DALIS that are not associated with the nucleus distinguishes them from PODs and the MTOC.
DALIS are Triton X-100 resistant and appear in the detergent-insoluble fraction (10). Therefore, we infected BMDCs again with influenza virus at different time points, followed by Triton X-100 extraction, SDS-PAGE, and FK2 Western blot. We observed an increase in ubiquitinated proteins in the detergent-insoluble fraction (top row), whereas the level of detergent-soluble ubiquitinated proteins (bottom row) did not change (Fig. 5C, left panels). These experiments support the observation that influenza virus infection induces DALIS formation. EL-4 cells do not show the formation of DALIS after influenza infection, staining with FK2 Ab, and confocal microscopy analysis (data not shown). Consequently, no differences between detergent-soluble and -insoluble fractions could be detected in this cell type (Fig. 5C, right panels). Treatment with the proteasome inhibitor epoxomicin led in all cell types to abnormal accumulation of ubiquitinated proteins, both in the insoluble fractions (Fig. 5C, top panels) as well as in the soluble fractions (bottom panels).
Influenza NP can be found in DALIS
If the formation of DALIS is connected to the delayed presentation of the influenza NP CTL epitope, it should be possible to detect an association of this protein with DALIS. Therefore, we performed colocalization studies after PR/8/34 infection of BMDCs. After 12 h, colocalization (yellow) of anti-influenza NP (red) and FK2 (green) Abs was observed (Fig. 6A, right panel). Control staining with FK2 (green) and anti-20S proteasome (red) Abs revealed no colocalization (Fig. 6A, left panel). Noninfected cells show neither DALIS nor influenza NP staining (data not shown). Furthermore, we were able to detect influenza NP in DALIS-enriched fractions by influenza NP-specific Western blotting (NP-WB) directly (Fig. 6B, left panel). In addition, NP-specific IP (NP-IP) followed by NP (Fig. 6B, second panel) or FK2 Western blotting (FK2-WB) revealed ubiquitinated NP (third panel). A similar result was obtained after IP of ubiquitinated proteins (FK2 Ab; FK2-IP) and Western blotting with NP-specific Abs (NP-WB) (Fig. 6B, right panel). These results are compatible with the idea that DALIS contain ubiquitinated influenza NP. In all cases, LPS-matured BMDCs were used as negative control.
Maturation and DALIS formation concur in BMDCs
Our experiments indicate that there is a temporal connection between maturation, Ag processing, and DALIS formation in BMDCs. To support this finding, we examined the expression of surface MHC class I molecules as well as CD80 and CD86 during influenza-induced BMDC maturation. We observed up-regulation of MHC class I molecules 8 h after infection of BMDCs with PR/8/34 (Fig. 7A). Up-regulation of the costimulatory molecules CD80 and CD86 started after 12 h and reached its maximum 18 h after adding the virus (Fig. 7B). Thus, our results suggest a close relationship between up-regulation of MHC class I, CD80, and CD86 molecules, DALIS formation, and Ag presentation during BMDC maturation.
Discussion
Our study reveals that the presentation of the immunodominant influenza NP-derived CTL epitope is delayed in DCs compared with nonprofessional APCs like EL-4 or MC57 cells. For BMDCs, the peak of influenza NP-specific CTL responses as determined by target cell lysis and IFN- production was detectable only 8–10 h after infection (Fig. 1, A and B). Ag presentation was abolished by the inhibition of protein synthesis using CHX or by UV inactivation of the virus (Fig. 1C). In contrast, maximum CTL recognition of EL-4 and MC57 cells was observed already after 4 h (Fig. 1, A and B). These differences are not due to reduced or delayed influenza NP expression in BMDCs (Fig. 2). Furthermore, our data show that influenza virus infection leads to DC maturation and DALIS formation, which is detectable between 4 and 18 h after adding the stimulus, whereas nonprofessional APCs (EL-4 and MC57 cells) show no DALIS formation. The maximum size of DALIS was reached 8–15 h after infection. DALIS formation in influenza-infected DCs is not only supported by the transient nature of the FK2-positive aggregates (Fig. 5A; Ref. 10) but also by the observation that multiple aggregates per cell can be detected that are not attached to the nucleus (Fig. 5B). These features additionally separate DALIS from PODs and the MTOC.
The kinetics of DALIS formation coincide with the up-regulation of MHC class I molecules and the costimulatory molecules CD80 and CD86 on the cell surface (Fig. 7). The connection of DALIS formation and delayed Ag presentation is also supported by the observation that DALIS colocalize with influenza NP (Fig. 6A). In addition, DALIS-enriched fractions contain increasing amounts of ubiquitinated proteins (Fig. 5C, top left panel) that also contain influenza NP (Fig. 6B, left panel). The ability to visualize this protein after NP-specific IP (NP-IP) in combination with Western blotting against ubiquitinated proteins (FK2-WB) or vice versa (Fig. 6B, right panels) suggests that DALIS contain ubiquitinated influenza NP. The detection of only full-length NP might be due the specificity of the Abs used and will be further investigated by mass spectrometry analysis of proteins present in DALIS.
An increase in the ability of influenza virus-infected DCs to stimulate CTLs over time was observed previously by Cella et al. (23) showing that human DCs infected for 24 h were 10-fold more efficient in CTL activation than DCs infected for 5 h. In contrast, Crowe et al. (24) observed that NP-derived peptide presentation is most efficient 2–4 h after infection, then decreases between 6 and 8 h and increases again 24 h after infection. The reasons for this difference are not clear but could be related to the fact that the latter study used a DC-like cell line, whereas in the study by Cella et al. (23) and our experiments, primary DCs were used. Whether or not these cells differ in their abilities for DALIS formation remains to be determined.
What might be the reason for DCs to delay the presentation of antigenic peptides?
In contrast to nonprofessional APCs, in which early generation and presentation of CTL epitopes is required to allow fast elimination of infected cells by activated CTLs, the situation for DCs is different. Although viral infection and expression of virus-derived proteins are obviously very fast, DCs need time to acquire an activated phenotype that allows optimal activation of naive T cells. The storage of Ags in DALIS is a possible mechanism to delay the presentation of CTL epitopes and to allow DCs to reach complete maturation before they encounter T cells. The coordinated regulation of DC maturation and Ag presentation is to be expected, because DCs perform different functions within the immune system depending on their maturational state and location (25, 26).
It has been shown that optimal activation of naive CD8+ T cells requires Ag recognition on mature DCs that display the highest expression levels of MHC, costimulatory, and adhesion molecules. In contrast, the interaction of naive T cells with immature DCs favors the induction of tolerance. The latter can be avoided if Ag processing and presentation is synchronized with DC maturation by storing Ags in DALIS.
Because only endogenous proteins are found in DALIS, this mechanism does not delay the MHC class I-restricted presentation of peptides from exogenous proteins. This process, which operates efficiently in DCs, is called "cross-presentation" and leads either to cross-tolerance or cross-priming (27, 28, 29, 30). For cross-priming, fully matured DCs are necessary. Nevertheless, cross-presentation is regulated during DC maturation by the selective control of Ag internalization and transport to the cytosol (7). In fact, immature DCs hardly cross-present Ags, whereas this pathway is very efficient 7 h after DC stimulation (6). Thus, not only presentation of peptides from endogenous proteins but also from exogenous proteins is delayed in DCs, preventing the presentation of T cell epitopes by immature DCs. During maturation, DCs migrate from the source of the inflammatory stimulus to the draining lymph nodes where they localize to the T cell areas (2). This change in migratory capacity of the maturing DCs is due to a dramatic change in chemokine receptor expression. Maturing DCs down-regulate CCR1 and CCR5 as fast as 1–2 h after exposure to maturation stimuli, and up-regulation of the lymph node homing receptor CCR7 starts 3–4 h after stimulation (31). In addition, it has been shown that respiratory DCs can be found in draining peribronchial lymph nodes 6 h after influenza infection (32), which strikingly parallels the initiation of Ag presentation observed in our experiments.
The capability of mature DCs to induce T cell immunity is not unlimited but of a temporary nature. There seems to be an optimal time frame in which mature DCs are able to fulfill their job to activate specific T cells, which is 8–18 h after DC stimulation. Recently, it has been shown that DCs pulsed with the CTL epitope TRP-2180–188 and stimulated with LPS for 8 h elicited a more powerful CTL response in C57BL/6 mice than untreated DCs or DCs that were stimulated for 48 h (33). Due to the kinetics of IL-12 expression, DCs that were stimulated for 8 h reached the draining lymph nodes at a time at which that cytokine is maximally produced. Accordingly, these DCs can rapidly induce CD4+ T cell activation, possibly Th1 polarization, and priming of CTLs. With the loss of IL-12 production, DCs switch from a Th1- to a Th2-inducing mode, although one has to consider other factors that can influence the Th1-Th2 polarization (9). It is possible that the narrow window of time in which mature DCs elicit a potent T cell response minimalizes the risk of overactivation of the induced immune response.
However, there is another reason for DCs to postpone Ag presentation until maturation is induced. Activated CTLs are able to rapidly induce apoptosis of their target cells by exocytosis of cytotoxic granula or by cross-linking of death receptors. Therefore, the DCs that induced the CTL response are at risk to be eliminated, which would limit the capacity of DCs to prime CTL immunity. Immature DCs are susceptible to CTL-induced apoptosis, whereas mature DCs are protected through expression of a serine protease inhibitor (SPI-6). Interestingly, SPI-6 expression can be detected 8 h after stimulation (34) and thereby provides protection from CTL-mediated apoptosis when Ag presentation is initiated.
In summary, DCs appear to be most effective in the induction of T cell immunity between 8 and 18 h after activation. This is exemplified by the fact that the up-regulation of costimulatory molecules as well as MHC class I molecules, IL-12 production, SPI-6, and CCR7 expression peak in this time frame. Now, we show that MHC class I-restricted Ag presentation is also delayed and is preceded by DALIS appearance, which coincides with DC maturation. We hypothesize that these processes are tightly regulated in DCs to achieve the optimal activation of CTLs.
Acknowledgments
We thank Drs. J. Yewdell and Oliver Planz for the donation of the influenza nucleoprotein Abs. We also thank Dr. S. Pleschka for providing us the influenza virus PR/8/34. We are grateful to Dr. S. Tenzer and M. Schatz for helpful discussions and critical reading of this manuscript.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by grants from the Deutsche Forschungsgemeinschaft (Sonderforschungsbereich 490, B7 (to H.S.) and Sonderforschungsbereich 510, C1 (to H.S.) and the European Union (EC Project QLK3-CT-1999-00064).
2 Address correspondence and reprint requests to Prof. Hansj?rg Schild, Institute for Immunology, Johannes Gutenberg University, Hochhaus am Augustusplatz, D-55131 Mainz, Germany. E-mail address: schild@uni-mainz.de
3 Abbreviations used in this paper: DC, dendritic cell; BMDC, bone marrow-derived DC; CHX, cycloheximide; DALIS, DC aggresome-like-induced structure; DRiP, defective ribosomal product; IP, immunoprecipitation; NP, nucleoprotein; POD, promyelocytic leukemia oncogenic domain; MTOC, microtubule organizing center.
Received for publication April 20, 2004. Accepted for publication May 6, 2005.
References
Banchereau, J., F. Briere, C. Caux, J. Davoust, S. Lebecque, Y. T. Liu, B. Pulendran, K. Palucka. 2000. Immunobiology of dendritic cells. Annu. Rev. Immunol. 18: 767-811.
Banchereau, J., R. M. Steinman. 1998. Dendritic cells and the control of immunity. Nature 392: 245-252.
Lanzavecchia, A., F. Sallusto. 2000. Dynamics of T lymphocyte responses: intermediates, effectors, and memory cells. Science 290: 92-97.
Pamer, E., P. Cresswell. 1998. Mechanisms of MHC class I-restricted antigen processing. Annu. Rev. Immunol. 16: 323-358.
Hosken, N. A., M. J. Bevan, F. R. Carbone. 1989. Class I-restricted presentation occurs without internalization or processing of exogenous antigenic peptides. J. Immunol. 142: 1079-1083.
elamarre, L., H. Holcombe, I. Mellman. 2003. Presentation of exogenous antigens on major histocompatibility complex (MHC) class I and MHC class II molecules is differentially regulated during dendritic cell maturation. J. Exp. Med. 198: 111-122.
Gil-Torregrosa, B. C., A. M. Lennon-Dumenil, B. Kessler, P. Guermonprez, H. L. Ploegh, D. Fruci, P. van Endert, S. Amigorena. 2004. Control of cross-presentation during dendritic cell maturation. Eur. J. Immunol. 34: 398-407.
Muraille, E., C. De Trez, B. Pajak, M. Brait, J. Urbain, O. Leo. 2002. T cell-dependent maturation of dendritic cells in response to bacterial superantigens. J. Immunol. 168: 4352-4360.
Langenkamp, A., M. Messi, A. Lanzavecchia, F. Sallusto. 2000. Kinetics of dendritic cell activation: impact on priming of TH1, TH2 and nonpolarized T cells. Nat. Immunol. 1: 311-316.
Lelouard, H., E. Gatti, F. Cappello, O. Gresser, V. Camosseto, P. Pierre. 2002. Transient aggregation of ubiquitinated proteins during dendritic cell maturation. Nature 417: 177-182.
Lelouard, H., V. Ferrand, D. Marguet, J. Bania, V. Camosseto, A. David, E. Gatti, P. Pierre. 2004. Dendritic cell aggresome-like induced structures are dedicated areas for ubiquitination and storage of newly synthesized defective proteins. J. Cell Biol. 164: 667-675.
Johnston, J. A., C. L. Ward, R. R. Kopito. 1998. Aggresomes: a cellular response to misfolded proteins. J. Cell Biol. 143: 1883-1898.(Sylvia Herter*,, Philipp )
Influenza virus infection induces maturation of murine dendritic cells (DCs), which is most important for the initiation of an immune response. However, in contrast to EL-4 and MC57 cells, DCs present viral CTL epitopes with a delay of up to 10 h. This delay in Ag presentation coincides with the up-regulation of MHC class I molecules as well as costimulatory molecules on the cell surface and the accumulation of newly synthesized ubiquitinated proteins in large cytosolic structures, called DC aggresome-like-induced structures (DALIS). These structures were observed previously after LPS-induced maturation of DCs, and it was speculated that they play a role in the regulation of MHC class I Ag presentation. Our findings provide the first evidence for a connection between DC maturation, MHC class I-restricted Ag presentation, and DALIS formation, which is further supported by the observation that DALIS contain ubiquitinated influenza nucleoprotein.
Introduction
Dendritic cells (DCs)3 play a central role in the induction of primary adaptive immune responses by providing unspecific and specific stimuli for T cells (1). These abilities require the development from an immature into a mature state of differentiation and the presentation of a diverse array of Ags (2). The former is accompanied by the increase in the expression of surface MHC and costimulatory molecules and the secretion of proinflammatory cytokines (3). The latter requires efficient Ag uptake and processing that involves proteasomal activity for most Ags presented by MHC class I molecules (4). For the activation of naive T cells, these two independent processes should take place in a coordinated fashion. In nonprofessional APCs for example, MHC class I-restricted peptide presentation after influenza virus infection reaches its maximum already after 2–4 h (5). In professional APCs, MHC class I-restricted Ag presentation should be slowed down to allow it to take place together with maximum expression of costimulatory molecules and secretion of proinflammatory cytokines that require 12–24 h for the former (6, 7, 8) and 6–12 h for the latter (9). Recently, a new feature of DCs was discovered that could represent a mechanism to enable coordinated maturation and Ag presentation in this cell type. After LPS-induced maturation, DCs transiently accumulate newly synthesized ubiquitinated proteins in an aggregated form in the cytosol (10, 11). These protein aggregates are called DC aggresome-like-induced structures (DALIS) because they were observed initially in DCs and resemble aggresomes. Both DALIS and aggresomes contain ubiquitinated proteins, but in contrast to DALIS, aggresomes result from microtubule-dependent conglomeration of smaller aggregates and redistribution of the intermediate filament protein vimentin (12). Furthermore, DALIS formation is not due to impaired proteolysis and does not affect the ubiquitin-proteasome pathway. Ubiquitin is a conserved 8.5 kDa protein that is involved in many cellular functions, including gene expression, ribosome biosynthesis, receptor expression, and stress response. Ubiquitin is covalently attached to proteins that bind to and are degraded by 26S proteasomes (13, 14). This process results in the generation of peptides from which ligands presented by class I molecules are selected.
It has been shown that protein neosynthesis is necessary for the presentation of a T cell epitope from a long-lived viral protein that is a property anticipated by the defective ribosomal product (DRiP) hypothesis (15). It was speculated that most of the peptide ligands for MHC class I molecules are made from DRiPs (16), which are newly synthesized polypeptides that never attain native structure (17, 18, 19). Normally, DRiPs are degraded rapidly by proteasomes (20). However, in DCs, DRiPs are stored in DALIS, which delay their degradation and stabilize them for 8–16 h (11). Recently, it was shown that DALIS formation can occur not only in DCs but also in stimulated macrophages with a similar time course (21).
By delaying degradation of newly synthesized misfolded proteins, maturing DCs might have the ability to delay MHC class I loading and peptide presentation until the secretion of proinflammatory cytokines and the expression of costimulatory molecules has been initiated. This feature would prevent the presentation of T cell epitopes by immature DCs and therefore avoid the induction of tolerance.
In this study, we show that activation of bone marrow-derived DCs (BMDCs) by influenza virus infection results indeed in a delayed processing and presentation of viral Ags to CTLs compared with Ag presentation in nonprofessional APCs. This delay in Ag presentation coincides with DALIS formation during BMDC maturation. Furthermore, we show that DALIS contain ubiquitinated influenza nucleoprotein (NP). Thus, our results demonstrate that MHC class I-restricted Ag presentation and formation of DALIS appear to be tightly linked in DCs.
Materials and Methods
Generation of BMDCs
Mouse immature BMDCs were generated from bone marrow of C3H/HeN, C3H/HeJ/TLR2–/– (TLR2/4–/–), and C57BL/6 mice. For the generation of mouse BMDCs, IMDM (BioWhittaker) supplemented with 2 mM L-glutamine (Seromed), 100 U/ml penicillin/streptomycin (BioWhittaker), 5% FCS (Vitromex), and 200 U/ml GM-CSF (PeproTech) was used. Bone marrow cells were incubated in GM-CSF-containing medium for 6 or 7 days, with full medium replacement on day 3 (removal of nonadherent cells) and feeding on day 5. On day 6 or 7, nonadherent and semiadherent cells were harvested. These cells were typically CD14– and >80% CD11c+ as determined by FACS analysis. Furthermore, the CD11c+ cells were CD40low, CD80low, CD86low, and H-2 Kb medium.
Generation of NP366–374-specific T cell line
A C57BL/6 mouse (8 wk of age) was immunized i.p. with 500 hemagglutinating units of human influenza virus A (PR/8/34) (obtained from Dr. S. Pleschka, Institute for Virology, University of Giessen, Giessen, Germany). After 9 days, the mouse was killed, and the splenocytes were stimulated with the H2-Db-restricted immunodominant epitope from the influenza nucleoprotein (NP366–374). Cells were split and restimulated in weekly intervals using modified Eagle’s medium (-modification) supplemented with 10% (v/v) heat-inactivated FCS, 2 mM L-glutamine, antibiotics, and 50 μM 2-ME. Specificity and reactivity were tested routinely by 51Cr release assay.
IFN- and 51Cr release assay
BMDCs and EL-4 cells were stimulated at different time points with 1 PFU/cell human influenza virus A (PR/8/34). Half of the activated cells were incubated with 50 ng/ml NP366–374 peptide for 30 min before washing and fixation with 0.05% glutaraldehyde. Fifty thousand targets per well were incubated overnight with 100,000 T cells specific for the immunodominant viral NP366–374 epitope. After 20 h, supernatants were collected and IFN- was measured using sandwich ELISA (see below). For 51Cr release assays, BMDCs, EL-4 cells, and MC57 cells were infected with 1 PFU/cell PR/8/34 at 37°C. Thereafter, cells were counted and washed twice with medium before labeling with [51Cr]Na2CrO4 plus 5% FCS for 20 min. To control the general susceptibility to CTL lysis, half of the cells were incubated with 500 pmol of ASNENMETM during radioactive labeling. After intense washing, 20,000 targets were used in a 4-h release assay with 80,000 or 400,000 effector cells from a NP366–374-specific T cell line.
To determine the influence of protein neosynthesis on NP peptide presentation, BMDCs were treated with or without 40 μM cycloheximide (CHX; Sigma-Aldrich) during 6 h of PR/8/34 infection or cells were incubated with UV-treated influenza virus. For UV treatment, PR/8/34 was exposed to short-wave UV light (254 nm) for 5 min.
To test the viability of the CHX-treated cells, BMDCs were incubated with 500 pmol of ASNENMETM during radioactive labeling.
Cytokine ELISAs and FACS staining
BMDCs and EL-4 cells were activated with 1 PFU/cell PR/8/34, 50 μg/ml poly(I)poly(C) (poly(I:C); Amersham Biosciences), or 1 μg/ml LPS (Salmonella typhimurium; Sigma-Aldrich). After 18 h, supernatants were collected, and IL-6 was measured using standard sandwich ELISA protocol. Briefly, the capture Ab was coated on an ELISA plate (MaxiSorb; Nunc) and the biotinylated detection Ab was revealed by streptavidin-conjugated HRP and tetramethylbenzidine substrate (Sigma-Aldrich). The assay was read at 450 nm after adding 2 M H2SO4 to stop the reaction.
For FACS staining, BMDCs and EL-4 cells were activated at different time points by adding 1 PFU/cell PR/8/34. Cells were stained either with FITC-conjugated anti-CD80 and PE-conjugated anti-CD86 Abs or biotinylated anti-H-2 Kb/H-2 Db and streptavidin-FITC. All Abs and the cytokine standard were obtained from BD Biosciences. All analyses were performed using FACSCalibur and CellQuest software (BD Biosciences).
Immunoprecipitation
For influenza NP immunoprecipitation (IP), 3 x 106 BMDCs were treated with 1 PFU/cell PR/8/34 or with 1 μg/ml LPS for 11 h and washed two times with PBS before cell lysis was performed in PBS supplemented with Complete protease inhibitors (Roche Applied Science). After five freeze-thaw cycles, lysed cells were centrifuged at 12,000 rpm for 5 min before 8 μg of anti-influenza Ab (clone N3F3; gift from Dr. O. Planz, Institute for Immunology, Bundesforschungsanstalt für Viruserkrankungen der Tiere, Tübingen, Germany) was added to the supernatants. After a 2-h rotation, 100 μl of protein A-Sepharose were added for 1 h. Afterward, the Sepharose was washed three times with PBS before it was resuspended in SDS loading buffer and separated by SDS-PAGE. Then Western blots using anti-influenza Ab (N1A4; gift from Dr. O. Planz) or FK2 Ab as primary Abs were performed as described below.
For FK2 IP, 10 x 106 BMDCs were treated as described above. After 11 h of incubation, cells were washed two times with PBS before cell lysis with Triton lysis buffer (20 mM Tris-HCl (pH 7.4), 1% Triton X-100, 1.5 mM MgCl2, and 10 mM NaCl) at 4°C for 20 min was performed. After centrifugation at 12,000 x g for 10 min, pellets were resuspended in PBS with Complete protease inhibitors (Roche Applied Science) before adding 5 μg of biotinylated FK2 Ab (FK2 Ab was biotinylated with D-biotinoyl--aminocaprons?ure-N-hydroxysuccinimidester (Boehringer Mannheim) according to the manufacturer’s instructions). After 1.5 h at 4°C, 10 μl of MACS streptavidin microbeads (Miltenyi Biotec) and the IPs were performed following the manufacturer’s instructions. Afterward, pellets were resuspended in SDS loading buffer and separated by SDS-PAGE, followed by anti-influenza NP Western blotting as described below.
Western blotting
BMDCs, EL-4 cells, and MC57 cells were infected with 1 PFU/cell PR/8/34 at different time points or activated with LPS (1 μg/ml). Cells were harvested and washed three times with PBS before they were treated with lysis buffer (20 mM Tris-HCl (pH 7.4), 1% Triton X-100, 1.5 mM MgCl2, and 10 mM NaCl) at 4 °C for 20 min. After centrifugation at 12,000 x g for 10 min, supernatants and pellets were resuspended separately in SDS loading buffer and separated by SDS-PAGE. After electrophoresis, proteins were transferred to a Hybond ECL membrane (Pharmacia Biotech) by semidry Western blotting. The membrane was blocked for 1 h with 50 mM Tris, 150 mM NaCl, 0.1% Nonidet P-40, 3% BSA, and 5 mM EDTA and then incubated with rabbit anti-influenza NP Abs (gift from Dr. J. Yewdell, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD) or FK2 Ab (Affiniti/Biomol), followed by HRP anti-rabbit IgG (Amersham Biosciences) or HRP anti-mouse IgG (Jackson Immunoresearch/Dianova). For actin Western blotting, anti-actin (C2)-HRP (Santa Cruz Biotechnology) was used. To analyze the influence of CHX on influenza NP expression, cells were treated simultaneously with or without 100 μM CHX (Sigma) and 1 PFU/cell PR/8/34 for 6 h at 37°C before intense washing and counting. Afterward, cell pellets were resuspended in SDS loading buffer and separated by SDS-PAGE. The proteasome inhibitor epoxomicin (Biomol) was added 10 min before infection of BMDCs and EL-4 cells with the influenza virus at a final concentration of 1 μM. All Western blots were developed using ECL substrate (Pierce/Perbio) and analyzed by the ChemiDoc XRS System and Quantity One software (Bio-Rad).
Immunofluorescence staining and Abs
BMDCs, EL-4 cells, or MC57 cells were treated with 1 PFU/cell human influenza virus A PR/8/34, 50 μg/ml poly(I:C) (Amersham Biosciences), or 1 μg/ml LPS (Sigma-Aldrich) at different times. Activated cells were washed in PBS, 0.5% BSA (Roth), and 0.1% Saponin (Sigma-Aldrich) and fixed with Cytofix/Cytoperm (BD Pharmingen) for 20 min at 4°C. For DALIS staining, biotinylated FK2 Ab (Affiniti) and streptavidin-FITC (Sigma-Aldrich) were used as secondary reagents. Influenza NP staining was done with polyclonal anti-influenza NP serum (generously provided by Dr. J. Yewdell) and Alexa 546- or Alexa 594-conjugated anti-rabbit IgG (Molecular Probes). All Abs were diluted in washing buffer, and cells were stained 30 min at 4°C. In some cases, Ab incubation was followed by 4',6-diamidino-2-phenylindole dihydrochloride (Molecular Probes) staining for 10 min at room temperature. For colocalization experiments, BMDCs were harvested after 12 h of influenza infection and coated on 1% alcian blue (Sigma-Aldrich)-treated coverslips for 10 min at 37°C. Cells were permeabilized with 1% Triton X-100 in PBS for 5 min at 4°C and fixed with 3% paraformaldehyde in PBS for 10 min at room temperature before staining with FK2 and anti-influenza NP serum (provided by Dr. J. Yewdell) or FK2 and anti-20S proteasome Ab (BioTrend).
A Zeiss LSM 510 laser-scanning microscope or the Olympus Cell System was used.
Results
Influenza virus infection of BMDCs results in delayed Ag presentation
To analyze the kinetics of Ag presentation, we used BMDCs, EL-4 cells, and MC57 cells that were infected with human influenza virus A PR/8/34 at different time points as target cells for T cells specific for the immunodominant viral NP366–374 epitope. After 20 h of coincubation of T cells with fixed target cells, supernatants were analyzed by standard ELISA for IFN- (Fig. 1A). Whereas EL-4 cells induced IFN- production as early as 4 h after influenza infection, we found a delay in BMDC-induced IFN- secretion. In this study, IFN- production was first detectable after 12 h of infection. 51Cr release assays confirmed this observation. Influenza infection of the target cells for 4 h before adding them to the T cells was sufficient to induce maximum lysis of EL-4 cells. Similar kinetics were obtained for MC57 cells infected with the influenza virus. In contrast, specific killing of BMDCs increased until 10 h of virus infection (Fig. 1B). Target cells infected with PR/8/34 for >10 h showed spontaneous 51Cr release levels above 50% and were excluded from the experiments.
PR/8/34 induces BMDC maturation
Recently, it has been reported that during LPS-induced maturation, DCs accumulate newly synthesized ubiquitinated proteins in large cytosolic structures called DALIS (10). It was speculated that these structures play a role in the regulation of MHC class I Ag presentation. To test the ability of human influenza virus PR/8/34 to induce BMDC maturation, we infected BMDCs with 1 PFU/cell PR/8/34 or activated them with 1 μg/ml LPS or 50 μg/ml poly(I:C) and compared production of the proinflammatory cytokine IL-6 and up-regulation of the costimulatory molecules CD80 and CD86. All stimuli tested induced IL-6 secretion (Fig. 3A) and up-regulation of CD80 and CD86 (Fig. 3B). Poly(I:C)- and LPS-activated BMDCs showed similar and higher IL-6 production than PR/8/34-infected cells. In contrast, PR/8/34 infection induced a higher up-regulation of CD80 and CD86 molecules than LPS. Accumulation of ubiquitinated proteins during BMDC maturation was analyzed by confocal microscopy using the FK2 mAb, which is specific for mono- and poly-ubiquitinated proteins. The Ab does not react with free ubiquitin. FK2-positive accumulations of proteins could be observed in PR/8/34-infected BMDCs as well as in LPS-activated cells (Fig. 3C) and poly(I:C)-treated cells (data not shown).
In contrast, LPS treatment of TLR2/4–/–-derived BMDCs did not result in either cell activation as judged by IL-6 secretion and up-regulation of costimulatory molecules or formation of FK2-positive protein accumulations, whereas both could be observed in wild-type C3H/HeN-derived BMDCs. These results indicate that PR/8/34-induced activation of BMDCs is not due to LPS contamination and suggest a tight coupling between TLR-mediated cell activation and formation of FK2-positive protein accumulations in BMDCs. Poly(I:C) treatment of TLR2/4–/– BMDCs was performed as positive control for the cell ability to get activated.
PR/8/34-infected BMDCs show DALIS formation
A characteristic feature of DALIS is their transient nature. They can be observed 4 h after LPS stimulation and start to disappear after reaching their maximum around 8 h (10). We investigated whether formation of accumulations of ubiquitinated proteins after PR/8/34 infection shows similar kinetics. BMDCs were infected with 1 PFU/cell at different time points, stained with FK2 Ab and analyzed by confocal microscopy. Aggregates of ubiquitinated proteins appeared as early as 4 h after infection in the cytosol of BMDCs and reached their maximum in size between 8 and 15 h. Afterward, they start to become smaller and are undetectable 43 h after virus infection (Fig. 5A).
To provide additional evidence that FK2 staining of BMDCs visualizes cytosolic DALIS located in the cytosol and not promyelocytic leukemia oncogenic domains (PODs) or microtubule organizing center (MTOC) (22), simultaneous staining of ubiquitinated protein aggregates with FK2 (green) and of the nucleus with 4',6-diamidino-2-phenylindole dihydrochloride (blue) was performed (Fig. 5B). The detection of several DALIS that are not associated with the nucleus distinguishes them from PODs and the MTOC.
DALIS are Triton X-100 resistant and appear in the detergent-insoluble fraction (10). Therefore, we infected BMDCs again with influenza virus at different time points, followed by Triton X-100 extraction, SDS-PAGE, and FK2 Western blot. We observed an increase in ubiquitinated proteins in the detergent-insoluble fraction (top row), whereas the level of detergent-soluble ubiquitinated proteins (bottom row) did not change (Fig. 5C, left panels). These experiments support the observation that influenza virus infection induces DALIS formation. EL-4 cells do not show the formation of DALIS after influenza infection, staining with FK2 Ab, and confocal microscopy analysis (data not shown). Consequently, no differences between detergent-soluble and -insoluble fractions could be detected in this cell type (Fig. 5C, right panels). Treatment with the proteasome inhibitor epoxomicin led in all cell types to abnormal accumulation of ubiquitinated proteins, both in the insoluble fractions (Fig. 5C, top panels) as well as in the soluble fractions (bottom panels).
Influenza NP can be found in DALIS
If the formation of DALIS is connected to the delayed presentation of the influenza NP CTL epitope, it should be possible to detect an association of this protein with DALIS. Therefore, we performed colocalization studies after PR/8/34 infection of BMDCs. After 12 h, colocalization (yellow) of anti-influenza NP (red) and FK2 (green) Abs was observed (Fig. 6A, right panel). Control staining with FK2 (green) and anti-20S proteasome (red) Abs revealed no colocalization (Fig. 6A, left panel). Noninfected cells show neither DALIS nor influenza NP staining (data not shown). Furthermore, we were able to detect influenza NP in DALIS-enriched fractions by influenza NP-specific Western blotting (NP-WB) directly (Fig. 6B, left panel). In addition, NP-specific IP (NP-IP) followed by NP (Fig. 6B, second panel) or FK2 Western blotting (FK2-WB) revealed ubiquitinated NP (third panel). A similar result was obtained after IP of ubiquitinated proteins (FK2 Ab; FK2-IP) and Western blotting with NP-specific Abs (NP-WB) (Fig. 6B, right panel). These results are compatible with the idea that DALIS contain ubiquitinated influenza NP. In all cases, LPS-matured BMDCs were used as negative control.
Maturation and DALIS formation concur in BMDCs
Our experiments indicate that there is a temporal connection between maturation, Ag processing, and DALIS formation in BMDCs. To support this finding, we examined the expression of surface MHC class I molecules as well as CD80 and CD86 during influenza-induced BMDC maturation. We observed up-regulation of MHC class I molecules 8 h after infection of BMDCs with PR/8/34 (Fig. 7A). Up-regulation of the costimulatory molecules CD80 and CD86 started after 12 h and reached its maximum 18 h after adding the virus (Fig. 7B). Thus, our results suggest a close relationship between up-regulation of MHC class I, CD80, and CD86 molecules, DALIS formation, and Ag presentation during BMDC maturation.
Discussion
Our study reveals that the presentation of the immunodominant influenza NP-derived CTL epitope is delayed in DCs compared with nonprofessional APCs like EL-4 or MC57 cells. For BMDCs, the peak of influenza NP-specific CTL responses as determined by target cell lysis and IFN- production was detectable only 8–10 h after infection (Fig. 1, A and B). Ag presentation was abolished by the inhibition of protein synthesis using CHX or by UV inactivation of the virus (Fig. 1C). In contrast, maximum CTL recognition of EL-4 and MC57 cells was observed already after 4 h (Fig. 1, A and B). These differences are not due to reduced or delayed influenza NP expression in BMDCs (Fig. 2). Furthermore, our data show that influenza virus infection leads to DC maturation and DALIS formation, which is detectable between 4 and 18 h after adding the stimulus, whereas nonprofessional APCs (EL-4 and MC57 cells) show no DALIS formation. The maximum size of DALIS was reached 8–15 h after infection. DALIS formation in influenza-infected DCs is not only supported by the transient nature of the FK2-positive aggregates (Fig. 5A; Ref. 10) but also by the observation that multiple aggregates per cell can be detected that are not attached to the nucleus (Fig. 5B). These features additionally separate DALIS from PODs and the MTOC.
The kinetics of DALIS formation coincide with the up-regulation of MHC class I molecules and the costimulatory molecules CD80 and CD86 on the cell surface (Fig. 7). The connection of DALIS formation and delayed Ag presentation is also supported by the observation that DALIS colocalize with influenza NP (Fig. 6A). In addition, DALIS-enriched fractions contain increasing amounts of ubiquitinated proteins (Fig. 5C, top left panel) that also contain influenza NP (Fig. 6B, left panel). The ability to visualize this protein after NP-specific IP (NP-IP) in combination with Western blotting against ubiquitinated proteins (FK2-WB) or vice versa (Fig. 6B, right panels) suggests that DALIS contain ubiquitinated influenza NP. The detection of only full-length NP might be due the specificity of the Abs used and will be further investigated by mass spectrometry analysis of proteins present in DALIS.
An increase in the ability of influenza virus-infected DCs to stimulate CTLs over time was observed previously by Cella et al. (23) showing that human DCs infected for 24 h were 10-fold more efficient in CTL activation than DCs infected for 5 h. In contrast, Crowe et al. (24) observed that NP-derived peptide presentation is most efficient 2–4 h after infection, then decreases between 6 and 8 h and increases again 24 h after infection. The reasons for this difference are not clear but could be related to the fact that the latter study used a DC-like cell line, whereas in the study by Cella et al. (23) and our experiments, primary DCs were used. Whether or not these cells differ in their abilities for DALIS formation remains to be determined.
What might be the reason for DCs to delay the presentation of antigenic peptides?
In contrast to nonprofessional APCs, in which early generation and presentation of CTL epitopes is required to allow fast elimination of infected cells by activated CTLs, the situation for DCs is different. Although viral infection and expression of virus-derived proteins are obviously very fast, DCs need time to acquire an activated phenotype that allows optimal activation of naive T cells. The storage of Ags in DALIS is a possible mechanism to delay the presentation of CTL epitopes and to allow DCs to reach complete maturation before they encounter T cells. The coordinated regulation of DC maturation and Ag presentation is to be expected, because DCs perform different functions within the immune system depending on their maturational state and location (25, 26).
It has been shown that optimal activation of naive CD8+ T cells requires Ag recognition on mature DCs that display the highest expression levels of MHC, costimulatory, and adhesion molecules. In contrast, the interaction of naive T cells with immature DCs favors the induction of tolerance. The latter can be avoided if Ag processing and presentation is synchronized with DC maturation by storing Ags in DALIS.
Because only endogenous proteins are found in DALIS, this mechanism does not delay the MHC class I-restricted presentation of peptides from exogenous proteins. This process, which operates efficiently in DCs, is called "cross-presentation" and leads either to cross-tolerance or cross-priming (27, 28, 29, 30). For cross-priming, fully matured DCs are necessary. Nevertheless, cross-presentation is regulated during DC maturation by the selective control of Ag internalization and transport to the cytosol (7). In fact, immature DCs hardly cross-present Ags, whereas this pathway is very efficient 7 h after DC stimulation (6). Thus, not only presentation of peptides from endogenous proteins but also from exogenous proteins is delayed in DCs, preventing the presentation of T cell epitopes by immature DCs. During maturation, DCs migrate from the source of the inflammatory stimulus to the draining lymph nodes where they localize to the T cell areas (2). This change in migratory capacity of the maturing DCs is due to a dramatic change in chemokine receptor expression. Maturing DCs down-regulate CCR1 and CCR5 as fast as 1–2 h after exposure to maturation stimuli, and up-regulation of the lymph node homing receptor CCR7 starts 3–4 h after stimulation (31). In addition, it has been shown that respiratory DCs can be found in draining peribronchial lymph nodes 6 h after influenza infection (32), which strikingly parallels the initiation of Ag presentation observed in our experiments.
The capability of mature DCs to induce T cell immunity is not unlimited but of a temporary nature. There seems to be an optimal time frame in which mature DCs are able to fulfill their job to activate specific T cells, which is 8–18 h after DC stimulation. Recently, it has been shown that DCs pulsed with the CTL epitope TRP-2180–188 and stimulated with LPS for 8 h elicited a more powerful CTL response in C57BL/6 mice than untreated DCs or DCs that were stimulated for 48 h (33). Due to the kinetics of IL-12 expression, DCs that were stimulated for 8 h reached the draining lymph nodes at a time at which that cytokine is maximally produced. Accordingly, these DCs can rapidly induce CD4+ T cell activation, possibly Th1 polarization, and priming of CTLs. With the loss of IL-12 production, DCs switch from a Th1- to a Th2-inducing mode, although one has to consider other factors that can influence the Th1-Th2 polarization (9). It is possible that the narrow window of time in which mature DCs elicit a potent T cell response minimalizes the risk of overactivation of the induced immune response.
However, there is another reason for DCs to postpone Ag presentation until maturation is induced. Activated CTLs are able to rapidly induce apoptosis of their target cells by exocytosis of cytotoxic granula or by cross-linking of death receptors. Therefore, the DCs that induced the CTL response are at risk to be eliminated, which would limit the capacity of DCs to prime CTL immunity. Immature DCs are susceptible to CTL-induced apoptosis, whereas mature DCs are protected through expression of a serine protease inhibitor (SPI-6). Interestingly, SPI-6 expression can be detected 8 h after stimulation (34) and thereby provides protection from CTL-mediated apoptosis when Ag presentation is initiated.
In summary, DCs appear to be most effective in the induction of T cell immunity between 8 and 18 h after activation. This is exemplified by the fact that the up-regulation of costimulatory molecules as well as MHC class I molecules, IL-12 production, SPI-6, and CCR7 expression peak in this time frame. Now, we show that MHC class I-restricted Ag presentation is also delayed and is preceded by DALIS appearance, which coincides with DC maturation. We hypothesize that these processes are tightly regulated in DCs to achieve the optimal activation of CTLs.
Acknowledgments
We thank Drs. J. Yewdell and Oliver Planz for the donation of the influenza nucleoprotein Abs. We also thank Dr. S. Pleschka for providing us the influenza virus PR/8/34. We are grateful to Dr. S. Tenzer and M. Schatz for helpful discussions and critical reading of this manuscript.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by grants from the Deutsche Forschungsgemeinschaft (Sonderforschungsbereich 490, B7 (to H.S.) and Sonderforschungsbereich 510, C1 (to H.S.) and the European Union (EC Project QLK3-CT-1999-00064).
2 Address correspondence and reprint requests to Prof. Hansj?rg Schild, Institute for Immunology, Johannes Gutenberg University, Hochhaus am Augustusplatz, D-55131 Mainz, Germany. E-mail address: schild@uni-mainz.de
3 Abbreviations used in this paper: DC, dendritic cell; BMDC, bone marrow-derived DC; CHX, cycloheximide; DALIS, DC aggresome-like-induced structure; DRiP, defective ribosomal product; IP, immunoprecipitation; NP, nucleoprotein; POD, promyelocytic leukemia oncogenic domain; MTOC, microtubule organizing center.
Received for publication April 20, 2004. Accepted for publication May 6, 2005.
References
Banchereau, J., F. Briere, C. Caux, J. Davoust, S. Lebecque, Y. T. Liu, B. Pulendran, K. Palucka. 2000. Immunobiology of dendritic cells. Annu. Rev. Immunol. 18: 767-811.
Banchereau, J., R. M. Steinman. 1998. Dendritic cells and the control of immunity. Nature 392: 245-252.
Lanzavecchia, A., F. Sallusto. 2000. Dynamics of T lymphocyte responses: intermediates, effectors, and memory cells. Science 290: 92-97.
Pamer, E., P. Cresswell. 1998. Mechanisms of MHC class I-restricted antigen processing. Annu. Rev. Immunol. 16: 323-358.
Hosken, N. A., M. J. Bevan, F. R. Carbone. 1989. Class I-restricted presentation occurs without internalization or processing of exogenous antigenic peptides. J. Immunol. 142: 1079-1083.
elamarre, L., H. Holcombe, I. Mellman. 2003. Presentation of exogenous antigens on major histocompatibility complex (MHC) class I and MHC class II molecules is differentially regulated during dendritic cell maturation. J. Exp. Med. 198: 111-122.
Gil-Torregrosa, B. C., A. M. Lennon-Dumenil, B. Kessler, P. Guermonprez, H. L. Ploegh, D. Fruci, P. van Endert, S. Amigorena. 2004. Control of cross-presentation during dendritic cell maturation. Eur. J. Immunol. 34: 398-407.
Muraille, E., C. De Trez, B. Pajak, M. Brait, J. Urbain, O. Leo. 2002. T cell-dependent maturation of dendritic cells in response to bacterial superantigens. J. Immunol. 168: 4352-4360.
Langenkamp, A., M. Messi, A. Lanzavecchia, F. Sallusto. 2000. Kinetics of dendritic cell activation: impact on priming of TH1, TH2 and nonpolarized T cells. Nat. Immunol. 1: 311-316.
Lelouard, H., E. Gatti, F. Cappello, O. Gresser, V. Camosseto, P. Pierre. 2002. Transient aggregation of ubiquitinated proteins during dendritic cell maturation. Nature 417: 177-182.
Lelouard, H., V. Ferrand, D. Marguet, J. Bania, V. Camosseto, A. David, E. Gatti, P. Pierre. 2004. Dendritic cell aggresome-like induced structures are dedicated areas for ubiquitination and storage of newly synthesized defective proteins. J. Cell Biol. 164: 667-675.
Johnston, J. A., C. L. Ward, R. R. Kopito. 1998. Aggresomes: a cellular response to misfolded proteins. J. Cell Biol. 143: 1883-1898.(Sylvia Herter*,, Philipp )