Differential Expression Regulation of the and Subunits of the PA28 Proteasome Activator in Mature Dendritic Cells1
http://www.100md.com
免疫学杂志 2005年第12期
Abstract
Activation of dendritic cells (DC) by Th-dependent (CD40) or -independent (LPS, CpG, or immune complexes) agonistic stimuli strongly enhances the expression of the proteasome activator PA28 complex. Upon activation of DC, increased MHC class I presentation occurred of the melanocyte-associated epitope tyrosinase-related protein 2180-188 in a PA28-dependent manner. In contrast to other cell types, regulation of PA28 expression in DC after maturation was found to be IFN- independent. In the present study, we show that expression of PA28 and subunits was differentially regulated. Firstly, PA28 expression is high in both immature and mature DC. In contrast, PA28 expression is low in immature DC and strongly increased in mature DC. Secondly, we show the presence of a functional NF-B site in the PA28 promoter, which is absent in the PA28 promoter, indicating regulation of PA28 expression by transcription factors of the NF-B family. In addition, glycerol gradient analysis of DC lysates revealed elevated PA28 complex formation upon maturation. Thus, induction of PA28 expression allows proper PA28 complex formation, thereby enhancing proteasome activity in activated DC. Therefore, maturation of DC not only improves costimulation but also MHC class I processing. This mechanism enhances the CD8+ CTL (cross)-priming capacity of mature DC.
Introduction
Dendritic cells (DC)4 are specialized APC that are able to present antigenic peptides in both MHC class II and I molecules for optimal priming of Ag-specific T lymphocytes. Efficient uptake of antigenic protein followed by processing to peptide fragments suited for presentation in MHC molecules are crucial functions of these cells. DC can efficiently present Ags in MHC class II molecules using the exogenous route of Ag uptake and degradation in the endosomal/lysosomal pathway (reviewed in Ref. 1). In contrast, the MHC class I presentation route in DC is less well understood but is in general dependent on proteasome activity, requiring antigenic proteins to enter the cytosol after exogenous uptake.
After Ag presentation in MHC class II molecules, specific Th cells can interact with DC to deliver signals that lead to an activated state of DC. This is an essential step for efficient priming of MHC class I-restricted CTL, which requires cognate interactions between CD40L (CD154) on Th cells and CD40 on DC (2, 3, 4). Th-independent (innate) signals such as bacterial LPS or CpG, triggering specific TLR, can also induce the DC maturation required for CTL priming (5, 6, 7, 8). This mature state is characterized by up-regulated expression of several cell surface molecules, including the costimulatory molecules CD80 and CD86 and MHC class II and class I. A recent report shows that triggering of TLR on bone marrow-derived DC (BMDC) can improve cross-presentation of protein-derived antigenic peptides in MHC class I (9). Delamarre et al. (10) showed that presentation of exogenous Ags on MHC class I and class II molecules is differentially regulated during DC maturation, although a molecular explanation for this was not provided.
The immunomodulatory cytokine IFN-, produced during an immune response by activated Th 1 lymphocytes, CD8+ CTL, and NK cells, enhances Ag presentation by up-regulation of MHC and TAP gene products, as well as proteasome subunits and regulators (11). IFN- alters proteasome activity qualitatively by induction of three newly synthesized proteasome subunits LMP2, LMP7, and MECL-1, which replace the three constitutive 1 (), 2 (MB1), and 5 (Z) subunits, respectively, in the core 20S proteasome. Several groups reported that these so-called immunosubunits are incorporated interdependently (12, 13, 14, 15). Taken together with the reported differences in fine specificity (16, 17, 18), these findings imply that two types of proteasomes exist: "constitutive" proteasomes, which are expressed in all somatic cells, and "immunoproteasomes," which are expressed under influence of immunomodulatory cytokines such as IFN-. However, immunoproteasomes can be constitutively expressed in some cells, in particular, lymphoid cells and DC (19, 20). In vitro studies showed that proteasomes equipped with LMP2, LMP7, and MECL-1 are more efficient in the generation of several antigenic viral peptides (21). Mice that were made deficient for LMP2 (22) or LMP7 (23) can be regarded as immunoproteasome-deficient mice. These mice showed impairment in the presentation of some Ags, indicating that immunoproteasomes are indeed involved in generation of immunologically relevant antigenic peptides in vivo. Recently, LMP2-deficient mice have revealed that immunoproteasomes can also affect shaping of the CTL repertoire (24).
In addition to proteasome immunosubunits, IFN- also regulates the expression of the proteasome-associated complex PA28 (REG or the 11S regulator). PA28 was identified as a proteasome activator that strongly increases the maximal velocity of the hydrolytic reaction and decreases the concentration of substrate required for cleavage by purified 20S proteasomes (25, 26). Isolated PA28 does not display enzymatic activity by itself. Three homologous PA28 subunits have been cloned thus far, of which, only the and subunits are strongly induced by IFN- (27). Protein chemical analyses have revealed that the and subunits form a tightly connected heterohexameric or heteroheptameric complex, composed of comparable amounts of both subunits (28). The PA28 heteromultimer is able to bind to the rings of the 20S core proteasome, creating bell-shaped cap structures on both ends (11) or on one end, combined with the PA700 (or 19S) cap (the ubiquitin-dependent targeting/unfolding ATPase complex) on the other end (29, 30), to form the so-called "hybrid proteasome." The mechanism of PA28-mediated enhancement of intrinsic proteasome activity is not fully clear, although the cocrystalization structure of PA26 (the trypanosome homologue of PA28) with the 20S catalytic core shows a conformational change in the subunits that leads to an opening of the exit gate of the proteasome (31).
Both in vitro and in vivo studies have provided evidence that PA28 was implicated in MHC class I Ag processing. In vitro digestion studies with long peptides comprising dominant CTL epitopes showed that purified PA28 can enhance cleavage by the 20S proteasome, leading to augmented epitope liberation (32). Furthermore, expression of PA28 in mouse fibroblasts increased the sensitivity for lysis by virus-specific CTL (33). We have shown that upon controlled expression of PA28 in a tetracycline-dependent system, functional presentation of some Moloney leukemia virus CTL epitopes was increased significantly (34). The absence of PA28 in vivo reduces epitope processing and CTL recognition of some Ags, as shown in PA28-deficient mice (35). Furthermore, PA28 expression in melanoma cells could rescue processing and presentation of a CTL epitope from the tyrosinase-related protein 2 (TRP-2) Ag that was normally not presented in these tumor cells (36).
Macagno et al. (19, 20) have shown in human DC that PA28 and expression was elevated after DC maturation. These reports show that PA28 expression is independent of assembly of constitutive and immunoproteasomes. However, the expression regulation of these PA28 subunits in maturing DC was not addressed.
So far, IFN- appears to be the main regulator of the processing and presentation pathway. We have now investigated the role of DC activation via Th-dependent (CD40) and Th-independent signals (LPS) on the MHC class I-processing pathway. For this study, we have used the well-characterized mouse DC cell line D1 (37). This cell line is a growth factor-dependent spleen-derived DC cell line that, upon appropriate culture, maintains a fully immature state associated with efficient Ag uptake capacity. When activated, it matures both phenotypically and functionally (5). We have observed that such activation is required for efficient MHC class I presentation of a viral CTL epitope or for presentation of the SIINFEKL epitope from the OVA protein as delivered to DC via Ab-Ag immune complexes (IC) (38). This way of cross-presentation in MHC class I is FcR and proteasome dependent (39). DC that have been pulsed with protein Ag via IC very efficiently cross-prime CD8+ CTL responses in vivo (38). Interestingly, Pierre and coworkers (40, 41) have shown the presence of ubiquitin aggregates (so-called DALIS) in mature DC, suggesting a role of proteasome activity in direct MHC class I presentation.
Our detailed analysis of the changes in the molecular machinery of MHC class I Ag processing during maturation of DC shows enhanced expression of the PA28-chain of the PA28 heteromeric, proteasome-associated complex. Both LPS- and CD40-mediated stimulation strongly induced RNA and protein expression of PA28, independently of IFN-. The induced PA28 complexes were functionally active in mature D1 DC. In immature cells, the assembly of PA28 complexes appeared to be limited by the low expression of PA28, whereas PA28 expression was already high. This study indicates differentially regulated gene expression of these two molecules.
These findings are compatible with the "license to kill model" in which maturation of DC is a crucial step for initiation of effective CTL immune responses. Our data show that maturation of DC not only improves "signal two" but also enhances "signal one."
Materials and Methods
Nuclear extracts of 10 x 106 nonstimulated and stimulated D1 cells were prepared as described previously (45). The total amount of protein was determined using the BCA Protein Assay kit (Pierce), according to the manufacturer’s instructions.
EMSA was performed by incubating nuclear extracts (5 μg of protein) in DNA protein-binding buffer (10 mM HEPES (pH 7.9), 60 mM KCl, 10% v/v glycerol, 1 mM DTT, 1 mM EDTA, 3 mM MgCl2, and 10 mM NaPi), with 200 ng of poly(dI:dC), 200 ng of sonicated single-stranded herring sperm DNA, and 1 ng of 32P-labeled probe for 30 min at 4°C. The samples were run on a 6% nondenaturing gel in 0.25x Tris-borate EDTA buffer at 200 V for 2 h. The gels were fixed with a 10% methanol, 10% acetic acid solution, dried onto Whatmann 3M paper, and exposed to a x-ray film. The ds-oligonucleotide encompassing the potential NF-B site within the promoter sequence of the murine PA28 gene (5'-TCCCCGGCGGGGAAGTCCCACCTTATC-3') was used as a probe for the EMSA (46). A known NF-B-binding sequence from HLA class I (45) was used as competitor probe.
For supershift assays, 1 μg of each Ab directed against members of the NF-B/Rel family of transcription factors was added to the nuclear extract and probe mixture and incubated for 1 h at 4°C. The Abs used were RelB (sc-226x), RelA (sc-109x), and p50 (sc-114x; Santa Cruz Biotechnology).
Results
DC maturation enhances CTL recognition in a PA28-dependent fashion
The involvement of PA28 in CTL epitope processing in maturing DC was tested by measuring the presentation of the PA28-dependent CTL epitope 180-188 of the melanoma Ag TRP-2. We prepared BMDC cultures from PA28-deficient mice (PA28KO) (35) and wild-type C57BL/6 (B6) mice. These DC were either left immature or treated with LPS for 48 h to induce maturation as indicated by elevated expression of CD86, CD40, and MHC class II and class I (data not shown). After 48 h, the cells were incubated with a processing-dependent synthetic peptide 180-193 harboring the TRP-2 epitope and a 5-aa C-terminal extension. CTL epitope presentation was monitored with a TRP-2-specific CTL clone (Fig. 1A). The data show that both B6 and PA28KO immature DC are very inefficient in processing and presentation of the TRP-2 epitope, whereas LPS-matured B6 DC showed very efficient presentation. In contrast, PA28KO DC showed strongly reduced capacity of CTL epitope presentation, indicating the direct involvement of PA28 in the efficiency of processing of this epitope. The same cells were incubated with the processing-independent, minimal 9-mer peptide, which was equally well presented to the CTL by all types of DC (Fig. 1B).
On the basis of these results, we performed an in depth molecular analysis of the PA28-proteasome-dependent Ag MHC class I-processing pathway in DC.
Constitutive high immunoproteasome expression in immature and mature DC
The well-defined murine DC line D1 develops functional maturation upon stimulation with bacterial LPS or CD40 ligation by agonistic mAb as evidenced by increased CD86, MHC class II and class I cell surface expression, and IL-12 production (5). Because of its homogeneity, this cell line is an excellent tool for biochemical studies. The expression of the proteasome immunosubunits LMP2, LMP7, and MECL-1 was determined in total cell lysates by Western blot analysis. LMP7 was already strongly expressed by immature D1 cells and was not detectably up-regulated by IFN-, LPS, or anti-CD40 treatment (Fig. 2A, upper panel). LMP2 and MECL-1 analysis showed similar results (data not shown). Also, the lymphoid cell lines RMA (T cell lymphoma) and 771 (B cell lymphoma; data not shown) expressed high basal levels of LMP7. In contrast, fibroblast-type MEC did not show detectable expression of LMP7 (Fig. 2A). In MEC, LMP7 was up-regulated by IFN- treatment but failed to be induced by LPS.
These results indicate that immature DC already possess high constitutive expression levels of immunoproteasomes that are only marginally enhanced upon DC activation.
PA28 expression is significantly increased after DC maturation
In the same Western blot experiments, we observed that the expression levels of PA28 and PA28 were markedly enhanced upon IFN- treatment but also following LPS and CD40 stimulation of D1 cells (Fig. 2A, lower panels). PA28 is already expressed at relatively high levels in resting DC and is only marginally increased after the different treatments. In contrast to this, PA28 expression is very low in immature DC and strongly enhanced after DC maturation. Next, we analyzed the expression kinetics of PA28 and PA28 after LPS or CD40 stimulation (Fig. 2B). This experiment shows that PA28 expression is already detectably increased after 24 h of LPS or anti-CD40 treatment and reaches strongly elevated levels after 72 h of activation. In contrast, PA28 expression level is constitutively high in nonstimulated cells and is only moderately increased upon stimulation. These results suggest a differential regulation of PA28 and gene expression in DC. Similar results were obtained with freshly isolated BMDC cultures that were matured with LPS or the TLR9 ligand CpG (Fig. 2C). The kinetics and differential up-regulation of protein expression of PA28 and in these primary DC is similar to that of D1 cells, showing that D1 is a representative DC line.
Previously, we have shown that FcR-mediated activation and FcR-facilitated Ag uptake strongly improved MHC class I presentation and CTL cross-priming (38). In the present study, we have analyzed the ability of Ab-Ag IC to induce PA28 expression in DC. Fig. 2D shows strong induction of PA28 in both D1 cell line and freshly isolated BMDC. As a control, PA28 expression was absent in PA28-deficient BMDC.
In comparison to DC, the expression levels of PA28 are intermediate in the lymphoid cell line RMA and low in MEC (Fig. 2A). The expression level of PA28 was low in RMA and almost undetectable in MEC. LPS or CD40 stimulation did not significantly influence PA28 or expression in these (non-DC) cell lines. However, IFN- could induce expression of both PA28 molecules in MEC fibroblast cells.
These data suggest that TLR- or CD40-induced expression of the (classically IFN--inducible) PA28 molecules in DC is independent of IFN-. This is supported by the fact that D1 DC do not produce detectable amounts of IFN-, not even after activation by LPS or via CD40 (determined by specific ELISA; data not shown). To exclude the possibility of a role of secreted IFN- by D1 cells, LPS or CD40 stimulation was also performed in the presence of neutralizing anti-IFN- Abs. This Ab could block the expression of PA28 and as induced by exogenously added IFN- but not the induction of expression of PA28 and by LPS or CD40 stimulation (data not shown). Importantly, BMDC of IFN-R-deficient mice were analyzed for PA28 and expression after maturation with LPS. Fig. 2E shows similarly strong induction of the proteasome activator subunits, indicating IFN- independence.
PA28 and molecules in mature DC are complexed and functional
PA28 and molecules are highly expressed in mature DC. To test whether these molecules are present in functional PA28 complexes, total lysates were subjected to glycerol gradient analysis. Under these conditions, PA28 complexes remain intact and functional. Fig. 3A shows the result of a Western blot analysis of the individual fractions of the glycerol gradients of immature D1 cells and CD40-matured D1 cells. Early fractions 3–5 contain the monomeric PA28 molecules, and fractions 7–10 contain the complexed PA28 molecules. The presence of PA28 is shown in Fig. 3A, left panel. This analysis shows that the total level of PA28 does not increase significantly upon DC activation by anti-CD40 Ab, but rather shifts from the monomeric form to the complexed form. In contrast, PA28 expression (Fig. 3A, right panel) strongly increases following DC activation, and virtually all of the expressed PA28 is then present in the complexed form (fractions 7–10). Fig. 3B shows a similar experiment with LPS-stimulated D1 cells. Although the shift of monomeric to complexed PA28 is less striking, the increased expression of PA28 is apparent. The increased PA28 complex formation is similar to that seen following CD40 stimulation of D1 and again appears to be limited by PA28 expression.
These complexes were tested for their functional activity by incubation of the individual glycerol gradient fractions with purified 20S proteasomes, allowing evaluation of the enhancing effect of the PA28 proteasome activator. The cleavage of a fluorigenic synthetic peptide substrate was measured as shown in Fig. 4A. This figure shows a strong increase in proteolytic activity in fractions 7–10 of both the lysates of CD40- and LPS-matured D1 cells in comparison to the same fractions of immature D1 cells. The basal activity of these fractions, in the absence of added purified 20S proteasomes, is low (Fig. 4B). Only in the later fractions (13, 14, 15, 16, 17) endogenous proteolytic activity can be measured. The early glycerol fractions 3–5, containing monomeric PA28 and subunits, do not exhibit strong functional activity.
These data indicate that PA28 is already expressed at high, saturated levels in immature D1 cells and that functional PA28 complex formation is limited by PA28 expression. Therefore, upon maturation of DC, the levels of PA28 are strongly elevated, enabling formation of functional PA28 complexes.
Gene expression of PA28 and is differentially regulated
Our experiments indicate that PA28 up-regulation by LPS or CD40 stimulation occurs in an IFN--independent fashion. We investigated transcriptional activity of both genes after LPS and CD40 stimulation and IFN- treatment of D1 cells. Fig. 5C shows that the levels of PA28 and transcripts are increased equally after IFN- treatment during 6 and 18 h. This is in line with the presence of an IFN-stimulated response element in the promoters of both PA28 and (46). In contrast, LPS and CD40 treatment strongly increase transcriptional activity of the PA28 gene already after 6 h (Fig. 5, A and B), whereas the level of PA28 transcripts was increased only marginally after 18 h. These data indicate that the transcription of both genes is regulated differentially. Therefore, we examined the published sequences of the promoter regions of both genes (46). It struck us that an important difference between the regulatory sequences of both genes was the presence of a NF-B binding site in the PA28 promoter sequence that was absent from the PA28 promoter sequence. As already shown before, LPS activation of D1 cells leads to increased levels of NF-B (47). We have confirmed this by band shift analysis of a known HLA class I oligonucleotide NF-B-binding sequence with nuclear extracts of LPS- and CD40-activated D1 cells in comparison with extracts of immature D1 (data not shown). To investigate the binding of NF-B to this putative NF-B site in the murine PA28 promoter sequence, we synthesized an oligonucleotide probe of this sequence and performed a band shift assay using nuclear extracts of immature and matured D1 cells (Fig. 6). This experiment shows that the PA28 B sequence binds protein complexes present in these extracts. Significantly more protein is bound in extracts from LPS- and CD40-stimulated D1 cells. The observed bands can be supershifted by addition of RelA-, RelB-, and p50-specific Abs, showing that the sequence indeed is a NF-B binding site. These data indicate that expression of the PA28 and PA28 genes is regulated differentially via the NF-B-signaling cascade and thus that PA28 gene transcription can be enhanced independently by LPS- or CD40-mediated maturation of D1 DC.
Discussion
In the present study, we show that maturation of DC can influence the processing of CTL epitopes by elevated expression of the proteasome activator. This is achieved independently of the action of the classical immunomodulatory cytokine IFN-. DC maturation, either in a Th cell-dependent fashion mimicked by the CD40 agonistic Ab or in an innate-type fashion via TLR triggers, results in increased functional expression of PA28 complexes. This indicates that mature DC acquire the capacity to increase their peptide quantities, as generated by the proteasome, for presentation in MHC class I molecules. Thus far, the function of the PA28 complex is thought to involve an increase in the activity of the 20S proteasome rather than a qualitative change in the function or fine specificity of the catalytic proteasome core. Although it cannot be fully excluded that interaction of the PA28 cap with the 20S core could also affect the individual activity of the proteolytic sites (a qualitative change in the proteolytic activity), the quantitative effect of PA28 on cleavage activity has been well described previously (26, 33, 48). The recent cocrystalization structure of PA26 (the trypanosome homologue of PA28) with the 20S catalytic core shows a conformational change in the subunits that leads to an opening of the exit gate of the proteasome (31). This structure appears to support a role of PA28 in facilitating the release of peptide products and thereby a quantitative increase in the generation of relevant peptide lengths. Therefore, we interpret our findings in DC as mainly a quantitative effect on the availability of MHC class I-presented peptides. This is an important physiological event in APC acquiring the capacity for efficient priming of CTL by improving the expression of TCR ligands, namely MHC class I molecules presenting the appropriate peptide epitopes. Several reports (2, 4), including our own (3, 5), have shown that Th-dependent and Th-independent DC maturation signals are essential for acquisition of effective CTL priming capacity. So far, this is mainly explained by the elevated expression of MHC molecules, as well as relevant costimulatory molecules and crucial cytokines. In the present study, we show that DC maturation is not only associated with improved costimulation and MHC expression but also with improved MHC class I processing via up-regulation of functional PA28 regulator complexes.
Our findings show that PA28 complex formation in mature DC is limited by the low expression of the PA28 molecule. This conclusion is based on several lines of evidence. First, Western blot analysis of total D1 cell lysates shows significantly more induction of PA28 than of PA28, upon DC maturation. Although this method is only a qualitative approach and has limitations caused by potential differences in affinity of the Abs used, it is clear from the glycerol gradient analysis that relatively more functional PA28 complexes are generated as a result of the presence of the PA28 subunit. In addition to the protein data, independent quantitative analysis of mRNA transcripts of both molecules indicates that PA28 is indeed more strongly up-regulated than PA28 upon maturation of DC. These findings are in agreement with the proteasome composition analysis in human monocyte-derived DC; higher levels of PA28 transcripts than of PA28 can be observed in the data shown (20). In contrast, classical IFN--induced transcription of both genes does not differ with respect to the level of up-regulation. Triggering of the CD40 molecule, as a member of the TNFR family, as well as LPS activation, mediated via TLR4 (49) or CpG via TLR9 (50) are known to signal via an NF-B-mediated cascade. The presence of a NF-B binding site in the PA28 promoter sequence and its absence in the corresponding PA28 sequence strongly suggests a differential mechanism of expression of both genes. In this study, we show the functionality of the PA28 promoter NF-B-binding sequence for its binding capacity to NF-B protein complexes present in nuclear extracts of LPS- and CD40-matured DC. Taken together, these data strongly indicate a PA28-mediated mechanism for controlled elevation of the PA28 complex in maturing DC. This quantitative rather than qualitative mode of MHC class I-binding ligand regulation is further emphasized by the lack of change in the subunits of the 20S proteasome upon maturation. Immature DC already express the immunosubunits LMP-2, LMP-7, and MECL-1 at high constitutive levels. Therefore, a qualitative change in the fine specificity of the proteasome upon DC maturation, although reported for human monocyte-derived DC (19), is not supported by our findings.
Recent analysis of Ag presentation in PA28-deficient mice show that hsp90 could compensate for enhancing presentation of the OVA epitope (51). This suggests that other processing pathways could also play a role for certain Ags. The role of hsp90 in DC will be important to study.
The sequence of events in DC following Ag uptake, allowing cognate interactions with specific lymphocytes to occur, is apparently tightly regulated both at a cellular level and in time. It appears that MHC class II presentation occurs relatively rapidly after DC activation. Within a few hours, pre-existing intracellular MHC compartments fuse with the cell membrane, leading to a strong increase of MHC class II cell surface expression (37, 52, 53, 54). This allows specific Th-DC interactions to occur. The resulting Th-dependent signals will further mature the DC by expression of costimulatory molecules, and the MHC class I-processing machinery will be optimally engaged in the ensuing time period. In agreement with this model, we have observed recently sequential waves of Ag-specific CD4+ and CD8+ T cells in the regression of retrovirus-induced sarcoma (55). In conclusion, DC use a series of mechanisms for optimal priming of peptide-specific CTL. The peptide-epitopes are generated by a well-controlled processing system allowing efficient MHC class I presentation and thereby induction of specific CTL-mediated immunity.
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 Netherlands Cancer Foundation Grant UL 2002-2724 and European Community Grants QLRT 1999-00064, QLK2-CT-2000-00470, and QLK3-CT-2001-00093.
2 F.O. and N.F. contributed equally to this work.
3 Address correspondence and reprint requests to Dr. Ferry Ossendorp, Department of Immunohematology and Blood Transfusion, Leiden University Medical Center, Leiden, The Netherlands. E-mail address: f.a.ossendorp{at}lumc.nl
4 Abbreviations used in this paper: DC, dendritic cell; BMDC, bone-marrow-derived DC; IC, immune complex; TRP-2, tyrosinase-related protein 2; MEC, mouse embryo cell.
Received for publication July 15, 2004. Accepted for publication April 12, 2005.
References
van Bergen, J., F. Ossendorp, R. Jordens, A. M. Mommaas, J. W. Drijfhout, F. Koning. 1999. Get into the groove! Targeting antigens to MHC class II. Immunol. Rev. 172: 87-96.
Bennett, S. R., F. R. Carbone, F. Karamalis, R. A. Flavell, J. F. A. P. Miller, W. R. Heath. 1998. Help for cytotoxic-T cell responses is mediated by CD40 signaling. Nature 393: 478-480.
Schoenberger, S. P., R. E. Toes, E. I. van der Voort, R. Offringa, C. J. Melief. 1998. T cell help for cytotoxic T-lymphocytes is mediated by CD40-CD40L interactions. Nature 393: 480-483.
Ridge, J. P., F. Di Rosa, P. Matzinger. 1998. A conditioned dendritic cell can be a temporal bridge between a CD4+ T-helper and a T-killer cell. Nature 393: 474-478.
Schuurhuis, D. H., S. Laban, R. E. Toes, P. Ricciardi-Castagnoli, M. J. Kleijmeer, E. I. van der Voort, D. Rea, R. Offringa, H. J. Geuze, C. J. Melief, F. Ossendorp. 2000. Immature dendritic cells acquire CD8+ cytotoxic T lymphocyte priming capacity upon activation by T helper cell-independent or -dependent stimuli. J. Exp. Med. 192: 145-150.
Kaisho, T., O. Takeuchi, T. Kawai, K. Hoshino, S. Akira. 2001. Endotoxin-induced maturation of myd88-deficient dendritic cells. J. Immunol. 166: 5688-5694.
Sparwasser, T., E. S. Koch, R. M. Vabulas, K. Heeg, G. B. Lipford, J. W. Ellwart, H. Wagner. 1998. Bacterial DNA and immunostimulatory CpG oligonucleotides trigger maturation and activation of murine dendritic cells. Eur. J. Immunol. 28: 2045-2054.(Ferry Ossendorp2,3,*, Nat)
Activation of dendritic cells (DC) by Th-dependent (CD40) or -independent (LPS, CpG, or immune complexes) agonistic stimuli strongly enhances the expression of the proteasome activator PA28 complex. Upon activation of DC, increased MHC class I presentation occurred of the melanocyte-associated epitope tyrosinase-related protein 2180-188 in a PA28-dependent manner. In contrast to other cell types, regulation of PA28 expression in DC after maturation was found to be IFN- independent. In the present study, we show that expression of PA28 and subunits was differentially regulated. Firstly, PA28 expression is high in both immature and mature DC. In contrast, PA28 expression is low in immature DC and strongly increased in mature DC. Secondly, we show the presence of a functional NF-B site in the PA28 promoter, which is absent in the PA28 promoter, indicating regulation of PA28 expression by transcription factors of the NF-B family. In addition, glycerol gradient analysis of DC lysates revealed elevated PA28 complex formation upon maturation. Thus, induction of PA28 expression allows proper PA28 complex formation, thereby enhancing proteasome activity in activated DC. Therefore, maturation of DC not only improves costimulation but also MHC class I processing. This mechanism enhances the CD8+ CTL (cross)-priming capacity of mature DC.
Introduction
Dendritic cells (DC)4 are specialized APC that are able to present antigenic peptides in both MHC class II and I molecules for optimal priming of Ag-specific T lymphocytes. Efficient uptake of antigenic protein followed by processing to peptide fragments suited for presentation in MHC molecules are crucial functions of these cells. DC can efficiently present Ags in MHC class II molecules using the exogenous route of Ag uptake and degradation in the endosomal/lysosomal pathway (reviewed in Ref. 1). In contrast, the MHC class I presentation route in DC is less well understood but is in general dependent on proteasome activity, requiring antigenic proteins to enter the cytosol after exogenous uptake.
After Ag presentation in MHC class II molecules, specific Th cells can interact with DC to deliver signals that lead to an activated state of DC. This is an essential step for efficient priming of MHC class I-restricted CTL, which requires cognate interactions between CD40L (CD154) on Th cells and CD40 on DC (2, 3, 4). Th-independent (innate) signals such as bacterial LPS or CpG, triggering specific TLR, can also induce the DC maturation required for CTL priming (5, 6, 7, 8). This mature state is characterized by up-regulated expression of several cell surface molecules, including the costimulatory molecules CD80 and CD86 and MHC class II and class I. A recent report shows that triggering of TLR on bone marrow-derived DC (BMDC) can improve cross-presentation of protein-derived antigenic peptides in MHC class I (9). Delamarre et al. (10) showed that presentation of exogenous Ags on MHC class I and class II molecules is differentially regulated during DC maturation, although a molecular explanation for this was not provided.
The immunomodulatory cytokine IFN-, produced during an immune response by activated Th 1 lymphocytes, CD8+ CTL, and NK cells, enhances Ag presentation by up-regulation of MHC and TAP gene products, as well as proteasome subunits and regulators (11). IFN- alters proteasome activity qualitatively by induction of three newly synthesized proteasome subunits LMP2, LMP7, and MECL-1, which replace the three constitutive 1 (), 2 (MB1), and 5 (Z) subunits, respectively, in the core 20S proteasome. Several groups reported that these so-called immunosubunits are incorporated interdependently (12, 13, 14, 15). Taken together with the reported differences in fine specificity (16, 17, 18), these findings imply that two types of proteasomes exist: "constitutive" proteasomes, which are expressed in all somatic cells, and "immunoproteasomes," which are expressed under influence of immunomodulatory cytokines such as IFN-. However, immunoproteasomes can be constitutively expressed in some cells, in particular, lymphoid cells and DC (19, 20). In vitro studies showed that proteasomes equipped with LMP2, LMP7, and MECL-1 are more efficient in the generation of several antigenic viral peptides (21). Mice that were made deficient for LMP2 (22) or LMP7 (23) can be regarded as immunoproteasome-deficient mice. These mice showed impairment in the presentation of some Ags, indicating that immunoproteasomes are indeed involved in generation of immunologically relevant antigenic peptides in vivo. Recently, LMP2-deficient mice have revealed that immunoproteasomes can also affect shaping of the CTL repertoire (24).
In addition to proteasome immunosubunits, IFN- also regulates the expression of the proteasome-associated complex PA28 (REG or the 11S regulator). PA28 was identified as a proteasome activator that strongly increases the maximal velocity of the hydrolytic reaction and decreases the concentration of substrate required for cleavage by purified 20S proteasomes (25, 26). Isolated PA28 does not display enzymatic activity by itself. Three homologous PA28 subunits have been cloned thus far, of which, only the and subunits are strongly induced by IFN- (27). Protein chemical analyses have revealed that the and subunits form a tightly connected heterohexameric or heteroheptameric complex, composed of comparable amounts of both subunits (28). The PA28 heteromultimer is able to bind to the rings of the 20S core proteasome, creating bell-shaped cap structures on both ends (11) or on one end, combined with the PA700 (or 19S) cap (the ubiquitin-dependent targeting/unfolding ATPase complex) on the other end (29, 30), to form the so-called "hybrid proteasome." The mechanism of PA28-mediated enhancement of intrinsic proteasome activity is not fully clear, although the cocrystalization structure of PA26 (the trypanosome homologue of PA28) with the 20S catalytic core shows a conformational change in the subunits that leads to an opening of the exit gate of the proteasome (31).
Both in vitro and in vivo studies have provided evidence that PA28 was implicated in MHC class I Ag processing. In vitro digestion studies with long peptides comprising dominant CTL epitopes showed that purified PA28 can enhance cleavage by the 20S proteasome, leading to augmented epitope liberation (32). Furthermore, expression of PA28 in mouse fibroblasts increased the sensitivity for lysis by virus-specific CTL (33). We have shown that upon controlled expression of PA28 in a tetracycline-dependent system, functional presentation of some Moloney leukemia virus CTL epitopes was increased significantly (34). The absence of PA28 in vivo reduces epitope processing and CTL recognition of some Ags, as shown in PA28-deficient mice (35). Furthermore, PA28 expression in melanoma cells could rescue processing and presentation of a CTL epitope from the tyrosinase-related protein 2 (TRP-2) Ag that was normally not presented in these tumor cells (36).
Macagno et al. (19, 20) have shown in human DC that PA28 and expression was elevated after DC maturation. These reports show that PA28 expression is independent of assembly of constitutive and immunoproteasomes. However, the expression regulation of these PA28 subunits in maturing DC was not addressed.
So far, IFN- appears to be the main regulator of the processing and presentation pathway. We have now investigated the role of DC activation via Th-dependent (CD40) and Th-independent signals (LPS) on the MHC class I-processing pathway. For this study, we have used the well-characterized mouse DC cell line D1 (37). This cell line is a growth factor-dependent spleen-derived DC cell line that, upon appropriate culture, maintains a fully immature state associated with efficient Ag uptake capacity. When activated, it matures both phenotypically and functionally (5). We have observed that such activation is required for efficient MHC class I presentation of a viral CTL epitope or for presentation of the SIINFEKL epitope from the OVA protein as delivered to DC via Ab-Ag immune complexes (IC) (38). This way of cross-presentation in MHC class I is FcR and proteasome dependent (39). DC that have been pulsed with protein Ag via IC very efficiently cross-prime CD8+ CTL responses in vivo (38). Interestingly, Pierre and coworkers (40, 41) have shown the presence of ubiquitin aggregates (so-called DALIS) in mature DC, suggesting a role of proteasome activity in direct MHC class I presentation.
Our detailed analysis of the changes in the molecular machinery of MHC class I Ag processing during maturation of DC shows enhanced expression of the PA28-chain of the PA28 heteromeric, proteasome-associated complex. Both LPS- and CD40-mediated stimulation strongly induced RNA and protein expression of PA28, independently of IFN-. The induced PA28 complexes were functionally active in mature D1 DC. In immature cells, the assembly of PA28 complexes appeared to be limited by the low expression of PA28, whereas PA28 expression was already high. This study indicates differentially regulated gene expression of these two molecules.
These findings are compatible with the "license to kill model" in which maturation of DC is a crucial step for initiation of effective CTL immune responses. Our data show that maturation of DC not only improves "signal two" but also enhances "signal one."
Materials and Methods
Nuclear extracts of 10 x 106 nonstimulated and stimulated D1 cells were prepared as described previously (45). The total amount of protein was determined using the BCA Protein Assay kit (Pierce), according to the manufacturer’s instructions.
EMSA was performed by incubating nuclear extracts (5 μg of protein) in DNA protein-binding buffer (10 mM HEPES (pH 7.9), 60 mM KCl, 10% v/v glycerol, 1 mM DTT, 1 mM EDTA, 3 mM MgCl2, and 10 mM NaPi), with 200 ng of poly(dI:dC), 200 ng of sonicated single-stranded herring sperm DNA, and 1 ng of 32P-labeled probe for 30 min at 4°C. The samples were run on a 6% nondenaturing gel in 0.25x Tris-borate EDTA buffer at 200 V for 2 h. The gels were fixed with a 10% methanol, 10% acetic acid solution, dried onto Whatmann 3M paper, and exposed to a x-ray film. The ds-oligonucleotide encompassing the potential NF-B site within the promoter sequence of the murine PA28 gene (5'-TCCCCGGCGGGGAAGTCCCACCTTATC-3') was used as a probe for the EMSA (46). A known NF-B-binding sequence from HLA class I (45) was used as competitor probe.
For supershift assays, 1 μg of each Ab directed against members of the NF-B/Rel family of transcription factors was added to the nuclear extract and probe mixture and incubated for 1 h at 4°C. The Abs used were RelB (sc-226x), RelA (sc-109x), and p50 (sc-114x; Santa Cruz Biotechnology).
Results
DC maturation enhances CTL recognition in a PA28-dependent fashion
The involvement of PA28 in CTL epitope processing in maturing DC was tested by measuring the presentation of the PA28-dependent CTL epitope 180-188 of the melanoma Ag TRP-2. We prepared BMDC cultures from PA28-deficient mice (PA28KO) (35) and wild-type C57BL/6 (B6) mice. These DC were either left immature or treated with LPS for 48 h to induce maturation as indicated by elevated expression of CD86, CD40, and MHC class II and class I (data not shown). After 48 h, the cells were incubated with a processing-dependent synthetic peptide 180-193 harboring the TRP-2 epitope and a 5-aa C-terminal extension. CTL epitope presentation was monitored with a TRP-2-specific CTL clone (Fig. 1A). The data show that both B6 and PA28KO immature DC are very inefficient in processing and presentation of the TRP-2 epitope, whereas LPS-matured B6 DC showed very efficient presentation. In contrast, PA28KO DC showed strongly reduced capacity of CTL epitope presentation, indicating the direct involvement of PA28 in the efficiency of processing of this epitope. The same cells were incubated with the processing-independent, minimal 9-mer peptide, which was equally well presented to the CTL by all types of DC (Fig. 1B).
On the basis of these results, we performed an in depth molecular analysis of the PA28-proteasome-dependent Ag MHC class I-processing pathway in DC.
Constitutive high immunoproteasome expression in immature and mature DC
The well-defined murine DC line D1 develops functional maturation upon stimulation with bacterial LPS or CD40 ligation by agonistic mAb as evidenced by increased CD86, MHC class II and class I cell surface expression, and IL-12 production (5). Because of its homogeneity, this cell line is an excellent tool for biochemical studies. The expression of the proteasome immunosubunits LMP2, LMP7, and MECL-1 was determined in total cell lysates by Western blot analysis. LMP7 was already strongly expressed by immature D1 cells and was not detectably up-regulated by IFN-, LPS, or anti-CD40 treatment (Fig. 2A, upper panel). LMP2 and MECL-1 analysis showed similar results (data not shown). Also, the lymphoid cell lines RMA (T cell lymphoma) and 771 (B cell lymphoma; data not shown) expressed high basal levels of LMP7. In contrast, fibroblast-type MEC did not show detectable expression of LMP7 (Fig. 2A). In MEC, LMP7 was up-regulated by IFN- treatment but failed to be induced by LPS.
These results indicate that immature DC already possess high constitutive expression levels of immunoproteasomes that are only marginally enhanced upon DC activation.
PA28 expression is significantly increased after DC maturation
In the same Western blot experiments, we observed that the expression levels of PA28 and PA28 were markedly enhanced upon IFN- treatment but also following LPS and CD40 stimulation of D1 cells (Fig. 2A, lower panels). PA28 is already expressed at relatively high levels in resting DC and is only marginally increased after the different treatments. In contrast to this, PA28 expression is very low in immature DC and strongly enhanced after DC maturation. Next, we analyzed the expression kinetics of PA28 and PA28 after LPS or CD40 stimulation (Fig. 2B). This experiment shows that PA28 expression is already detectably increased after 24 h of LPS or anti-CD40 treatment and reaches strongly elevated levels after 72 h of activation. In contrast, PA28 expression level is constitutively high in nonstimulated cells and is only moderately increased upon stimulation. These results suggest a differential regulation of PA28 and gene expression in DC. Similar results were obtained with freshly isolated BMDC cultures that were matured with LPS or the TLR9 ligand CpG (Fig. 2C). The kinetics and differential up-regulation of protein expression of PA28 and in these primary DC is similar to that of D1 cells, showing that D1 is a representative DC line.
Previously, we have shown that FcR-mediated activation and FcR-facilitated Ag uptake strongly improved MHC class I presentation and CTL cross-priming (38). In the present study, we have analyzed the ability of Ab-Ag IC to induce PA28 expression in DC. Fig. 2D shows strong induction of PA28 in both D1 cell line and freshly isolated BMDC. As a control, PA28 expression was absent in PA28-deficient BMDC.
In comparison to DC, the expression levels of PA28 are intermediate in the lymphoid cell line RMA and low in MEC (Fig. 2A). The expression level of PA28 was low in RMA and almost undetectable in MEC. LPS or CD40 stimulation did not significantly influence PA28 or expression in these (non-DC) cell lines. However, IFN- could induce expression of both PA28 molecules in MEC fibroblast cells.
These data suggest that TLR- or CD40-induced expression of the (classically IFN--inducible) PA28 molecules in DC is independent of IFN-. This is supported by the fact that D1 DC do not produce detectable amounts of IFN-, not even after activation by LPS or via CD40 (determined by specific ELISA; data not shown). To exclude the possibility of a role of secreted IFN- by D1 cells, LPS or CD40 stimulation was also performed in the presence of neutralizing anti-IFN- Abs. This Ab could block the expression of PA28 and as induced by exogenously added IFN- but not the induction of expression of PA28 and by LPS or CD40 stimulation (data not shown). Importantly, BMDC of IFN-R-deficient mice were analyzed for PA28 and expression after maturation with LPS. Fig. 2E shows similarly strong induction of the proteasome activator subunits, indicating IFN- independence.
PA28 and molecules in mature DC are complexed and functional
PA28 and molecules are highly expressed in mature DC. To test whether these molecules are present in functional PA28 complexes, total lysates were subjected to glycerol gradient analysis. Under these conditions, PA28 complexes remain intact and functional. Fig. 3A shows the result of a Western blot analysis of the individual fractions of the glycerol gradients of immature D1 cells and CD40-matured D1 cells. Early fractions 3–5 contain the monomeric PA28 molecules, and fractions 7–10 contain the complexed PA28 molecules. The presence of PA28 is shown in Fig. 3A, left panel. This analysis shows that the total level of PA28 does not increase significantly upon DC activation by anti-CD40 Ab, but rather shifts from the monomeric form to the complexed form. In contrast, PA28 expression (Fig. 3A, right panel) strongly increases following DC activation, and virtually all of the expressed PA28 is then present in the complexed form (fractions 7–10). Fig. 3B shows a similar experiment with LPS-stimulated D1 cells. Although the shift of monomeric to complexed PA28 is less striking, the increased expression of PA28 is apparent. The increased PA28 complex formation is similar to that seen following CD40 stimulation of D1 and again appears to be limited by PA28 expression.
These complexes were tested for their functional activity by incubation of the individual glycerol gradient fractions with purified 20S proteasomes, allowing evaluation of the enhancing effect of the PA28 proteasome activator. The cleavage of a fluorigenic synthetic peptide substrate was measured as shown in Fig. 4A. This figure shows a strong increase in proteolytic activity in fractions 7–10 of both the lysates of CD40- and LPS-matured D1 cells in comparison to the same fractions of immature D1 cells. The basal activity of these fractions, in the absence of added purified 20S proteasomes, is low (Fig. 4B). Only in the later fractions (13, 14, 15, 16, 17) endogenous proteolytic activity can be measured. The early glycerol fractions 3–5, containing monomeric PA28 and subunits, do not exhibit strong functional activity.
These data indicate that PA28 is already expressed at high, saturated levels in immature D1 cells and that functional PA28 complex formation is limited by PA28 expression. Therefore, upon maturation of DC, the levels of PA28 are strongly elevated, enabling formation of functional PA28 complexes.
Gene expression of PA28 and is differentially regulated
Our experiments indicate that PA28 up-regulation by LPS or CD40 stimulation occurs in an IFN--independent fashion. We investigated transcriptional activity of both genes after LPS and CD40 stimulation and IFN- treatment of D1 cells. Fig. 5C shows that the levels of PA28 and transcripts are increased equally after IFN- treatment during 6 and 18 h. This is in line with the presence of an IFN-stimulated response element in the promoters of both PA28 and (46). In contrast, LPS and CD40 treatment strongly increase transcriptional activity of the PA28 gene already after 6 h (Fig. 5, A and B), whereas the level of PA28 transcripts was increased only marginally after 18 h. These data indicate that the transcription of both genes is regulated differentially. Therefore, we examined the published sequences of the promoter regions of both genes (46). It struck us that an important difference between the regulatory sequences of both genes was the presence of a NF-B binding site in the PA28 promoter sequence that was absent from the PA28 promoter sequence. As already shown before, LPS activation of D1 cells leads to increased levels of NF-B (47). We have confirmed this by band shift analysis of a known HLA class I oligonucleotide NF-B-binding sequence with nuclear extracts of LPS- and CD40-activated D1 cells in comparison with extracts of immature D1 (data not shown). To investigate the binding of NF-B to this putative NF-B site in the murine PA28 promoter sequence, we synthesized an oligonucleotide probe of this sequence and performed a band shift assay using nuclear extracts of immature and matured D1 cells (Fig. 6). This experiment shows that the PA28 B sequence binds protein complexes present in these extracts. Significantly more protein is bound in extracts from LPS- and CD40-stimulated D1 cells. The observed bands can be supershifted by addition of RelA-, RelB-, and p50-specific Abs, showing that the sequence indeed is a NF-B binding site. These data indicate that expression of the PA28 and PA28 genes is regulated differentially via the NF-B-signaling cascade and thus that PA28 gene transcription can be enhanced independently by LPS- or CD40-mediated maturation of D1 DC.
Discussion
In the present study, we show that maturation of DC can influence the processing of CTL epitopes by elevated expression of the proteasome activator. This is achieved independently of the action of the classical immunomodulatory cytokine IFN-. DC maturation, either in a Th cell-dependent fashion mimicked by the CD40 agonistic Ab or in an innate-type fashion via TLR triggers, results in increased functional expression of PA28 complexes. This indicates that mature DC acquire the capacity to increase their peptide quantities, as generated by the proteasome, for presentation in MHC class I molecules. Thus far, the function of the PA28 complex is thought to involve an increase in the activity of the 20S proteasome rather than a qualitative change in the function or fine specificity of the catalytic proteasome core. Although it cannot be fully excluded that interaction of the PA28 cap with the 20S core could also affect the individual activity of the proteolytic sites (a qualitative change in the proteolytic activity), the quantitative effect of PA28 on cleavage activity has been well described previously (26, 33, 48). The recent cocrystalization structure of PA26 (the trypanosome homologue of PA28) with the 20S catalytic core shows a conformational change in the subunits that leads to an opening of the exit gate of the proteasome (31). This structure appears to support a role of PA28 in facilitating the release of peptide products and thereby a quantitative increase in the generation of relevant peptide lengths. Therefore, we interpret our findings in DC as mainly a quantitative effect on the availability of MHC class I-presented peptides. This is an important physiological event in APC acquiring the capacity for efficient priming of CTL by improving the expression of TCR ligands, namely MHC class I molecules presenting the appropriate peptide epitopes. Several reports (2, 4), including our own (3, 5), have shown that Th-dependent and Th-independent DC maturation signals are essential for acquisition of effective CTL priming capacity. So far, this is mainly explained by the elevated expression of MHC molecules, as well as relevant costimulatory molecules and crucial cytokines. In the present study, we show that DC maturation is not only associated with improved costimulation and MHC expression but also with improved MHC class I processing via up-regulation of functional PA28 regulator complexes.
Our findings show that PA28 complex formation in mature DC is limited by the low expression of the PA28 molecule. This conclusion is based on several lines of evidence. First, Western blot analysis of total D1 cell lysates shows significantly more induction of PA28 than of PA28, upon DC maturation. Although this method is only a qualitative approach and has limitations caused by potential differences in affinity of the Abs used, it is clear from the glycerol gradient analysis that relatively more functional PA28 complexes are generated as a result of the presence of the PA28 subunit. In addition to the protein data, independent quantitative analysis of mRNA transcripts of both molecules indicates that PA28 is indeed more strongly up-regulated than PA28 upon maturation of DC. These findings are in agreement with the proteasome composition analysis in human monocyte-derived DC; higher levels of PA28 transcripts than of PA28 can be observed in the data shown (20). In contrast, classical IFN--induced transcription of both genes does not differ with respect to the level of up-regulation. Triggering of the CD40 molecule, as a member of the TNFR family, as well as LPS activation, mediated via TLR4 (49) or CpG via TLR9 (50) are known to signal via an NF-B-mediated cascade. The presence of a NF-B binding site in the PA28 promoter sequence and its absence in the corresponding PA28 sequence strongly suggests a differential mechanism of expression of both genes. In this study, we show the functionality of the PA28 promoter NF-B-binding sequence for its binding capacity to NF-B protein complexes present in nuclear extracts of LPS- and CD40-matured DC. Taken together, these data strongly indicate a PA28-mediated mechanism for controlled elevation of the PA28 complex in maturing DC. This quantitative rather than qualitative mode of MHC class I-binding ligand regulation is further emphasized by the lack of change in the subunits of the 20S proteasome upon maturation. Immature DC already express the immunosubunits LMP-2, LMP-7, and MECL-1 at high constitutive levels. Therefore, a qualitative change in the fine specificity of the proteasome upon DC maturation, although reported for human monocyte-derived DC (19), is not supported by our findings.
Recent analysis of Ag presentation in PA28-deficient mice show that hsp90 could compensate for enhancing presentation of the OVA epitope (51). This suggests that other processing pathways could also play a role for certain Ags. The role of hsp90 in DC will be important to study.
The sequence of events in DC following Ag uptake, allowing cognate interactions with specific lymphocytes to occur, is apparently tightly regulated both at a cellular level and in time. It appears that MHC class II presentation occurs relatively rapidly after DC activation. Within a few hours, pre-existing intracellular MHC compartments fuse with the cell membrane, leading to a strong increase of MHC class II cell surface expression (37, 52, 53, 54). This allows specific Th-DC interactions to occur. The resulting Th-dependent signals will further mature the DC by expression of costimulatory molecules, and the MHC class I-processing machinery will be optimally engaged in the ensuing time period. In agreement with this model, we have observed recently sequential waves of Ag-specific CD4+ and CD8+ T cells in the regression of retrovirus-induced sarcoma (55). In conclusion, DC use a series of mechanisms for optimal priming of peptide-specific CTL. The peptide-epitopes are generated by a well-controlled processing system allowing efficient MHC class I presentation and thereby induction of specific CTL-mediated immunity.
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 Netherlands Cancer Foundation Grant UL 2002-2724 and European Community Grants QLRT 1999-00064, QLK2-CT-2000-00470, and QLK3-CT-2001-00093.
2 F.O. and N.F. contributed equally to this work.
3 Address correspondence and reprint requests to Dr. Ferry Ossendorp, Department of Immunohematology and Blood Transfusion, Leiden University Medical Center, Leiden, The Netherlands. E-mail address: f.a.ossendorp{at}lumc.nl
4 Abbreviations used in this paper: DC, dendritic cell; BMDC, bone-marrow-derived DC; IC, immune complex; TRP-2, tyrosinase-related protein 2; MEC, mouse embryo cell.
Received for publication July 15, 2004. Accepted for publication April 12, 2005.
References
van Bergen, J., F. Ossendorp, R. Jordens, A. M. Mommaas, J. W. Drijfhout, F. Koning. 1999. Get into the groove! Targeting antigens to MHC class II. Immunol. Rev. 172: 87-96.
Bennett, S. R., F. R. Carbone, F. Karamalis, R. A. Flavell, J. F. A. P. Miller, W. R. Heath. 1998. Help for cytotoxic-T cell responses is mediated by CD40 signaling. Nature 393: 478-480.
Schoenberger, S. P., R. E. Toes, E. I. van der Voort, R. Offringa, C. J. Melief. 1998. T cell help for cytotoxic T-lymphocytes is mediated by CD40-CD40L interactions. Nature 393: 480-483.
Ridge, J. P., F. Di Rosa, P. Matzinger. 1998. A conditioned dendritic cell can be a temporal bridge between a CD4+ T-helper and a T-killer cell. Nature 393: 474-478.
Schuurhuis, D. H., S. Laban, R. E. Toes, P. Ricciardi-Castagnoli, M. J. Kleijmeer, E. I. van der Voort, D. Rea, R. Offringa, H. J. Geuze, C. J. Melief, F. Ossendorp. 2000. Immature dendritic cells acquire CD8+ cytotoxic T lymphocyte priming capacity upon activation by T helper cell-independent or -dependent stimuli. J. Exp. Med. 192: 145-150.
Kaisho, T., O. Takeuchi, T. Kawai, K. Hoshino, S. Akira. 2001. Endotoxin-induced maturation of myd88-deficient dendritic cells. J. Immunol. 166: 5688-5694.
Sparwasser, T., E. S. Koch, R. M. Vabulas, K. Heeg, G. B. Lipford, J. W. Ellwart, H. Wagner. 1998. Bacterial DNA and immunostimulatory CpG oligonucleotides trigger maturation and activation of murine dendritic cells. Eur. J. Immunol. 28: 2045-2054.(Ferry Ossendorp2,3,*, Nat)