| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print April 7, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Laboratoire d’Immunologie Moléculaire, Département de Microbiologie et Immunologie, Université de Montréal, Québec, Canada;
Laboratoire d’Immunologie, Hôpital Saint-Luc, Centre de Recherche du Centre Hospitalier de l’Université de Montréal, Québec, Canada;
Division of Hematology and Immunodeficiency Service, Royal Victoria Hospital, McGill University Health Centre, Montréal, Québec, Canada;
Institut de recherche en biotechnologie, Montréal, Québec, Canada; and
¶ Laboratoire d’immunologie, Hôpital Notre-Dame, Centre hospitalier de l’Université de Montréal, Québec, Canada
1 Correspondence: Laboratoire d’Immunologie Moléculaire, Département de Microbiologie et Immunologie, Université de Montréal, CP 6128, Succursale Centre-Ville, Montréal, Québec, Canada, H3C 3J7. E-mail: Jacques.Thibodeau{at}umontreal.ca
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Transduction of DOß alone resulted in the accumulation of a small pool of DO in DM+ CD63+ vesicles and at the plasma membrane of mature DC. The cell-surface increase in class II-associated invariant chain peptide (CLIP)/class II complexes is in line with an inhibitory role of DO on DM. Cotransduction of DO and DOß only slightly increased CLIP and DO levels at the cell surface. Together with the fact that a large fraction of transduced DO remains in the endoplasmic reticulum, this suggests that DM is limiting in these conditions. DO expression did not affect a mixed lymphocyte reaction but reduced presentation of the exogenous gp100 antigen to a specific T cell clone. These results show that transduced DO modulates antigen presentation in human mature MoDC, evoking the possible use of this chaperone for immunotherapy.
Key Words: HLA-DR • HLA-DM • MHC • CLIP • invariant chain
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and ß chains assemble in the endoplasmic reticulum (ER) and form a complex with the invariant chain (Ii) before being targeted into the endocytic pathway [3
]. In endosomes, Ii is degraded until only a class II-associated Ii fragment (CLIP) remains in the peptide-binding groove [4
]. Humal leukocyte antigen (HLA)-DM (DM) is a nonclassical class II molecule residing in endosomal/lysosomal compartments [5
, 6
]. It catalyzes the dissociation of CLIP and is consequently required for efficient peptide loading of class II molecules [7
, 8
]. Finally, the class II molecules are transported to the cell surface, where they serve as ligands for the T cell receptors (TcRs) on T cells.
B lymphocytes express still another chaperone, HLA-DO (DO; H2-O in the mouse) [9
, 10
]. This intracellular, nonclassical class II molecule needs to associate with DM to egress the ER and to be transported to the endosomes [11
12
13
14
]. DO was first described as a negative modulator of DM activity, as its overexpression in various cell lines increased the amount of CLIP-associated MHC class II molecules at the cell surface [15
16
17
]. The influence of DO on the peptide repertoire would come, at least in part, from its ability to inhibit DM and CLIP release in early endocytic compartments but not at the low pH of late endosomes [18
19
20
]. Experiments by the group of Karlsson [18
] using H2-O-deficient mice suggested that this nonclassical class II molecule can inhibit the presentation of some antigens taken up by fluid-phase endocytosis and processed in early endosomes. Later, the same group reported that the effect of H2-O was even more drastic on antigens taken up by the specific surface immunoglobulin (sIg) receptor. It is surprising that the presentation of some epitopes was increased in the presence of DO, suggesting that the nature of the antigen and/or the sIg will dictate DO dependency [21
]. Also, peptide repertoire analyses between DO+-transfected cells and their DO– counterpart confirmed that various antigens benefit from the presence of DO. At steady state, 
10% of the class II peptide repertoire was specific to DO+ cells [17
]. Accordingly, only a partial overlap was detected in the class II peptide pools eluted from H2-O–/– and +/+ animals [22
].
Given its demonstrated immunomodulatory properties and nonpolymorphic nature, DO might become a valuable tool to shape the class II peptide repertoire at the surface of APC. Its ability to increase CLIP display could also be exploited. Indeed, recently, the group of Kropshofer [23
] has suggested that mature dendritic cells (mDC) up-regulate the expression of CLIP-class II complexes to antagonize T helper cell type 1 (Th1) responses. However, DO expression is limited mostly to B cells and ill-defined subsets of thymic epithelial cells [12
, 24
, 25
]. As it is coregulated with classical class II genes, other cell types such as DC probably express only a small amount of DO. However, the DOB gene is regulated differently, and the absence of DOß precludes the inhibition of DM [12
, 24
, 26
27
28
].
Recently, the group of Denzin [29 ] modulated the activity of H2-M through the ectopic expression of human DO in transgenic mice DC. The potential activity of DO in human DC remains to be assessed, as important differences exist in the organization and function of the endocytic pathway between various species and even cell types. Professional APC finely tune their antigen-presentation machinery according to their maturation state and their highly specialized physiological function in the afferent or effector limb of the immune response [30 ]. Differences in the class II maturation pathway between B cells and DC have been documented (reviewed in ref. [31 ]). For example, in B cells, cross-linking of the BcR triggers within minutes the formation of vesicles inside MHC class II vesicles and the accumulation of MHC class II-Ii complexes in these remodeled compartments [32 , 33 ]. As for resting B cells, immature DC (iDC), found in peripheral tissues, are poor T cell stimulators. Although they capture antigens efficiently, peptide-loading is inefficient. This might explain the presence at the plasma membrane of empty class II molecules that can be loaded from the outside [34 ]. Murine iDC awaiting maturation signals retain class II inside lysosomes by blocking the processing of Ii. The regulation of antigen processing is different in human, as iDC efficiently recycle class II from the surface [35 ]. When DC receive danger signals, they migrate and transport antigens to secondary lymphoid organs, where they complete their maturation. mDC no longer capture antigens but proceed to load class II molecules with antigenic peptides and redistribute most MHC class II and costimulatory molecules to the cell surface for efficient T cell stimulation [36 , 37 ]. It is interesting that the DM molecules purified from mDC seem to have a reduced activity [23 ]. Altogether, these features highlight the cell type-specific control over antigen processing.
Other observations suggest that the cellular background could affect the function and fate of DO. The absence of DO was shown to affect the presentation of antigens differently, depending on the strain genetic background or the coexpression of various class II molecules [22 , 38 ]. Also, although the H2-Oß staining was negative on splenocytes from H2-Oa knockout (KO) mice, its pattern of expression was intriguingly unaltered in the thymus [18 ].
To gain further insights into the function of DO, we developed adenoviral vectors coding for the ß chain alone or the and ß chains. Our results demonstrate that transduced DO can effectively modulate antigen presentation in human monocyte-derived DC (MoDC).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
L243 (IgG2a) monoclonal antibody (mAb) binds a conformational epitope on DR [39
, 40
]. Anti-CD14 and -CD83 mAb were from Immunotech (Marseille, France). HKC5 mAb reacts against the cytoplasmic tail of DOß [41
]. MaP.DM1 (DM-specific; PharMingen, Oakville, ON) has been described previously [42
]. Mags.DO5 (IgG1) recognizes a conformational determinant of DO [43
]. Cer.CLIP (IgG1) is directed at the N-terminal segment of CLIP bound to class II molecules [42
] (PharMingen). H5C6 (IgG1) is specific to CD63 molecules found in late endosomes [Developmental Studies Hybridoma Bank, National Institute of Child Health and Human Development (NICHD), University of Iowa, Iowa City]. The polyclonal rabbit anticalnexin recognizes the C-terminal part of the protein (Stressgen, Victoria, BC). Goat anti-mouse IgG coupled to Alexa Fluor® 488, goat anti-rabbit IgG coupled to Alexa Fluor® 594, and streptavidin coupled to Alexa Fluor® 594 were from Molecular Probes (Eugene, OR). The rabbit antisera against the cytoplasmic tail of DO and DM have been described previously [13
]. Anti-actin (IgG1) is specific for the N-terminal of the molecule (Chemicon International, Temecula, CA). Peroxidase-coupled goat anti-rabbit and goat anti-mouse secondary antibodies were from Jackson ImmunoResearch Laboratories (Mississauga, ON).
Adenoviral vectors and DC infection
The empty adenoviral vector (Ad0) is a replication-deficient, recombinant adenovirus derived from serotype 5 with the deletion of E1 and E3 regions [44
]. AdDOß and AdDO encode the DOß chain and DO and DOß, respectively. The cDNAs were cloned in the pAdenovator CMV5(CuO), in which the transgenes are driven by the repressible cumate promoter CMV5(CuO) [45
]. Functional viruses were produced as described [46
]. Briefly, following recombination in bacteria, recombinant adenoviral DNAs were transfected in HEK 293 cells, which complement for the deleted E1 gene. The cell monolayers were then covered with agarose and incubated until formation of viral plaques. Viruses were eluted from agarose plugs, propagated by several rounds of 293 cell infections, and purified by double centrifugation on CsCl gradients. Virus aliquots were stored at –80°C until use.
For adenoviral transduction, DC were distributed in flat-bottom, 96-well plates (Costar Corning Inc., Corning, NY) at 2.5 x 105 cells/well in 50 µl RPMI-1640 medium to favor virus adsorption. Viruses were added at a multiplicity of infection (MOI) of 250 and incubated at 37°C in the presence of 5% CO2 for 3 h. Complete media supplemented with 2 ng/ml lipopolysaccharide (LPS; Sigma-Aldrich, Munich, Germany) were added on cells in a final volume of 200 µl and incubated for 2 days to allow complete maturation of DC.
Isolation and culture of DC
Peripheral blood mononuclear cells (PBMC) were purified from leukapheresis of healthy donors by Ficoll-PaqueTM PLUS gradient centrifugation (Amersham Biosciences, UK). PBMC were cryopreserved in RPMI medium containing 20% FBS and 10% dimethyl sulfoxide. Monocytes were isolated by adherence or positive selection. Similar results were obtained using cells prepared by these two methods. Briefly, PBMC were plated at 9 x 106 cells/well in six-well plates (Costar Corning Inc.) in RPMI-1640 medium without serum at 37°C in the presence of 5% CO2. After 1 h, repeated washings with phosphate-buffered saline (PBS) removed nonadherent cells. Alternatively, monocytes were purified using Easysep CD14+ cell separation kit (StemCell Technologies, Inc., Vancouver, Canada) and placed in six-well plates (Costar Corning Inc.) at 2 x 106 cells/well. To generate iDC, monocytes were cultured for 7 days in complete medium containing 1 000 U/ml human granulocyte macrophage-colony stimulating factor (GM-CSF; Cangene Corp., Winnipeg, MB) and 300 U/mL interleukin (IL)-4 (Sigma-Aldrich). Every 2 days, fresh, complete medium containing GM-CSF and IL-4 was added. To produce mDC, fresh, complete medium supplemented with 2 ng/ml LPS (Sigma-Aldrich) was added for 48 h on day 7.
Flow cytometry
Cells were stained for the presence of cell-surface DR, CD86, CD83, CD14, and CLIP using L243, anti-CD86, anti-CD83, anti-CD14, and Cer.CLIP mAb, respectively. For intracellular staining, cells were first permeabilized using saponin as described previously [13
]. We used a secondary goat anti-mouse IgG coupled to Alexa Fluor® 488 for surface and intracellular stainings. Cells were analyzed on a FACSCalibur® (Becton Dickinson, Mississauga, ON).
Western blotting
Cells were left untreated or transduced with Ad0, AdDOß, or AdDO. Cells were washed with PBS and lysed at 4°C for 30 min in lysis buffer (20 mM Tris, pH 7, and 150 mM NaCl) containing protease inhibitors (Roche Diagnostics, Laval, Quebec, Canada) and 2% Triton X-100. After centrifugation, supernatants were harvested and diluted in reducing buffer. One-half of each sample was boiled, and the other half was boiled and treated with endoglycosidase H (EndoH; New England Biolabs, Ltd., Mississauga, ON), according to the manufacturer’s protocols before sodium dodecyl sulfate-polyacrylamide gel electrophoresis. After protein transfer, the membranes (Amersham Biosciences) were blotted with the various antibodies. A secondary peroxidase-coupled antibody was added for 1 h before detection (chemiluminescence-blotting substrate, Roche Diagnostics).
Reverse transcriptase-polymerase chain reaction (RT-PCR) amplifications
Total RNA was extracted from HeLa cells and DC using Tri-Reagent isolation solution (Sigma-Aldrich) according to the manufacturer’s recommendations. cDNAs were prepared by RT with random nonamer primers and Moloney murine leukemia virus RT (Sigma-Aldrich). PCR was performed for 40 cycles with denaturation at 94°C for 30 s, annealing at 56°C for glyceraldehyde 3 phosphate-dehydrogenase (GAPDH) and DOß, or 52°C for DO for 45 s and elongation at 68°C for 2 min. Three pairs of oligonucleotide PCR primers targeting DO, DOß, and GAPDH genes were synthesized and the sequences were as follows: DO (5'-CCA CTC AAA GTC AGC ACA GCG-3', 5'-TGG TCC TGG GGT TCC AC-3'); DOß (5'-CAG TGA CTA CCT GAG CAT TTG-3', 5'-AAC CAG AGG TGA CAG TGT-3'); GAPDH (5'-TGA AGG TCG GAG TCA ACG GAT TTG GT-3', 5'-CAT GTG GGC CAT GAG GTC CAC CAC-3').
Fluorescence microscopy
Cells were permeabilized and stained as described previously [13
]. After centrifugation on microscope slides at 0.2–1 x 105/slides, cells were examined with a Leica TCS-SP1 confocal microscope using a x100 planapochromat objective.
Functional analysis
For mixed lymphocyte reactions (MLR), allogeneic T cells were purified from PBMC using Easysep CD4+ negative selection kit (StemCell Technologies, Inc., Vancouver, BC). Approximately 2.5 x 104 mDC transduced on day 7 with AdDO, AdDOß, or Ad0 were mixed with increasing numbers of CD4+ T cells in flat-bottom, 96-well plates in a final volume of 200 µl. After 5 days of incubation, 1 µCi [3H] thymidine (Amersham Biosciences) was added overnight, and its uptake was measured the day after.
Pulsing of DC and recognition by anti-gp100 T cell clone
Lysates from melanoma cell lines expressing gp100 (MelS gp100) or green fluorescent protein (GFP; MelS GFP) and the CD4+ T cell clone specific to a DRß1*0701-restricted gp100 epitope were prepared as described previously [47
]. MelS gp100 and MelS GFP were derived from a melanoma cell line retrovirally transduced with gp100 or GFP cDNAs [48
]. MoDC from a DRß1*0701 healthy donor were not transduced or transduced with AdDOß, AdDO, or Ad0 on day 6. On day 8, cells were plated at 2.5 x 104 cells/well in 96-well plates in 150 µl complete medium. MelS gp100 or control MelS GFP lysates were added for 10 h. Then, LPS was added overnight in the presence of MelS cell lysates. Then, 4 x 104 gp100-specific T cells were added for an additional 24 h. Supernatants were harvested, and interferon- 
(IFN-) was measured by enzyme-linked immunosorbent assay (ELISA).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
mRNA as compared with the constitutively expressed GAPDH control. Conversely, DC do not express any mRNA for DOß in this semiquantitative assay (Fig. 1B
, middle panel), in line with the reported lack of DO proteins in DC [28
, 51
].
Adenoviral vectors do not trigger full maturation of DC
The above-described results confirmed that the discordant regulation of the endogenous and ß chains precludes DO expression in DC. To express the DOß heterodimer in these cells, we used recombinant adenoviruses. These gene-transfer vectors have many advantages for immunotherapy [52
]. Above all, recombinant adenoviruses are not dependent on active cell division and are thus efficient in nondividing, primary cells such as DC. Additionally, high titers and high levels of gene expression can be achieved.
Two approaches were used to restore DO expression in DC (Fig. 2A
). Given the significant expression of the DOA gene (see Fig. 1
), we reasoned that one could complement the endogenous pool of the DO chain by transducing the DOß chain alone. However, to maximize expression of the heterodimer, we also constructed a cassette bearing the DO and DOß cDNAs separated by an IRES. The shuttle vectors were allowed to recombine with the adenoviral plasmid to produce recombinant AdDOß and AdDO viruses.
|
Transduction of DO in DC
iMoDC were transduced with the control Ad0, AdDOß, or AdDO. After a maturation period of 48 h in the presence of LPS, cell lysates were prepared, and the DO chain expression was analyzed by Western blotting. Lysates were also treated with EndoH to reveal the presence of resistant forms, indicating their passage through the Golgi (Fig. 3A
). As described previously, control HeLa cells transfected with DOß do not contain EndoH-resistant molecules, reflecting the efficient ER retention in the absence of DM [11
]. Indeed, upon coexpression of DM, both molecules end up in the endocytic pathway, and their carbohydrates become resistant to EndoH processing. Despite the presence of its mRNA (see Fig. 1
), the endogenous DO chain was undetectable in untransduced DC or those transduced with the control Ad0. This most likely results from the combination of a low level of expression and the absence of a ß-chain partner. It is interesting that transduction of DOß in these DC was not sufficient to allow the detection of DO, even upon overexposure of the blots (data not shown). Conversely, EndoH-resistant forms of DMß were detected in all conditions. These results suggest that low levels of endogenous DO proteins are present in DC. The possibility remains that our rabbit antiserum made against a peptide of the cytoplasmic tail does not bind the endogenous DO chain in DC. However, this is unlikely, as it recognizes the endogenous form found in primary B cells or in CIITA transfectants (ref. [47
] and Fig. 3A
). Also, the size of the cDNA fragment covering the cytoplasmic tail-coding region was similar in all cells tested (see Fig. 1
).
|
chain was detected on Western blots only upon transduction of both chains as part of the AdDO construction (Fig. 3A)
. A strong signal was detected, and although the majority of the molecules was EndoH-sensitive, the presence of a faint band comigrating with the untreated material suggests that a fraction has acquired complex sugars and has egress from the ER en route to the endocytic pathway (Fig. 3A
, top-right panel). The apparent overexpression of DM in cells transduced with AdDO is in line with the stabilization of H2-M in mouse DC expressing DO [29
]. However, the increase in DM expression was not so apparent in all experiments (data not shown). Altogether, these results indicate that AdDO can efficiently transduce DO in DC.
We next investigated the expression of DO by flow cytometry using mAb HKC5 specific for a linear epitope in the DOß cytoplasmic tail [41
]. Analysis of permeabilized cells revealed that the AdDOß and AdDO vectors are highly efficient in transducing the ß chain (Fig. 3B)
. No DOß chain was detected in mock cells transduced with the Ad0 vector, in agreement with the absence of DOß transcripts or proteins in DC (see Fig. 1
and refs. [28
, 51
]). From the HKC5 staining profile, we estimate that more than 90% of the DC have been transduced and express levels of DOß equal or superior to B cell lines such as Raji (data not shown). The strong expression of DOß was confirmed by Western blotting, and again, the presence of EndoH-resistant forms is almost undetectable (Fig. 3B
, right panel). Transduced cells were then analyzed by flow cytometry using the Mags.DO5 mAb, which is specific for DOß. It is interesting that there was little difference between the signals obtained for the DC transduced with DOß alone as compared with cells transduced with the ß chains together (Fig. 3B) . As this antibody recognizes DO molecules, this result indicates that the amount of the endogenous DO chain is nearly sufficient to saturate the available pool of free DOß. Finally, the fact that the Mags.DO5 profile is rather homogenous as compared with the wide distribution seen with HKC5 indicates that little DO is formed. This would be consistent with the fact that only a small proportion of DO acquired resistance to EndoH.
Intracellular distribution of transduced DO molecules
To ensure that DO molecules expressed in DC behave as in B cells, we determined their subcellular localization by confocal microscopy using HKC5 and Mags.DO5. DOß+ DC stained with HKC5 display a diffuse pattern that overlaps the one obtained for calnexin, an ER marker (Fig. 4A
, 4a
4b
4c
). This strong colocalization between calnexin and DOß supports the conclusion reached above that the transduced molecules were produced in vast excess. The ER signal is strong, but it is possible to detect weak colocalization with the CD63 marker of late endosomes (Fig. 4A
, 4d
4e
4f
). The first clear indication that DO gains access to endosomes in DC comes from the subcellular distribution observed with Mags.DO5. Colocalization was obvious with CD63, confirming that DM is highly efficient in releasing DO from ER retention (Fig. 4A , 4g
4h
4i
4j
4k
4l
). These experiments were repeated with similar results using DC transduced with AdDO (Fig. 4B)
. In addition, a strong colocalization was observed in vesicular structures between Mags.DO5 and the DM-specific antibody (Fig. 4B , 4j
4k
4l
).
|
and ß chains are produced in vast excess and that DO molecules associating with DM are transported rapidly to the endosomes, possibly after internalization from the plasma membrane.
DO modulates antigen processing in DC
The subcellular localization of transduced DO inside the endocytic pathway of DC validates further studies regarding its potential role in antigen processing. First, we monitored the inhibition of CLIP release by flow cytometry using the Cer.CLIP antibody specific for class II-CLIP complexes [42
]. Our results show the endogenous DO and transduced DOß are sufficient to inhibit DM and increase the surface expression of CLIP (Fig. 5A
). DC transduced with the control Ad0 construct only showed basal levels of CLIP. Overexpression of DO and -ß chains slightly and reproducibly increased the levels of CLIP, proportionally to the increase in MagsDO staining (Fig. 3B)
. These results suggest that DO, as in B cells and various transfectants, gains access to early endosomes, where the pH allows its inhibitory activity to manifest.
|
52-63/H-2Ab using H2-O-deficient cells or transfectants overexpressing the chaperone [20
, 38
]. Another study demonstrated that the expression of H2-O does not affect the presentation in vivo and in vitro of endogenous antigens by class II+ tumor cells [55
]. To gain further insights into this issue, we tested the potency of our transduced human DC in a MLR. In this assay, T cell proliferation relies, at least in part, on the recognition of foreign class II molecules bound to DM-dependent, natural peptides [56
]. DC were incubated at various ratios with purified CD4+ T cells from an unrelated donor. T cell activation was assessed by 3H-thymidine incorporation. Although DC preparations did not show any proliferation in the absence of added T cells, a strong alloresponse was measured when T cells and DC were coincubated (Fig. 5B)
. Nontransduced and transduced DC were equally effective in these conditions. Similar results were obtained with different combinations of donors (data not shown). Thus, the consequence of DO expression on CLIP release and antigen processing does not affect, at least quantitatively, the alloresponse triggered by transduced DC. It has been proposed that the inhibition of H2-M/DM by "O" molecules in early endocytic compartments could prevent the presentation of antigens internalized by fluid-phase endocytosis [18 ]. As DC are highly efficient in the capture and internalization of antigens [35 ], we addressed the possible role of DO in modulating the presentation of exogenous gp100. We have recently suggested that the down-regulation of DO in CD40-activated B cells could contribute to the enhanced presentation of this tumor antigen [47 ]. Lysates were prepared from the MelS melanoma cell line transfected with gp100 and pulsed on transduced DC. The cells were treated with LPS to complete their maturation and cocultured with the gp100-specific, DRB1*0701-restricted T cell clone. Results show that the DO+ DC were less efficient in presenting the gp100 epitope than cells transduced with the empty Ad0 (Fig. 5C , left panel). The effect was reproducibly more pronounced when using a limiting amount of antigens. The negative control represents a lysate from MelS cells transfected with GFP. We also compared the effect of titrating the amount of DO by comparing the stimulation obtained with AdDOß- versus AdDO-transduced DC (Fig. 5C , right panel). The results show that cells transduced with both chains of DO (AdDO), which displayed more Mags.DO-positive complexes and more cell-surface CLIP (see Fig. 2 ), were less potent in processing and presenting the gp100 peptide. Again, this effect of DO was more pronounced when using limited amounts of antigens. Finally, a DO molecule with a mutation reducing the interaction with DM did not inhibit the presentation of gp100 (A. Bellemare-Pelletier, Francis Deshaies, J. Thibodeau, unpublished data). Altogether, these results demonstrate the capacity of DO to modulate the peptide repertoire of DC and to affect the processing of soluble, exogenous antigens.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Manipulation of the class II pathway offers interesting, possible avenues for the development of tumor vaccines. More specifically, Ii has been modified to encode and deliver T cell epitopes to peptide-loading compartments of the endocytic pathway [60 , 61 ]. Also, inhibition of Ii expression in class II+ cancer cells has been used to derive efficient cell-based vaccines against unknown tumor antigens in animal models [62 , 63 ]. DM affects the peptide repertoire independently of the presence of Ii, suggesting that it might also serve as a target to modulate the efficiency of vaccination [64 ]. Moreover, DM was shown to determine the immunodominant fate of T cell epitopes [65 ]. Inhibition of DM in class II+ cancer cells could allow the presentation of cryptic epitopes from tumor antigens and boost the anti-tumor immune response. Indeed, the immunological response to antigens is usually directed against a narrow set of immunodominant peptides derived from complex antigens [66 ]. Other epitopes are hidden because of inadequate processing or low affinity for MHC molecules. The unmasking of cryptic determinants through epitope spreading has been described and could have important practical implication for the development of immunization strategies [67 ]. In this context, it is interesting that B lymphocytes express a natural regulator of DM and of antigen processing. One could thus envisage using DO to modulate DM in tumor antigen-pulsed DC and to activate new T cell specificities capable of recognizing tumors expressing low levels of cryptic epitopes. Indeed, under rules that have yet to be defined, DO was shown to up-regulate the presentation of some peptides. Deciphering the molecular aspects of its regulatory activity could help the design of powerful immunomodulators, not only for B cell responses but also for DC used in immunotherapy or vaccination.
So far inseparable from its immunomodulatory properties, the ability of DO to inhibit CLIP release from class II molecules could be clinically relevant. Auto-bone marrow transplant (BMT) following chemotherapy or irradiation represents an efficient, curative therapy for many patients with acute or even chronic leukemias, as well as for patients with lymphomas or breast cancer [68 69 70 ]. Following auto-BMT and cyclosporin A treatment, an anti-tumor activity develops in parallel to the establishment of a graft-versus-host disease [71 ]. It is surprising that in a rat model, the CD8+ effector T lymphocytes responsible for such graft-versus-leukemia (GVL) activity are directed at CLIP-associated MHC class II molecules [72 ]. Combining BMT and adoptive immunotherapies based on DO+, CLIP+ DC might boost the GVL efficiency and therapeutic benefits of auto-BMT.
Expression of endogenous DO and -ß in DC has been controversial for some time, probably as a result of the use of various cell types or the choice of antibodies [12
, 25
, 28
]. The differential regulation of the and ß chains warrants precautions in the interpretation of the data [24
, 73
]. Clearly, the and ß chains are required to form an active molecule capable of binding to DM and of gaining access to the endocytic pathway [11
]. Our RT-PCR, Western blotting, and confocal microscopy analyses confirm that DC do not express functional DOß heterodimers. The expression of the endogenous chain seems low in DC, as it was not detected by Western blotting. This could be a result of the relatively short half-life of its mRNA and to rapid degradation in the absence of a stabilizing ß chain [26
]. Indeed, transduction of DOß alone was sufficient to generate functional DO molecules, as judged by the presence of Mags.DO-positive molecules in the endocytic pathway of transduced DC and the increased surface expression of CLIP (Figs. 4
and 5)
.
Many evidences suggested that the amount of DM molecules available for interacting with DO is low in our experimental system: Very few DO and DOß chains acquired resistance to EndoH, which is particularly striking when comparing with transfected HeLa cells (Fig. 3A)
; transducing and ß chains of DO did not dramatically increase, over transduction of DOß alone, the Mags.DO+ species or the surface expression of CLIP; a large fraction of the transduced DOß colocalizes with the ER resident protein calnexin; all the DM molecules are EndoH-resistant on the Western blot analysis and are found in the endocytic vesicles by immunofluorescence microscopy (Figs. 3A and 4B)
; and the amount of CLIP is not as important as what is seen, for example, on B cell lines or transfectants expressing DO (data not shown). These results could be explained by a slowing down of the DM transcription upon DC transduction and maturation. Indeed, it was recently shown that LPS-induced maturation of DC results in the arrest of CIITA gene transcription and class II expression [74
]. Transduction by the adenoviral vector initiated the maturation of the DC (Fig. 2)
and could have shut down CIITA transcription. There is still a controversy as to the effect of viral vectors on iDC (see ref. [75
] and references therein). Our results are in agreement with those of Rea and collaborators [76
], showing that adenoviral vectors trigger some phenotypic changes in iDC, as judged by the increase in DR expression and the modest up-regulation of CD83.
In the light of the previously reported CIITA down-regulation in mDC and to maximize the yield of active DO molecules, we are considering an alternative approach, where the dispensable DO cDNA would be replaced by CIITA in our AdDO construct [77
, 78
]. This way, a larger pool of every molecule involved in antigen processing (including DM and DOß) would be synthesized simultaneously. Also, we could use a DOß mutant devoid of its cytoplasmic dileucine-sorting motif [13
, 14
]. This mutation was shown in Mel JuSo cells to further inhibit CLIP release and to redistribute DM from the internal vesicles to the limiting membrane of multivesicular bodies [14
]. Again, one has to bear in mind that the results obtained in cell lines might not be reproducible in DC [79
]. Finally, we have recently generated a functional mutant DO molecule that escapes ER retention, even in the absence of DM (F. Deshaies et al., submitted). Such DO could saturate the entire DM pool, including the molecules that were already sorted to the endosomes. Additional experiments will be needed to maximize the impact of DO in DC. For example, the need for LPS and the timing of transduction and of antigen delivery are issues that need to be carefully tuned.
DO was shown to play a preponderant role for the presentation of antigens taken up by the specific, high-affinity surface Ig on B lymphocytes [21 ]. The effect of DO, if any, on the processing of antigens internalized through other receptors such as the complement receptor, DEC 205, or the macrophage mannose receptor remains to be analyzed [80 , 81 ].
Even upon saturation of the entire DM pool, the inhibition by DO will never parallel the dramatic phenotype observed in DM-negative cells or mice. This is because DO does not inhibit DM in acidic compartments [18 , 28 ]. Thus, the role of DO could not be reproduced by the mere inhibition of DM, for example, with siRNA. As opposed to DM-negative APC from KO mice, which triggered an allogeneic response in syngenic wild-type animals, DO– splenocytes did not affect T cell recognition. Accordingly, our results confirmed that DO+ DC do not alter the intensity of the alloresponse [29 ]. The antigens eliciting such a response are not affected by the presence of DO, or they are replaced by peptides having similar properties. Qualitative variations might be revealed once the repertoire of alloreactive T cells toward DO-positive and -negative APC is compared carefully [64 ].
Collectively, our results demonstrate the functionality of DO expressed in human MoDC. Although we have not demonstrated the capacity of DO to increase the presentation of gp100, this system should allow the identification of some therapeutically relevant antigens showing enhanced presentation in these conditions. The gp100 epitope recognized by this T cell clone is not characterized and is most probably generated and loaded in early endosomes, where DM is inactivated. Other cryptic epitopes might benefit from DM inhibition. A global analysis of the TcR repertoire and clonotypes of primary cells responding to an antigen in the context of DO+ or DO– DC could shed light on this issue. Still, the demonstrated ability of DO to reduce processing could favor the activation of high-avidity T cells. It is interesting that CLIP can polarize T cell responses toward the Th2 phenotype [23 , 82 ]. If a dramatic DO-induced CLIP increase at the surface of transduced DC could qualitatively affect the response remains to be determined. Further studies will be needed to understand the rules governing the immunomodulatory effects of DO and to maximize its potential in DC.
|
ACKNOWLEDGEMENTS |
|---|
Received January 13, 2005; revised February 15, 2005; accepted February 25, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
ß dimers and facilitates peptide loading Cell 82,155-165[CrossRef][Medline]
chain DZ is distinct from genes in the DP, DQ and DR subregions EMBO J. 4,2231-2237[Medline]
/DO ß heterodimer J. Immunol. 165,1410-1416 chain residues critical for binding of the toxic shock syndrome toxin superantigen J. Exp. Med. 176,1779-1784 J. Exp. Med. 179,1109-1118 v ß 3 and v ß 5 promote adenovirus internalization but not virus attachment Cell 73,309-319[CrossRef][Medline]
