(Received for publication, August 27, 1996)
From the Centre d'Immunologie INSERM-CNRS de Marseille-Luminy, 13, 288 Marseille, France
We noticed that B cell receptor ligation or
phorbol 12-myristate 13-acetate treatment induced intracellular
vesicles containing major histocompatibility complex (MHC) class II and
invariant chain (Ii), and increased the amount of transmembrane p12 Ii
fragments coimmunoprecipitated with class II molecules. To determine
the influence of protein kinase C activation on the MHC class II
presentation pathway, we analyzed the subcellular distribution of Ii,
the induction of SDS-stable forms of class II molecules, and their
ability to present different antigens. Ii chains visualized with
luminal and cytoplasmic directed antibodies appeared in early endosomal compartments accessible to transferrin in response to phorbol 12-myristate 13-acetate treatment, whereas transmembrane Ii degradation products equivalent to the p12 Ii fragments were colocalized with the B
cell receptors internalized after cross-linking. Protein kinase C
activation delayed in parallel the formation of SDS-stable forms of
class II molecules and reduced the presentation of antigenic determinants requiring newly synthesized class II -Ii complexes. These data indicate that B cell activation affects Ii processing and
MHC class II peptide loading in endosomal compartments intersecting the
biosynthetic pathway.
MHC1 class II molecules bind in their
groove antigenic peptides generally derived from endocytosed proteins
for their presentation to specific CD4+ T lymphocytes
(reviewed in Refs. 1-3). Shortly after their biosynthesis, the MHC
class II and
chains assemble with Ii through a sequence from
amino acid 81 to 104 of Ii called CLIP for class II associated Ii
peptide (4-6), and form (
)3Ii3 nonameric
structures (7, 8). Ii prevents the association of antigenic peptides
with class II
heterodimers (9-11) through CLIP peptide (5),
which occupies the peptide-binding groove in crystallized MHC class II
molecules (12). The trimerization of Ii cytoplasmic domains, containing
two dileucine motives each, drives the targeting of newly synthesized
class II molecules to specialized endosomal compartments (13-15),
which were further characterized as compartments of antigen processing
and of peptide loading (16-19). The amount of class II molecules
located in these intracellular compartments can differ to a
considerable extent, depending on the cell type (20-23), and
proteolytic cleavage of Ii occurs at this stage as indicated by
ultrastructural observations (22). Upon removal of CLIP from class II
molecules catalyzed by HLA-DM or its murine equivalent H2-M class II
molecules (24-26), peptide-loaded MHC class II complexes are exported
to the cell surface by a poorly defined pathway. Recycling of class II
molecules, from the plasma membrane through endosomes, has also been
observed in B cells (27-29), allowing binding and presentation of
different antigens to T cells (27, 30). MHC class II molecules can
therefore gain access to the antigen-processing compartments by at
least two different routes, a direct targeting of newly synthesized
-Ii complexes and an indirect targeting of resident
heterodimers. These two pathways of antigen presentation coexist in B
cells. One is sensitive to protein synthesis and membrane transport
inhibitors (31-33), and requires the expression of Ii (34). The other
one requires preexisting
class II heterodimers with intact
cytoplasmic domains (30).
In B lymphocytes, the most efficient pathway of antigen uptake is
mediated by the BCR (reviewed in Ref. 35). Upon BCR ligation, antigens
are internalized independently of a phosphorylation of the
immunoreceptor tyrosine-based activation motives present on the
cytoplasmic tail of the coreceptors Ig and Ig
molecules (36).
Ligation of the BCR transduces activation signals, through the Ig
and Ig
coreceptors, leading to a cascade of protein tyrosine kinase
activation (reviewed in Refs. 37 and 38) and to the production of
second messengers, such as inositol triphosphate, and PKC
activators, such as diacylglycerol (39). Several routes dependent on
protein tyrosine kinase and PKC are then converging at the level
of the activation of the mitogen-activated protein kinase pathway
(40).
Since the BCR delivers signals leading to PKC activation (37, 41), it was of interest to analyze the effects of BCR cross-linking and of direct PKC activation, through phorbol ester, on the biosynthetic transport and the function of MHC class II molecules. Our results show that different mechanisms of PKC activation decreased the endosomal degradation of MHC class II associated Ii chains and induced an intracellular accumulation of p12 Ii protein fragments previously identified in MHC class II-Ii transfectants (20) and equivalent to the p10 Ii protein fragment SLIP (23, 42). This is correlated with a selective modulation of antigen presentation, and a regulation of vesicular traffic as shown for the uptake of transferrin (43).
Anti-Ii rabbit polyclonal Abs were
raised against synthetic peptides of the cytoplasmic and CLIP
domain of the mouse Ii (Cyt.Ii and
CLIP). Anti-MHC class II
rabbit polyclonal Ab was raised against cytoplasmic domains of the
mouse I-A
molecule (
Cyt.IA
). The corresponding amino
acid sequences of the peptides used for immunization were:
M1DDQRDLISNHEQLPILGNRPREPESRCSRY31,
YR77MKLPKSAKPVSQMRMATPLLMRPMSMDNMLLG109,
and YH222RSQKGPRGPPPAGLLQ238 for the
Cyt.Ii,
CLIP, and
Cyt.IA
respectively. Peptides were coupled to keyhole limpet hemocyanin as a protein carrier with the
cross-linker bisdiazobenzidine. The biochemical characterization of
CLIP and
Cyt.IA
Abs was previously performed (44). The rabbit
antiserum
Cyt.Ii was conjugated to the NHS-LC-biotin (Pierce) for
double staining with other rabbit antiserum. The mouse hybridoma 10.2.16 producing an I-Ak-specific mAb was obtained from
the American Type Culture Collection (Rockville, MD). The rabbit
antiserum against luminal domain of human Ii (
Lum.Ii), the mouse
anti-lgp110-B mAb GL2A7 (45), and the mouse anti-Golgi apparatus mAb
CTR433 (46) were kindly provided by Drs. Salamero, Amigorena, and
Bornens (Curie Institute, Paris) respectively. The secondary reagents
(donkey anti-mouse Igs, anti-rabbit Igs, and streptavidin) coupled to
FITC, Texas Red, or Cyanine 5 suitable for multiple labeling
experiments and unlabeled donkey anti-mouse Igs to cross-link the BCR
were purchased from Jackson Immunoresearch (West Grove, PA). Human
transferrin (Sigma) was conjugated to FITC and purified on G25 column.
PMA and leupeptin were from Sigma and ICN respectively.
The F6 B cell lymphoma was derived from the M12C3 B
cell line, I-A negative, transfected for expression of I-Ak
molecules (47) and the B lymphoma, the CH27 cell line used in antigen
presentation experiments expressed H-2k molecules (48). The
3A9 T cell hybridoma (49) and the TS12 T cell hybridoma (50) are both
I-Ak-restricted and specific for HEL 46-61 and RNase A
43-56 peptides, respectively. Cell lines were cultured in Dulbecco'
modified Eagle's medium (Life Technologies, Inc.) supplemented with
10% fetal calf serum, 20 µM -mercaptoethanol, 1 mM sodium pyruvate, and 1 mM glutamine.
F6 cells cultured to 75% confluence on glass coverslips were treated or untreated with 50 ng/ml PMA or with 100 µM leupeptin. Cells were fixed for 15 min at room temperature with 4% paraformaldehyde in PBS. After washing, cells were permeabilized with PBS containing 0.05% saponin, then incubated for 30 min at room temperature with primary Abs diluted in PBS containing 0.05% saponin and 1% bovine serum albumin. Unbound Abs were removed by washing in the same medium, then cells were incubated for 30 min with secondary labeled Abs. After washing in PBS and distilled water, the coverslips were mounted onto glass slides with Mowiol plus 1,4-diazabicyclo[2.2.2]octane (Sigma). For double immunofluorescence using two rabbit antisera, the first staining was performed with the first primary Ab followed by the secondary labeled Ab as described above. The cells were then incubated with rabbit preimmune serum, washed, and post-fixed with 2% paraformaldehyde before being incubated for 30 min with the second primary antibody coupled to biotin. The cells were then rinsed and labeled with FITC or Texas Red-conjugated streptavidin. To label the endosomal compartments, FITC-coupled transferrin was endocytosed for 20 min, or surface Igs were cross-linked with anti-Ig Ab for 30 min at 37 °C, just before fixation and multiple labeling of other internal molecules. Confocal microscopy was performed using the Confocal Laser Scanning Microscopy TCS 4D (Leica Lasertechnik GmbH, Heidelberg Germany) interfaced with an argon/krypton ion laser and with fluorescence filters and detectors allowing to record simultaneously FITC, Texas Red, and Cyanine 5 markers. For the visualization of triple labeling, we have avoided the combination of red, green, and blue that produces too many hues leading to color interpretation problems. Instead, we defined a fluorescence threshold of 10% with an image processing program that automatically draws a black line around FITC-positive structures; the black contours are then superimposed to the conventional red-green pseudocolor look-up table representing the Texas Red and Cyanine 5 fluorescent markers (from D. Demandolx, CIML, Marseille).
Western BlottingF6 cells were either untreated or treated
with 100 µM leupeptin, with 50 ng/ml PMA, or with 10 µg/ml donkey anti-Ig Ab to cross-link the BCR for 1 h at
37 °C. Cells were washed in ice-cold PBS and were solubilized in 1 ml of lysis buffer (1% Nonidet P-40, 150 mM NaCl, 1 mM EDTA, 50 mM Tris-HCl, pH 7.5) containing a
mixture of protease inhibitors (0.25 mM
phenylmethylsulfonyl fluoride, 0.5 mM iodoacetamide, and 1 µg/ml leupeptin, pepstatin, and aprotinin). Half of the lysate was
directly charged on 15% SDS-polyacrylamide gel electrophoresis (PAGE).
Separated proteins were transferred onto Immobilon-P membrane
(Millipore). Ii and Ii fragments containing intact cytoplasmic tail
were detected with the Cyt.Ii Ab. The second half of the lysate was
incubated for 2 h with 10.2.16 mAb bound to the protein
A-Sepharose beads (Pharmacia Biotech Inc.). Then immunoprecipitates
were washed, and pellets were suspended in reducing sample buffer (10 mM Tris-HCl, 2 mM EDTA, 33% glycerol, 2% SDS,
5%
-mercaptoethanol). Immunoprecipitates were boiled in SDS and run
on 15% SDS-PAGE. After transferring onto membrane, proteins were
blotted with
Lum.Ii,
Cyt.Ii, or
CLIP Abs. After washing,
membranes were incubated with anti-rabbit Ab conjugated to horseradish
peroxydase (Jackson Immunoresearch). Labeled proteins were detected
using the ECL immunodetection kit (Amersham Corp.).
F6 cells were washed
twice and incubated for 45 min at 37 °C in cysteine/methionine-free
RPMI medium (Life Technologies, Inc.). In 3 ml of this medium
containing 5% dialyzed FCS, cells were incubated for 30 min at
37 °C with 0.6 mCi of [35S]cysteine/methionine in the
presence or in the absence of 50 ng/ml PMA. When indicated,
pulse-labeled cells were chased for different periods at 37 °C in
RPMI medium containing cold cysteine and methionine with or without
PMA. Cells were washed with ice-cold PBS and solubilized in lysis
buffer previously described. Lysates were precleared with protein
A-Sepharose beads, and the supernatants were mixed with Cyt.Ii Ab or
10.2.16 mAb previously bound to protein A-Sepharose beads.
Immunoprecipitates were extensively washed and pellets were resuspended
in reducing sample buffer. To detect SDS-stable
dimers,
immunoprecipitates were left at room temperature 1 h before being
resolved on 15% acrylamide SDS-PAGE. Radiolabeled proteins were
intensified and revealed by autoradiography.
F6 and CH27 antigen-presenting cells (APC) were untreated or treated with 50 ng/ml PMA for 1 h then washed. Cells were pulsed with various doses of HEL and RNase A for 2 h. After extensive washing, 5·104 APCs were cocultured for 24 h at 37 °C with 105 3A9 or 105 TS12 T cell hybridomas. Interleukin-2 production in culture supernatants was measured using thiazolyl blue 3-(4,5-dimethyl thiazol-2-xy)-2,5-diphenyl tetrazolium bromide (Sigma) to evaluate the growth of the interleukin-2-dependent CTLL-2 cell line.
To determine whether B cell activation influenced the
accumulation of Ii and of class II molecules into specialized
intracellular compartments, we used confocal microscopy and
immunofluorescence labeling of I-Ak and Ii molecules in F6
B lymphoma cells. In untreated cells, MHC class II molecules labeled
with Cyt.IA
polyclonal Ab were localized mainly at the cell
surface (Fig. 1A). Ii chain visualized with
Cyt.Ii polyclonal Ab gave mostly a reticulated intracellular staining consistent with the endoplasmic reticulum (Fig.
1B). After cross-linking of the surface Igs, MHC class II
molecules appeared in intracellular vesicles (Fig. 1C,
arrows) colocalized with Ii (Fig. 1D,
arrows). A similar colocalization of MHC class II and Ii was
found in PMA-treated cells (Fig. 1, E and F,
arrows).
The presence in class II-rich compartments of Ii proteins and Ii
degradation fragments was reported in human B cell lines (22) and in
murine B lymphoma cells treated with the leupeptin protease inhibitor
(23). To visualize the subcellular distribution and the processing of
Ii molecules in PMA (Fig. 2B)- and in
leupeptin (Fig. 2A)-treated cells, we double labeled the
cells with rabbit Cyt.Ii and
Lum.Ii Abs. A set of vesicles was
found to contain colocalized luminal and cytoplasmic Ii epitopes,
likely from full length Ii proteins (Fig. 2, A and
B, yellow vesicles), while many peripheral
vesicles were labeled with the
Cyt.Ii Ab only (Fig. 2, A
and B, red vesicles). These vesicles contain only
short Ii fragments lacking their luminal domains. Since distinct
vesicular compartments contained different forms of Ii, it was of
importance to localize Ii fragments with respect to Golgi, lysosomal,
and endosomal markers in PMA-treated cells. The mouse Golgi CTR433 mAb
which recognizes a marker present in the Golgi complex (46), was
partially colocalized with the
Lum.Ii (Fig. 2C,
yellow vesicles) and weakly colocalized with
Cyt.Ii
labeling (Fig. 2D). Some vesicles positive for the
Lum.Ii
Ab and many peripheral vesicles, containing Ii degradation products
recognized by the
Cyt.Ii Ab, were not coincident with this Golgi
marker (Fig. 2, C and D, red
vesicles). Next, to identify the lysosomal compartments, we used a
mAb directed against the lgp110-B marker (45) and found no
superimposition of the
Cyt.Ii Ab after PMA treatment (Fig.
2E), suggesting that Ii fragments do not reach or do not
persist in the lysosomes (51). Since antigens bound to the BCR are
transported to peptide-loading compartments for processing and
presentation by the MHC class II molecules (35), it was of importance
to determine whether the BCR could reach vesicles containing the
different forms of Ii. Internalization of the BCR was triggered by
cross-linking with anti-Ig Ab in PMA-treated cells. The anti-Ig Ab
internalized for 0.5 h were highly colocalized with peripheral
vesicles containing cytoplasmic Ii fragments (Fig. 2F).
To further define the intracellular localization of Ii degradation
products with respect to the early endosomes in activated cells,
FITC-coupled transferrin was internalized through its receptor during
PMA treatment. The cells were then fixed and processed for double
immunofluorescence staining with the Lum.Ii and
Cyt.Ii Abs. We
found a limited level of colocalization between the early endosomes
loaded with FITC-transferrin and the Ii positive vesicles labeled with
the
Cyt.Ii Ab (Fig. 2G, yellow vesicles). To determine whether the Ii-positive early endosomes contained also the luminal Ii
epitope, we used an image processing program to surround
FITC-transferrin-positive structures above a fluorescence value of 10%
with black contours. These contours were then superimposed on the
double fluorescence image obtained with
Lum.Ii and
Cyt.Ii Abs
(Fig. 2H). This representation shows triple positive
vesicles as double labeled objects surrounded with black contours. Some
early endosomes contain the two Ii epitopes (Fig. 2H, yellow
vesicles with black contours), while only a few contain short Ii
cytoplasmic fragments (Fig. 2H, red vesicles with black
contours). We think that newly synthesized MHC class II-Ii complexes
bearing unprocessed Ii chains gain access to early endosomal
compartments. PKC activation also generates Ii fragments that are
colocalized with internalized surface Igs in later elements of the
endosomal pathway (Fig. 2F).
The cytoplasmic Ii fragments identified in endosomal
compartments after PKC activation are probably equivalent to
cytoplasmic derived SLIP Ii protein fragment associated with
intracellular class II molecules (23, 42) and to the p12 Ii fragments
identified in -Ii-transfected fibroblasts (20). To test this
hypothesis, we performed Western blotting with
Cyt.Ii,
Lum.Ii,
and
CLIP antisera in total cell lysates from leupeptin (100 µM)- and PMA (50 ng/ml)-treated cells (Fig.
3A, lanes L and P,
respectively). The
Cyt.Ii polyclonal Ab recognized the p10, p12,
p31, and to a lesser extent the p41 forms of Ii in the total cell
lysate. After leupeptin or PMA treatment, there was no change in the
amount of p10 and p31 Ii forms; however, a slight increase in the p12 Ii form was detected in PMA treated cells (Fig. 3A). Since
Ii is produced in excess compared to class II molecules, we identified also I-Ak-associated Ii proteins on Western blots after
immunoprecipitation with the nonconformational 10.2.16 anti-I-Ak mAb (Fig. 3B). Compared to the
controls, no change occurred for p31 and p41 Ii forms associated with
class II molecules; however, PMA and leupeptin treatments increased the
class II-associated p12 Ii fragments, revealed with rabbit
Cyt.Ii
and
CLIP Abs (Fig. 3B, lanes L and
P). The cytoplasmic derived p10 Ii fragments identified in
the total cell lysates do not bind to class II molecules (Fig. 3,
A and B). In another set of experiments, we
compared the effect of PMA treatment and BCR ligation (Fig.
4, lanes P and B) on the pattern
of MHC class II-associated Ii fragments revealed with the
Cyt.Ii Ab
(Fig. 4A) and the
CLIP Ab (Fig. 4B). Both PKC activation pathways increased the level of association of p12 Ii
fragments with class II molecules (Fig. 4, A and
B).
To test whether PKC activation modified Ii turnover, we performed a
35S pulse-chase labeling followed by SDS-PAGE analysis of
Ii products immunoprecipitated with the Cyt.Ii Ab. In untreated
cells, Ii was rapidly degraded and a p12 Ii fragment appeared after 30 min of 35S pulse (Fig. 5A) analog
to SLIP. Faint p10 Ii fragments were detected after 2 h of chase,
and most of the p31 and p41 Ii forms were degraded within 3 h. In
cells treated with PMA, degradation of Ii was consistently reduced with
more p12 fragment at 2 h and more p31 form at 3 h of chase
compared with the untreated cells (Fig. 5A). Rather than
increasing the biosynthesis of Ii, PMA reduced the rate of Ii turnover.
This could affect peptide loading on newly synthesized
class II
dimers as in leupeptin-treated cells (52). To analyze the kinetics of
antigen binding we monitored the resistance of class II complexes to
denaturation by SDS detergent at 20 °C reflecting the presence of
peptide loaded class II
heterodimers (53). In untreated cells,
newly synthesized class II molecules acquired resistance to SDS 1 h after their biosynthesis; the mature form of class II
chains
(
m) appeared after 30 min of chase, while immature
chains (
i) disappeared after 1 h (Fig. 5B). Immature
chains (
i) disappeared after
30 min of chase, but since mature
chains (
m) migrate
at the same position as the p31 Ii chains, we could not really detect
the initial rate of
chain maturation on these SDS-PAGE. In
PMA-treated cells, SDS-resistant
dimers were delayed for
2.5 h following biosynthesis (Fig. 5B), but PMA did not
affect class II
and
chain maturation patterns. We think
therefore that PKC activation affects the peptide loading process
thought to occur in specialized endosomal compartments (16-19) in
which the persistence of Ii fragments can reduce the rate of
formation of SDS-resistant class II
dimers.
Effect of PMA on Antigen Presentation
In order to determine the effects of PKC activation on the MHC class II antigen presentation pathway, we studied the response of two T cell hybridoma requiring different pathways of antigen processing. The I-Ak-restricted presentation of HEL to the HEL 46-61 specific 3A9 T cell hybridoma (49) is sensitive to protein synthesis and membrane transport inhibitors (31-33), and requires Ii expression (34). The presentation of RNase A to the RNase A 43-56-specific TS12 T cell hybridoma requires neither protein synthesis nor Ii chain expression (34, 50). These two antigenic determinants are thought to use different pathways of processing for their presentation to helper T cells. Using cycloheximide-treated F6 and CH27 B lymphoma cells, we confirmed that, in B cells, protein synthesis inhibitor blocked the presentation of HEL to the 3A9 T cell hybridoma, but not the presentation of RNase A to the TS12 T cell hybridoma.2 The RNase A 43-56 peptide is therefore presented here by resident MHC class II molecules, whereas the HEL 46-61 peptide is presented by newly synthesized class II molecules.
In order to assess the function of newly synthesized and of resident
class II molecules depending on PKC activation, we delivered the
antigen by uptake from the fluid phase for 2 h in F6 B lymphoma cells (Fig. 6, A and B), and in
the nonadherent CH27 B lymphoma cells for the sake of comparison (Fig.
6, C and D). At low doses of antigen, the
presentation of HEL to the I-Ak-restricted 3A9 T cell
hybridoma was reduced by about 10-fold when the APCs were pretreated
for 1 h with 50 ng/ml PMA (Fig. 6, A and C).
The presentation of HEL became insensitive to PMA at high concentration
of antigen. In contrast to HEL, the efficiency of the presentation of
RNase A to the I-Ak-restricted TS12 specific T cell
hybridoma was not affected by pretreatment with PMA for 1 h in F6
and in CH27 B cell lines (Fig. 6, B and D). PKC
activation altered selectively the presentation pathway of the
HEL-derived 46-61 peptides, which are thought to meet newly
synthesized MHC class II molecules after processing of the protein.
The ligation of activation receptors such as surface Igs and the direct activation of PKC generate a cascade of phosphorylation events in B cells (37-40) which influences protein-protein interactions in the cytoplasm, activates gene transcription and modifies the cell morphology. Our goal was to investigate whether B cell activation influences the MHC class II presentation pathway. Initial experiments showed that surface Ig cross-linking and phorbol ester treatment can induce the appearance of MHC class II/Ii-containing intracellular vesicles in murine B lymphoma cells. These observations led us to evaluate the effects of PKC activation on Ii processing and transport, MHC class II peptide loading, and antigen presentation.
Redistribution of Ii Protein Fragments in PMA-treated B CellsUsing anti-peptide Abs raised against the cytoplasmic N-terminal and the luminal C-terminal portion of Ii, we found by immunofluorescence microscopy a reticulated cytoplasmic staining in untreated cells and a vesicular pattern in cells activated by BCR ligation and PMA treatment. Most of the Ii-positive vesicles were labeled with anti-MHC class II Ab and probably contain newly synthesized MHC-Ii complexes since they appeared after 30 min and were susceptible to protein synthesis inhibitors.2 The Ii-positive vesicles were further divided into two subsets according to the stage of Ii processing. One subset contained colocalized luminal and cytoplasmic Ii epitopes and the other contained only the cytoplasmic Ii epitope. Short term treatment with the leupeptin serine protease inhibitor induced a similar redistribution of Ii proteins and Ii fragments (Fig. 2, A and B).
Most of the vesicular compartments containing the luminal Ii epitope
were also labeled with the CTR433 mAb, which is characterized at the
ultrastructural level as a marker of the mid-Golgi compartment (46).
However, many vesicles containing only transmembrane Ii fragments were
not labeled with this Golgi marker and all of them were distinct from
the lysosomes identified with the lgp110-B marker (45). Ii-positive
vesicles were also distinct from the H2-M molecules present in the
lysosomes (54), and from the late endosomes2 defined by the
anti-Rab7 mAb and by the presence of the cation-independent mannose
6 phosphate receptor (55). The absence of colocalization with lysosomal
and late endosomal markers indicates that Ii cytoplasmic fragments are
present in vesicles equivalent to endosomal compartments defined by
subcellular fractionation in murine B cells (16, 17). Using multiple
immunofluorescence and confocal microscopy, we found that PKC
activation induces the redistribution of intact Ii from the endoplasmic
reticulum to the Golgi complex and to other elements of the endosomal
pathway. After cross-linking and internalization of the BCR, numerous
endocytic vesicles defined by their Ig content were labeled with the
Cyt.Ii Ab.
To determine whether the vesicles in which MHC class II-Ii complexes
are delivered correspond to early endosomes, we performed a double
immunofluorescence staining with Lum.Ii and
Cyt.Ii Abs following
internalization of FITC-transferrin (Fig. 2, G and H). Some early endosomes labeled with transferrin contain no
Ii molecules, whereas other ones are labeled with both luminal and cytoplasmic Ii-directed Abs. Intact Ii chains are probably reaching the
early endosomal compartments first, whereas later elements of endosomal
pathway are loaded with Ii fragments partially colocalized with
internalized Igs. This scheme is compatible with the steady state
distribution of Ii degradation products found in distinct compartments
in human B cell lines at the ultrastructural level (22). Our results
are also in agreement with recent analyses of subcellular fractions
derived from leupeptin-treated cells (23) and with subcellular
fractionation experiments performed in murine B lymphoma cells (56),
since class II molecules were colocalized with Ii proteins and Ii
fragments after BCR engagement or after PMA treatment. In conclusion,
PKC activation triggers the accumulation of intact Ii presumably
associated with newly synthesized MHC class II molecules in vesicular
structures accessible to transferrin uptake. Later elements of
the endosomal pathway, in which surface Igs are internalized, contain
apparently many Ii fragments, suggesting an initial targeting of
-Ii complexes to transferrin positive compartments as reported in
leupeptin treated cells (23, 56).
Looking at the
biosynthetic pathway of MHC class II transport, we report here that PKC
activation reduces Ii turnover and delays the induction of class II SDS
stable forms. PMA treatment also triggers the accumulation of CLIP
containing p12 Ii fragments previously identified in fibroblast
transfectants (20) and analogous to p10 cytoplasmic derived fragments
obtained in leupeptin-treated B cells (23, 42). The p12 Ii fragments
contain in their cytoplasmic portion two dileucine based targeting
motives able to direct Ii chain and Ii chimeric constructs to endosomal
compartments (13, 14). The fact that Ii proteins and p12 Ii fragments
remain associated with heterodimers for a longer time in PMA
treated cells provides an explanation for the intracellular retention
of
-Ii complexes and for the delay in the induction of SDS stable
forms. Moreover, intracellular Ii and MHC class II molecules remained
highly colocalized in PKC activated cells, indicating that the CLIP
portion of p12 Ii fragments presumably lies in the groove of newly
synthesized MHC class II molecules and impairs peptide loading in the
endosomes.
To identify whether PKC activation influences the antigen presentation
capacity of B cells in relation to the biochemical perturbation of Ii
processing, we have analyzed the presentation efficiency of
Ii-dependent and Ii-independent epitopes. In the B cell
populations selected here, newly synthesized class II molecules are
required to present HEL to the HEL 46-61 specific I-Ak
restricted, 3A9 T cell hybridoma (31, 32), and this antigen presentation pathway is critically dependent on Ii expression in B
cells (34). We showed here that this presentation event is partially
impaired by PMA treatment. However, we found no correlation with a
reduction in class II surface expression analyzed by flow cytometry and
no correlation with a reduction in the presentation of the HEL 46-61
peptide.2 The MHC class II recycling pathway of antigen
presentation was resistant to protein synthesis inhibitors here. This
pathway does not require Ii (34) but requires the integrity of MHC
class II cytoplasmic domains (30). As an example of this class II recycling pathway, we analyzed the presentation of RNase A to the RNase
A 43-56 specific I-Ak restricted, TS12 T cell hybridoma
(31, 32) and observed no influence of PKC activation, indicating in
addition that reorganization of the plasma membrane is not responsible
for the inhibition of HEL presentation to the 3A9 T cell hybridoma.
From our results, we think that the biosynthetic pathway of class II
presentation requires a rapid degradation of Ii fragments to be fully
functional. In splenic B cells derived from mice lacking H2-M
molecules, a similar defect occurs for the presentation of antigenic
determinants requiring both Ii expression and MHC class II synthesis
(44). In the absence of H2-M which catalyze the exchange of CLIP
peptides, Ii processing is normal but MHC class II molecules present at the cell surface are massively loaded with CLIP and do not reach their
final SDS stable forms (44). In PKC activated cells in which Ii
processing is partially inhibited, we found a higher degree of
intracellular retention of MHC class II and Ii molecules. This is
probably due to the fact that dileucine motives present in multiple
copies in the ()3-Ii3 and the
(
)3-(p12Ii)3 nonamers can retain the
complexes in endosomal compartments (13-15).
The molecular mechanism leading to the reduction of Ii degradation and
Ii-dependent antigen presentation after PMA treatment and
BCR ligation remains, however, unclear. We looked for a PMA-induced phosphorylation of MHC class II and Ii, but we had no evidence in favor
of this hypothesis. As shown previously in human B cells (57), Ii
chains can be phosphorylated on serine residues lying in the
cytoplasmic domain, and this event may regulate
Ii-dependent antigen presentation. We have detected a basal
level of 32P incorporation in p31 and p12 forms of Ii
immunoprecipitated with Cyt.Ii Ab, but found no modification of Ii
and MHC class II phosphorylation after PMA treatment.2 We
cannot exclude that PMA induced another type of post-translational modification of Ii, or reduced the activity and the expression of
proteases involved in Ii degradation (58, 59). Many endosomal proteases
are synthesized as precursors and need a proteolytic step occurring in
the trans-Golgi network and in subsequent compartments to become fully
active (60). PMA could also influence the maturation of proteases
resulting in a slower degradation and the accumulation of Ii fragments
partially associated with class II molecules. It should be noted that
inhibitors of protein phosphatases such as okadaic acid can regulate
endocytic vesicle fusion in vitro (61) and inhibits
endocytosis in Hela cells (62) consistent with other reports showing
that PMA can modulate the uptake of transferrin and of fluid phase
markers (43). Additional events such as a reorganization of the
cytoskeleton perturbing the exocytic traffic of MHC class II-Ii
complexes remain also possible.
Using BCR ligation and PKC activation, we observed for the first time a
negative regulation of the MHC class II biosynthetic pathway of antigen
presentation. Phorbol esters reduce peptide loading and conversion of
class II molecules into SDS-resistant heterodimers as in cells treated
with the leupeptin serine protease inhibitor (23, 52). However, the
mechanisms of action of PKC stimulation and of vacuolar proteases
inhibition differ to a considerable extent. The inhibition of vacuolar
proteases affects both the processing of the antigens and the
degradation of MHC class II associated Ii. In contrast to leupeptin,
which blocks the induction of SDS-stable forms of class II and their
transport to the cell surface (52), PKC activation through PMA
treatment modulates the presentation of Ii-dependent
epitopes. It is tempting to speculate that the sequence of events we
describe, i.e. delayed Ii degradation, accumulation of Ii
products in endosomal compartments involved in peptide loading of
heterodimers, delayed maturation of SDS-resistant
dimers,
and impaired presentation of Ii-dependent exogenous antigens, corresponds to a selective desensitization of the MHC class
II antigen presentation pathway in cells in which many BCR molecules
are engaged simultaneously.
We thank Drs J.-P. Gorvel and S. Meresse for their constant support during the course of the experiments, D. Demandolx for providing us with new image processing software, Drs. S. Amigorena, M. Bornens and J. Salamero for their gift of antibodies. We are also grateful to Drs. S. Amigorena, P. Chavrier, P. Golstein, L. Leserman, A. Manfredi, and P. Rovere for critical reading of the manuscript. We are indebted to Leica SA France for their financial support of the color prints.