| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Divisions of
*
Tumor Biology,
Cell Biology, and
Cellular Biochemistry, Netherlands Cancer Institute, Amsterdam, The Netherlands
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and the class II-associated invariant chain, the
DO cytoplasmic tail contains potential lysosomal targeting signals.
The DO signals, however, are not essential for internalization of
the DM/DO complex from the plasma membrane or targeting to the MIICs.
Instead, the DO tail determines the distribution of both DM/DO and
class II within the multivesicular MIIC by preferentially localizing
them to the limiting membrane and, in lesser amounts, to the internal
membranes. This distribution augments the efficiency of class II
antigenic peptide loading by affecting the efficacy of lateral
interaction between DM/DO and class II molecules. Sorting of DM/DO and
class II molecules to specific localizations within the MIIC represents
a novel way of regulating MHC class II Ag
presentation. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In B cells the composition of the class II antigenic peptide repertoire is controlled by yet another class II-like molecule, termed HLA-DO (7, 8, 9). DO tightly associates with DM, which is required for endoplasmic reticulum (ER) exit and interferes with its catalytic function (10, 11, 12). DO-expressing cells accumulate class II/CLIP complexes at the cell surface (10, 11, 12), yet formation of stable class II/peptide complexes is only reduced by half. The efficacy of DM inhibition by DO is pH dependent, being more effective at endosomal pH than at the acidic pH of the MIICs. The specific inhibition of DM by DO in endosomes results in a modulation of the class II peptide repertoire by preventing the presentation of certain peptides while inducing the presentation of others (12, 13).
Whereas class II molecules only transiently linger in the MIICs to pick
up their Ags on their way to the cell surface (14, 15),
both DM (16) and DO (9) are considered to be
MIIC residents. The mechanism behind this bifurcation in the
intracellular trafficking pathways of class II molecules and DM/DO is
still poorly understood. Internalization and intracellular targeting of
transmembrane proteins is mediated by discrete motifs in their
cytoplasmic domains. In general, these motifs are either tyrosine based
(YxxØ or NpxY, where x is any amino acid, and Ø is an amino acid with
a bulky hydrophobic group) or di-leucine-based (17, 18).
Class II molecules use the Ii di-leucine-based motif to reach the MIICs
(1). Lysosomal targeting of DM is dependent on a DM
tyrosine-based motif (YTPL) (19, 20). Interestingly, the
cytoplasmic domain of human DO contains two potential targeting
motifs (21, 22), the function of which is as yet
unclear.
The cellular trafficking pathway of DM/DO complexes is investigated
using the green fluorescent protein (GFP) tagged to DO. We show that
DM/DO complexes are constitutively delivered from MIICs to the cell
surface together with class II molecules, and are immediately
internalized to recycle back to the MIICs. Although we demonstrate that
two potential lysosomal targeting motifs in the DO cytoplasmic
domain are functional in a chimera with CD8, they are not of importance
for cellular trafficking of the DM/DO complex. However, the cytoplasmic
domain of DO redistributes class II molecules and DM/DO complexes
within the multivesicular MIIC, which strikingly affects the efficiency
of peptide exchange. Thus, we define a novel sorting mechanism that
regulates the efficacy of the class II-mediated immune response by
specialized protein sorting within the MIICs.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The human melanoma cell line Mel JuSo (HLA-A1, HLA-B8, HLA-Cw7, HLA-DR3, and HLA-DQ2 by DNA typing) and its stable transfectants are grown in IMEM with 10% FCS in the presence or the absence of 2000 µg/ml G418, 600 µg/ml hygromycin (Life Technologies, Gaithersburg, MD), and/or 500 µg/ml Zeocin (Invitrogen, Breda, The Netherlands). Continuous expression of the GFP-tagged proteins is ensured by regular selection of the GFP-positive cells by FACS.
The rabbit polyclonal antiserum against class II was previously
described (15), and the sera against DO
(A10 3.0) and
DO (A64 5.1) were raised against fusion proteins of GST with DO
or DO, respectively. The HLA-DR-specific mAb L243 (American Type
Culture Collection, Manassas, VA) (23) and 1B5
(24), the class II/CLIP-specific Ab CERCLIP.1
(25), the DR3-antigenic peptide specific mAb 16.23
(26, 27), the DM-specific mAb 5C1 (28),
and the anti-transferrin receptor mAb (B3/25; Roche, Indianapolis,
IN) were previously described. The anti-actin mAb Ab-1 was
purchased from Oncogene Research Products (San Diego, CA), the
anti-human CD8 OKT8 mAb was obtained from the American Type Culture
Collection (CRL-8014), and the Texas Red-conjugated secondary Abs were
purchased from Molecular Probes (Eugene, OR).
Quantification of Western blots was performed with the Alpha Innotech (San Leandro, CA) digital imaging system and Phoretix 1D Advanced software (Alpha Innotech).
DNA constructs
The DO, DO-GFP, and nuclear localization signal (NLS)-GFP
constructs have been previously described (10, 29). The
cDNAs encoding human dynamin (type 1), both wild-type (wt) and dominant
negative (K44A, negative for nucleotide binding of the GTPase),
provided by A. M. van der Bliek (30), were recloned
into pcDNA3. DNAs were introduced into the Mel JuSo cell line using
calcium phosphate or via microinjection.
The CD8-DM, CD8-DO, and CD8-DO chimeras were generated by
ligation of SalI/BamHI-digested PCR products
encoding the cytoplasmic tails of DM, DO, or DO to
SalI/BamHI-digested pCMUIV-CD8 vector
(31). The CD8-DO chimeras containing the tyrosine to
alanine and/or the di-leucine to alanine substitution were generated by
cloning hybridized complementary primers encoding the mutated
cytoplasmic DO tail into the SalI/BamHI site
of pCMUIV-CD8.
The pcDNA3 IRES-NLS-GFP vector was generated by ligation of part of the
LZRSneo vector (containing a multiple stop sequence, a multiple cloning
site, and an internal ribosomal entry site (IRES)) into the
BamHI and EcoRI sites of the pcDNA3 NLS-GFP
vector. The cDNA of DO was cloned into the XhoI site of
pcDNA3 IRES-NLS-GFP. The DO-CD8 chimeric construct was generated by
a fusion PCR between the extracellular to transmembrane region of DO
and the cytoplasmic domain of CD8, followed by cloning into the
EcoRI and XhoI sites of pcDNA3 IRES-NLS-GFP.
Subcellular fractionation by density gradient electrophoresis
Subcellular fractionation on wtDO and DOCD8 transfectants
incubated with 2 mg/ml HRP (type VIA; Sigma, St. Louis, MO) for 30 min
at 37°C was essentially performed as previously described
(32). The fractions collected were assayed for total
protein (bicinchoninic acid; Pierce, Rockford, IL), HRP activity, and
-hexosaminidase. Plasma membrane fractions were separated from ER
fractions using wheat-germ agglutinin (32). Proteins from
the different fractions were precipitated with TCA and analyzed by
SDS-PAGE and immunoblotting with A64 5.1 (DO).
FACS analysis
For FACS analysis, 1 x 106 cells were stained with saturating amounts of unlabeled primary Ab and PE-conjugated F(ab')2 rabbit anti-mouse IgG (H+L; Zymed, San Francisco, CA) and analyzed on a FACScan flow cytometer (BD Biosciences, Mountain View, CA).
Microinjection and confocal microscopy
Fifty to 100 cells were typically microinjected intranuclearly with a mix of the cDNA encoding NLS-GFP (0.2 ng/ml) and the cDNAs indicated (0.1–0.2 µg/µl) on a heated xy stage of a Zeiss Anxiovert 135 M microscope (New York, NY) equipped with an Eppendorf manipulator 5171/transjector 5246 system as previously described (29). After injection, cells were allowed to recover overnight at 37°C for expression of the microinjected cDNAs.
Living cells were analyzed in a heated tissue culture chamber at
37°C. Brefeldin A (BFA; 5 µg/ml; Sigma) was added 1.5–4.5 h before
confocal laser scanning microscopy (CLSM) analysis. CLSM analyses were
performed using a Bio-Rad MRC600 confocal microscope equipped with an
argon/krypton laser (Bio-Rad, Hercules, CA). Single-channel green
fluorescence was detected at 
>515 nm after excitation at
= 488 nm. For dual analyses, green fluorescence was detected at
= 520–560 nm, and red fluorescence was detected at >585
nm after excitation at = 568 nm.
For immunofluorescence labeling, cells were fixed with 3.7% formaldehyde in PBS, permeabilized (0.1% Triton X-100 in PBS), and incubated for 1 h with Abs diluted in PBS containing 0.5% BSA, followed by washing and 1-h incubations with Texas Red-conjugated secondary Abs and additional repeated washing.
Cell surface Ag presentation assay
Cells were incubated in medium with 200 µM cycloheximide at 37°C for 4 h. After trypsinization, cells were washed twice in phosphate-citrate buffer (137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, and 1.14 mM citric acid, adjusted to pH 5 or 7, supplemented with 0.05% NaN3, 50 µM 2-deoxyglucose, and 0.1% BSA). Subsequently, cells were incubated in the appropriate phosphate-citrate buffer with protease inhibitors in the absence or the presence of 20 µM biotinylated ApoB2877–2894 at 37°C for 2.5 h. After washing with cold PBS/0.1% BSA, bound biotinylated peptides were detected by streptavidin-conjugated PE (Molecular Probes) followed by FACS analysis.
Immunoelectron microscopy
Cells were fixed in 0.5% glutaraldehyde and 4%
paraformaldehyde in 0.1 M phosphate buffer (pH 7.2) and processed for
ultrathin cryosectioning as previously described (14). For
immunolabeling, sections were incubated with purified mAb
anti-human DM (5C1), followed by subsequent incubations with
rabbit anti-mouse IgG and 10 nm protein A-conjugated colloidal
gold. Next, the sections were treated with 1% glutaraldehyde, followed
by further incubation with rabbit anti-human class II serum and 15
nm protein A-conjugated colloidal gold. After embedding in a mixture of
methyl-cellulose and uranyl acetate, sections were analyzed with a
Philips CM10 electron microscope (Eindhoven, The Netherlands). For
quantification of class II and DM distribution patterns in the
multivesicular bodies (MVB), gold label in apparent contact with the
perimeter membrane was considered perimeter membrane labeling.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To study the intracellular transport of the DM/DO complex we used
a stable transfectant of the melanoma cell line Mel JuSo expressing a
heterodimeric complex of DO and a chimera of GFP fused to the DO
cytoplasmic tail (10). The DOGFP complex
quantitatively associates with DM without affecting endogenous class II
and DM levels (10). Confocal analysis shows that
DM/DOGFP complexes accumulate in perinuclear, lysosome-like MIICs
(10), which is indistinguishable from that of DRGFP
(Fig. 1
, -BFA).
|
GFP-containing vesicles to the cell surface (14).
Although DM/DO resides in similar vesicles, it is not observed at the
cell surface (Fig. 1
). DM/DO may be selectively retained in MIICs,
which implies a differential intracellular sorting of class II and
DM/DO complexes. Alternatively, DM/DO is cotransported to the cell
surface and immediately reinternalized. The route of entry into the
MIICs of DM/DOGFP is compared with that of DRGFP in living
cells by addition of BFA to block the transport of newly synthesized
molecules out of the ER. One and a half hours of BFA incubation do not
visibly affect the quantity and distribution of the
DOGFP/DM-containing vesicles (data not shown). This indicates
that the vesicles do not form part of the constitutive secretory
pathway from the ER to the cell surface, as the secretory vesicles get
depleted of newly synthesized material within this time frame. Similar
observations were made for DRGFP-containing vesicles
(14). Prolonged BFA incubation for an additional 3 h
enhances fluorescent staining of the reticulate ER network due to
accumulation of newly synthesized GFP-tagged proteins and depletes
the cells from intracellular vesicular DRGFP (Fig. 1, +BFA).
Interestingly, DM/DOGFP complexes remain present in
intracellular vesicles, albeit in slightly reduced numbers (Fig. 1, +BFA). Similar findings were obtained using cycloheximide (data not
shown). Thus, whereas DRGFP is transported through the
endosomal/lysosomal compartments and incorporated into the plasma
membrane within 4.5 h, DM/DOGFP complexes continue to
prevail in intracellular compartments.
To discriminate between selective retention of DM/DO in MIICs and
deposition at the plasma membrane followed by rapid internalization,
cells are microinjected with cDNAs encoding wild-type dynamin or a
dominant negative dynamin mutant that inhibits clathrin-mediated
endocytosis (30). Coinjection with a cDNA encoding GFP
tagged with a strong NLS (33) allows identification of
microinjected cells by nuclear GFP fluorescence and simultaneously
assesses the viability of the cells. The wt dynamin does not affect the
distribution of either DM/DOGFP or MHC class II complexes (Fig. 2A, wtDyn). Overexpression of dominant negative dynamin
induces membrane ruffling, typical for cells inhibited in endocytosis.
The distribution of fluorescence in DRGFP-expressing cells is not
significantly affected by mutant dynamin. Fluorescent DM/DOGFP
complexes, however, accumulate at the plasma membrane (indicated by
arrows) of the microinjected cells, but not in control cells (Fig. 2A, asterisks, DN Dyn-BFA). Cell surface expression of
HLA-DO was confirmed by staining intact nonpermeabilized
dominant-negative dynamin-expressing cells with DO, followed by
CLSM analysis (data not shown). These results demonstrate that in
normal cells DM/DO is exported to the cell surface together with class
II, followed by immediate reinternalization by clathrin-coated
vesicles, whereas the majority of class II complexes remain at the cell
surface. The very transient appearance of low amounts of DM/DOGFP
complexes at the plasma membrane can be confirmed by cell surface
iodination and recovery of iodinated DM/DOGFP via
immunoprecipitation (data not shown), subcellular fractionation (see
below, Fig. 4D) and FACS analysis (see later). Iodinated
DM/DOGFP is still recovered after 4-h cycloheximide treatment,
confirming that the DM/DO pool that transiently appears at the plasma
membrane is not derived from the constitutive secretory pathway, as
this pathway gets depleted of proteins within 1.5 h (data not
shown). Effective inhibition of clathrin-mediated endocytosis by
mutant dynamin was demonstrated by immunofluorescence detection of the
transferrin receptor, as it accumulates at the plasma membrane of
microinjected cells instead of in early endosomes (Fig. 2B).
Uptake of the fluid phase marker sulforhodamine 101 was not affected by
mutant dynamin, indicating that it specifically blocks
clathrin-mediated endocytosis (data not shown).
|
|
A, DN Dyn-BFA). Indeed, additional BFA treatment depletes
cells blocked in endocytosis of all vesicular DM/DOGFP, whereas
control cells retain a pool of fluorescent vesicles as observed
previously (Fig. 2A, DN Dyn + BFA). In contrast, BFA
treatment depletes both microinjected cells and control cells from
intracellular DRGFP at similar rates, indicating that
replenishment of the endocytic pathway via recycling DRGFP is not
a major class II trafficking route in these cells (Fig. 2A,
DN Dyn + BFA).
Together, these data suggest that after deposition into MIICs all
DM/DOGFP complexes continuously recycle between the cell surface
and intracellular Ag loading compartments. DM/DO thus enters the MIICs
via two major routes, a direct input from the biosynthetic pathway and
reuse of exocytosed material via continuous recycling through the
endocytic pathway.
The cytoplasmic tail of DO contains lysosomal targeting signals
Targeting of DM to the MIICs is dependent on a tyrosine-based
motif in the cytoplasmic tail of DM (19, 20). Although
this signal may well suffice to target the DM/DO complex to the MIICs,
the putative targeting motifs in the DO-chain (Fig. 3A) could contribute to cellular trafficking of DM/DO. The
functionality of these motifs is investigated using a chimera in which
the DO cytoplasmic tail is fused to the extracellular and
transmembrane domains of CD8 (Fig. 3A). In addition,
CD8-DM and CD8-DO chimeras are generated. The localization of the
chimeras is examined after expression of the constructs in the Mel JuSo
DOGFP transfectant. The wt CD8 localizes to the plasma membrane
(Fig. 3B), consistent with the lack of a functional
targeting motif that could deviate the protein from the default
secretory pathway. Likewise, the CD8-DO chimera lacks a targeting
signal, as it primarily localizes to the plasma membrane (Fig. 3B). In contrast, the CD8-DM chimera is targeted to
perinuclear vesicles and colocalizes with DOGFP (Fig. 3B). In agreement with previous studies (19, 20), these vesicles are CD63 positive, mannose-6-phosphate
negative, and negative for the early endosome marker EEA1 (data not
shown), emphasizing the lysosomal nature of these compartments. The
CD8-DO chimera is also targeted to perinuclear vesicles that are
positive for DOGFP. Thus, the DO cytoplasmic tail contains a
functional targeting motif.
|
cytoplasmic tail contains a membrane-proximal tyrosine
residue and two membrane-distal leucine residues (Fig. 3A),
both potential lysosomal targeting motifs. To investigate whether
either of these residues is involved in targeting the CD8-DO
chimera, constructs are made in which the tyrosine and/or the two
leucines are mutated into alanine residues (Fig. 3A).
Elimination of the tyrosine, the two leucine residues, or both results
in accumulation of the chimeras at the plasma membrane (Fig. 3C). Thus, both the tyrosine residue and the two leucines
are involved in lysosomal targeting of the CD8-DO chimera.
The DO cytoplasmic tail is not essential for lysosomal targeting
of the DM/DO complex
As DM and DO tightly associate, the benefit of a targeting signal
in addition to the DM signal is unclear. Possibly the additional
sorting information in DO distinguishes the subcellular targeting of
the DM/DO complex from that of DM alone. To investigate the
contribution of the DO motif to trafficking of the DM/DO complex,
the targeting information is removed by replacing the DO tail in the
DM/DO complex by that of CD8 (Fig. 4A). The wt DO and chimeric DOCD8 chains are expressed
from one bicistronic transcript together with the NLS-GFP protein,
allowing a correlation of DO expression levels to those of GFP.
Stable clonal Mel JuSo transfectants are generated expressing DO in
conjunction with either wtDO or the DOCD8 chimera. Similar
expression levels of the DO complexes are obtained by FACS sorting of
the transfectants for equal GFP levels. Western blot analysis and
subsequent quantification confirms the similar DR, DM, and DO
expression levels (Fig. 4B).
To examine whether the tail of DO has an effect on subcellular
localization of the DM/DO complex, wtDO and DOCD8 transfectants are
subjected to subcellular fractionation by density gradient
electrophoresis. To correct for experimental variation, the
localization of lysosomes (-hexosaminidase), endosomes (HRP), and
ER/plasma membrane (total protein) in the gradient was determined in
each independent experiment. Subsequent Western blot analysis of the
different fractions revealed that in both wtDO and DOCD8 transfectants
DO (and DM, data not shown) primarily localizes to
-hexosaminidase-positive fractions (lysosomes), and only a small
portion is detectable in plasma membrane fractions (Fig. 4D). Since DO requires DM for ER exit (9), part
of the DO pool remains in the ER due to relative overexpression. This
assures quantitative association of DM with DO, as validated by
immunoprecipitation (10). Lysosomal localization of wtDO
and DOCD8 is confirmed by CLSM analysis (Fig. 4C). Thus, the
relative intracellular distribution of DM/DO is not affected by
DO tail modification. This indicates that the DM/DO complex does not
require the additional DO sorting signal for proper targeting to the
MIICs.
The cytoplasmic tail of DO influences peptide exchange
To investigate whether the DO tail has an effect on Ag
presentation, the composition of the class II/peptide complexes was
analyzed in the Mel JuSo, wtDO, and DOCD8 transfectants by FACS. The
total amount of class II molecules at the cell surface was relatively
unvaried (L243, Fig. 5A). In line with the DM modulatory function of DO
(10, 11, 12), a large portion of these class II molecules was
occupied by CLIP in the wtDO transfectant (CERCLIP, Fig. 5A). Surprisingly, however, the DOCD8 transfectant showed a
substantial further increase in class II/CLIP levels (CERCLIP; Fig. 5A). The relative amount of class II/CLIP complexes was even
more than doubled in the DOCD8 transfectant compared with the wtDO
transfectant (ratio CERCLIP/L243, Fig. 5A). The reduced
release of CLIP from class II in the DOCD8 transfectant was paralleled
by a 50% reduction in the relative amount of class II molecules
presenting stably binding peptides, as demonstrated by the mAb 16.23
that recognizes these complexes (ratio 16.23/L243, Fig. 5A).
|
B). Again, the DOCD8 transfectant shows an even greater
reduction in the amount of stable class II/peptide complexes vs
unstable class II molecules (Fig. 5B), demonstrating that
omission of the DO cytoplasmic tail is deleterious for the
efficiency of CLIP-antigenic peptide exchange on class II molecules.
The observed differences in Ag presentation between the wtDO and DOCD8
transfectant are consistent for all other wtDO and DOCD8 clones
(n > 5) tested. Furthermore, the catalytic activity of
DM/DO complexes is not affected by the absence of the DO tail as
determined by an in vitro peptide exchange assay (data not shown).
Taken together, these results demonstrate that the tail of DO
affects the efficiency of DM/DO-mediated catalysis of class II
antigenic peptide loading, although it does not influence the
subcellular localization or the catalytic activity of the DM/DO
complex.
The DO tail alters the distribution of DR and DM/DO within the
multivesicular body
How does the cytoplasmic tail of DO affect the process of Ag
presentation, as it is not involved in the localization of the DM/DO
complex to the MIICs? It could be that DO affects localization of the
DM/DO complex at a level that cannot be detected by subcellular
fractionation studies. Therefore, the intracellular localization of
DM/DO and class II in Mel JuSo and both transfectants is analyzed in
more detail by immunoelectron microscopy. Intracellularly both DM(/DO)
and class II primarily localize to MVB, a characteristic morphology for
the MIICs in this cell line (Fig. 6A). Strikingly, the distribution between the limiting
perimeter membrane and the internal vesicles of both DM(/DO) and class
II within the multivesicular MIICs is altered in the wtDO transfectant
compared with the DOCD8 transfectant or Mel JuSo itself (Fig. 6A). Quantification of the distribution demonstrates that in
Mel JuSo and the DOCD8 transfectant DR is equally distributed over
perimeter membrane and internal membranes, whereas DR is enriched on
the perimeter membrane in the wtDO transfectant (Fig. 6B).
The redistribution to the perimeter membrane is also seen for DM(/DO),
going from 30% in Mel JuSo and the DOCD8 transfectant to 50% in the
wtDO transfectant (Fig. 6B). The redistribution seems to be
specific for DR and DM/DO complexes, since the distribution of the
lysosomal marker CD63 residing in the same MIICs is not altered (data
not shown). Apparently, the DO tail determines the intracompartment
localization of the DM/DO complex as well as that of the interacting
class II molecules (34, 35).
|
tail in the same planar membrane, wtDO and DOCD8 transfectants
were used in a cell surface peptide exchange assay. Cell surface class
II and DM/DO reside in the same planar membrane and should therefore
have the same capability for lateral interaction in both wtDO and DOCD8
transfectants. To enrich cell surface DM/DO complexes, cells were
treated for 4 h with cycloheximide, as we noted previously that
this impedes the efficiency of DM/DO internalization (data not shown),
but not of exocytosis. FACS analysis demonstrated equal levels of
DM/wtDO and DM/DOCD8 at the plasma membrane (data not shown). Cells
were then incubated with or without the biotinylated antigenic peptide
ApoB2877–2894 at the indicated pHs. Bound
biotinylated peptides were detected with streptavidin-conjugated PE in
FACS analysis. At pH 7, Mel JuSo, wtDO, and DOCD8 bind ApoB with
similar efficiency (Fig. 6C), in line with the low catalytic
activity of both DM and DM/DO at neutral pH (13). When
incubated at pH 5, Mel JuSo shows a strong increase in the loading of
ApoB, while the wtDO or DOCD8 transfectant show a reduced increase
(Fig. 6C), in concordance with the inhibitory activity of DO
on DM and the in vitro peptide exchange assays published
(13). When wtDO and DOCD8 are compared, both are equally
efficient in loading ApoB (Fig. 6C), again confirming that
the catalytic activity of DM/DO is not affected by the absence of the
DO tail. Thus, the difference in class II loading between wtDO and
DOCD8 (Fig. 5) is relieved when the ability for lateral interaction
between class II and DM/DO is the same. This strongly suggests that the
DO-mediated redistribution of class II and DM toward the perimeter
membrane of MIICs provides an optimal setting for lateral interaction
between the DM/DO complex and class II molecules. As a consequence,
this results in a more efficient class II peptide exchange. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
An important consequence of the continuous recycling of DM/DO is that these proteins traverse all compartments of the endocytic pathway before they accumulate in the MIIC. Thus, the modulatory activity of DM(/DO) on class II peptide loading is not confined to the MIIC, but extends to both the transport pathway from MIIC to plasma membrane as well as the entire endocytic pathway. This observation is reminiscent of the presence of DM in compartments other than the classical MIIC (37). The formation of class II/peptide complexes occurs in different compartments of the endosomal/lysosomal pathway; some epitopes are preferentially loaded onto class II molecules in early compartments, others in later compartments (38, 39, 40, 41). Although DM can potentially catalyze the loading of peptides in all these compartments, to date early endosomal class II peptide loading has been reported to be DM independent (40, 42, 43, 44). The present study demonstrates that this cannot be explained by the absence of DM from these compartments. The suboptimal pH in early endosomes for DM catalysis, and the high transit rate of proteins through early endosomes may account for the apparent DM-independent peptide loading in these compartments. B cells that express DO may have an additional mechanism to prevent DM-catalyzed peptide loading in early endosomes by the pH-dependent inhibition of DM by DO (12, 13). Indeed, peptide loading of Ags that are taken up by the B cell receptor occurs in acidified compartments and is dependent on newly synthesized class II molecules (45, 46, 47). B cells may therefore use DO to confine DM function to these compartments, as previously suggested (12).
Both class II molecules and DM/DO are directly sorted from the
trans-Golgi network into the endosomal pathway. The class II
molecules are targeted into the endosomal/lysosomal pathway via a
di-leucine-based lysosomal targeting signal in the associated Ii
(1). DM uses a different targeting signal (tyrosine-based)
present in the DM cytoplasmic tail for lysosomal targeting
(19, 20). We here demonstrate that DO contains two
targeting motifs that have the potential to mediate lysosomal
targeting, only when acting in consortium. This in contrast to the
tyrosine- and di-leucine-based motifs described to date that are
individually sufficient for lysosomal targeting (48).
Recently, Brunet et al. (49) reported an exclusive role of
the DO di-leucine motif for intracellular targeting of their class
II reporter protein in HeLa cells. Their data also show that
elimination of this di-leucine motif resulted in some cell surface
expression of the class II reporter molecule, with a considerable
fraction still colocalizing with LAMP-1 (e.g., Fig. 5, G–L)
(49). It is possible that further elimination of the
tyrosine residue would have resulted in exclusive cell surface
expression. Our data clearly show that elimination of either the
tyrosine or di-leucine motif is sufficient for cell surface expression
of the CD8 reporter molecules with a concomitant exclusion from
DOGFP vesicles (Fig. 3C). Thus, the DM/DO complex may
represent a unique example of a protein complex with multiple targeting
signals, a tyrosine-based motif for DM and both a tyrosine and a
di-leucine motif in the DO-chain.
The targeting information in DO is not required for correct
lysosomal targeting of the DM/DO complex, since replacing the DO
cytoplasmic domain for that of CD8 does not abolish lysosomal
localization of the DM/DO complex. The lysosomal targeting signal of DM
is probably sufficient for proper intracellular sorting of the DM/DO
complex. These observations were recently confirmed by Brunet et al.
(49). Although not required for targeting, the cytoplasmic
tail of DO influences the efficiency of peptide exchange. These
effects on Ag presentation correlate with the unique ability of DO to
determine the localization of class II and DM within the multivesicular
MIIC. Since DO is stably associated to DM (35), DO can
easily codistribute DM molecules within the MVB. Apparently, class II
molecules interact with DM/DO complexes at least transiently, because
they are also redistributed.
How can the observed redistribution of class II molecules and DM,
induced by the cytoplasmic domain of DO be beneficial for peptide
exchange? The presence of a protein on either the perimeter membrane or
the internal vesicles can determine further intracellular trafficking
of such a protein. The epidermal growth factor receptor is internalized
upon ligand binding and sorted toward the internal vesicles of MVBs,
resulting in its degradation. Kinase-dead mutants of the epidermal
growth factor receptor are also internalized, but are sorted to the
perimeter membrane of MVBs, followed by recycling to the plasma
membrane (50). Thus, the redistribution of class II
molecules and DM/DO toward the perimeter membrane would be beneficial
for directing class II molecules to the cell surface. The cell surface
levels of class II and the total amount of class II, however, do not
significantly differ between wtDO and the DOCD8 transfectant, probably
because class II is resistant to lysosomal conditions.
Another explanation for the enhanced efficiency of peptide exchange in
wtDO cells may be that redistribution of class II and DM/DO toward the
perimeter membrane enhances the probability of interaction through free
lateral mobility. Previous studies demonstrated that class II and DM
need to physically interact to catalyze peptide exchange (28, 51). Given the total number of class II molecules (
100,000 as
suggested for dendritic cells (52)) expressed by an APC
and the 20:1 molar ratio of total class II and DM (53), a
single MVB (100/cell) contains approximately 50 DM molecules. This
implies that only a subset of the internal vesicles will contain DM.
Since the internal vesicles of a MVB may be relatively secluded
subcompartments with restrained transport between internal vesicles (J.
Calafat et al., manuscript in preparation), lateral interaction between
class II and DM/DO on internal vesicles may be severely restricted. On
the other hand, the concurrent presence of class II and DM/DO on the
same perimeter membrane allows free lateral mobility and thereby
enhances the probability of interaction. As demonstrated here, DM and
DR indeed laterally interact, resulting in peptide exchange, consistent
with the observation of small amounts of catalytically active DM at the
cell surface (36). The redistribution of class II and
DM(/DO) toward the perimeter membrane of the MIICs is reminiscent of a
similar redistribution of class II molecules in Chediak-Higashi
syndrome B cells (54). This redistribution correlates to
an enhanced interaction between DM and class II molecules and a more
efficient removal of CLIP (55).
In conclusion, we have now demonstrated that DM/DO complexes
constitutively recycle between the plasma membrane and the MIIC.
Although the DM/DO complex contains multiple targeting signals,
retrieval of DM/DO complexes from the plasma membrane is primarily
controlled by the lysosomal targeting signal in DM. The targeting
information of DO is instead used for the distribution of class II
and DM/DO within the MVB. This distribution determines the efficiency
of peptide exchange and thereby forms a novel way to regulate class II
peptide loading. Thus, we have defined here for the first time that a
cytoplasmic domain (of DO) determines the localization of class II
and DM within the MVB and regulates class II loading at the
ultrastructural level.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 M.v.L. and M.v.H. contributed equally to this work.
3 Address correspondence and reprint requests to Dr. Marieke van Ham, Unit of Experimental Oncopathology, Department of Pathology, Free University Medical Center, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands. E-mail address: sm.vanham@vumc.nl or jneefjes{at}nki.nl
4 Abbreviations used in this paper: Ii, invariant chain; ApoB, apolipoprotein B; BFA, brefeldin A; CLIP, class II-associated invariant chain peptides; CLSM, confocal laser scanning microscopy; ER, endoplasmic reticulum; GFP, green fluorescent protein; IRES, internal ribosomal entry site; MIIC, MHC class II compartment; MVB, multivesicular body; NLS, nuclear localization signal; wt, wild type.
Received for publication November 9, 2000. Accepted for publication May 11, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
dimers and facilitates peptide loading. Cell 82:155.[Medline]
: a new chain gene in HLA-D with a distinct regulation of expression. EMBO J. 4:2839.[Medline]
chain directs HLA-DM to MHC class II compartments. J. Cell Biol. 131:351. and a chimeric DO chain. Int. Immunol. 11:99.-chain subunits of human Ia alloantigens. Immunology 50:613.[Medline]
. J. Biol. Chem. 275:37062.This article has been cited by other articles:
|
|
 |
|
  N. Lapaque, M. Jahnke, J. Trowsdale, and A. P. Kelly The HLA-DR{alpha} Chain Is Modified by Polyubiquitination J. Biol. Chem., March 13, 2009; 284(11): 7007 - 7016. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. van Lith and A. M. Benham The DM{alpha} and DMbeta Chain Cooperate in the Oxidation and Folding of HLA-DM J. Immunol., October 15, 2006; 177(8): 5430 - 5439. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Bellemare-Pelletier, J. Tremblay, S. Beaulieu, M.-R. Boulassel, J.-P. Routy, B. Massie, R. Lapointe, and J. Thibodeau HLA-DO transduced in human monocyte-derived dendritic cells modulates MHC class II antigen processing J. Leukoc. Biol., July 1, 2005; 78(1): 95 - 105. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Deshaies, A. Brunet, D. A. Diallo, L. K. Denzin, A. Samaan, and J. Thibodeau A point mutation in the groove of HLA-DO allows egress from the endoplasmic reticulum independent of HLA-DM PNAS, May 3, 2005; 102(18): 6443 - 6448. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Matsumoto, K. Funami, M. Tanabe, H. Oshiumi, M. Shingai, Y. Seto, A. Yamamoto, and T. Seya Subcellular Localization of Toll-Like Receptor 3 in Human Dendritic Cells J. Immunol., September 15, 2003; 171(6): 3154 - 3162. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Muntasell, M. Carrascal, L. Serradell, P. v. Veelen, F. Verreck, F. Koning, G. Raposo, J. Abian, and D. Jaraquemada HLA-DR4 Molecules in Neuroendocrine Epithelial Cells Associate to a Heterogeneous Repertoire of Cytoplasmic and Surface Self Peptides J. Immunol., November 1, 2002; 169(9): 5052 - 5060. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Li, K. Siemasko, M. R. Clark, and W. Song Cooperative interaction of Ig{alpha} and Ig{beta} of the BCR regulates the kinetics and specificity of antigen targeting Int. Immunol., October 1, 2002; 14(10): 1179 - 1191. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. Weber, C. T. Dao, J. Jun, J. L. Wigal, and P. E. Jensen Transmembrane Domain-Mediated Colocalization of HLA-DM and HLA-DR Is Required for Optimal HLA-DM Catalytic Activity J. Immunol., November 1, 2001; 167(9): 5167 - 5174. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Chen, O. Laur, T. Kambayashi, S. Li, R. A. Bray, D. A. Weber, L. Karlsson, and P. E. Jensen Regulated Expression of Human Histocompatibility Leukocyte Antigen (HLA)-DO During Antigen-dependent and Antigen-independent Phases of B Cell Development J. Exp. Med., April 15, 2002; 195(8): 1053 - 1062. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, Mel JuSo; 
, Mel JuSo/wtDO; black line, Mel JuSo/DOCD8; gray
line, secondary Ab only). The relative ratio of median values obtained
by CERCLIP.1 and 16.23 vs total class II levels (L243) is plotted in
the right panel. Shown are representative results from
three independent experiments, with mean fluorescence intensities
depicted in arbitrary units on a logarithmic scale. B,
Total cell lysates of Mel JuSo, Mel JuSo/wtDO, and Mel JuSo/DOCD8 were
boiled (b) in reducing sample buffer or left at room temperature (nb).
Samples were subsequently analyzed by SDS-PAGE and Western blotting.
SDS-stable class II complexes and the DR