| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
National Institute of Immunology, New Delhi, India;
National Centre for Biological Sciences, UAS-GKVK Campus, Bangalore, India; and
Department of Immunology, University of Washington School of Medicine, Seattle, WA 98195
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
52–68 yields an epitope that is
similar to the one generated during constitutive presentation of I-E
as an endogenous transmembrane protein, but is subtly different from
the one generated in the exogenous pathway. Constitutive MHC class
II-mediated presentation of the endogenous transmembrane protein I-E
is also specifically inhibited over time by inhibitors of cytosolic
proteolysis. Thus, Ag processing in the cytoplasm appears to be
essential for the efficient presentation of endogenous proteins, even
transmembrane ones, on MHC class II, and the proteolytic pathways
involved may differ from those used for MHC class I-mediated
presentation. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
However, MHC class I can present peptides from exogenous proteins, either by cytosolic leakage of endosomal contents (reviewed in Ref. 7) or via direct endolysosomal processing (reviewed in Ref. 8). Conversely, MHC class II can present peptides from cytosolic proteins (reviewed in Refs. 9 ; see also Refs. 10, 11). If both of these situations were exceptions, most peptides on MHC class I would still be from endogenous cellular sources and those on MHC class II from extracellular sources. Analyses of these peptide repertoires show that this is largely true for MHC class I (reviewed in Refs. 12, 13). However, major peptides on MHC class II also appear to be mainly from cellular rather than extracellular proteins (14, 15, 16). Although this could be explained partly by endolysosomal processing of membrane proteins, many of these peptides are from cytosol-resident proteins. The mechanisms for the generation of peptides from endogenous proteins for loading on MHC class II molecules are thus poorly understood.
It has been shown recently that processing pathways can also determine
the precise nature of a single peptide epitope presented by MHC class
II. For example, an epitope generated from the constitutively expressed
transmembrane protein I-E is 2 aa residues longer (aa 52–68) than
the one generated by its exogenous processing via forced
internalization (aa 52–66) (17). Such results suggest
that the same protein may be differently processed and generate
distinct peptide epitopes depending on whether it is endogenously or
exogenously encountered. However, peptides from both endogenous and
exogenous sources bind to MHC class II in the endolysosomal vesicles
(18, 19, 20, 21), although there are differences regarding the
precise characterization of this loading compartment (4).
Thus, it is possible that the major processing pathway for MHC class
II-restricted presentation of cellular proteins may differ from the
classical endolysosomal processing pathway described for presentation
of extracellular proteins on MHC class II.
In this context, we have examined the pathways of MHC class II-restricted presentation for cytosolically introduced proteins as well as a constitutively expressed transmembrane protein, and we show evidence here that the MHC class II-restricted presentation of both categories of proteins is dependent on cytosolic proteolysis, both proteasomal and nonproteasomal.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The mouse strains used, C3H/HeJ (H-2k), C57BL/6 (H-2b), B10.A(3R) (H-2i3), and B10.A(5R) (H-2i5), were bred in the small animal facility of the National Institute of Immunology (New Delhi, India) and used at 6–12 wk of age. All experiments were done with the approval of the Institutional Animal Ethics Committee.
Antigens
OVA or conalbumin (CA) (Sigma, St. Louis, MO) were maleylated
with maleic anhydride (Sigma) at alkaline pH as previously described
(22), dialyzed against PBS, and the degree of maleylation
was estimated as the loss of free 
amino groups measured by
2,4,6-nitrobenzenesulfonic acid. Maleylated proteins were used
only when maleylation was <90%. Demaleylation was done at pH 3.5 as
described elsewhere (22) and showed regeneration of 90%
of the free amino groups. A fusion protein (GST-I-E peptide (aa
52–68 (EapL)-myc) consisting of GST, residues aa 52–68 of the mouse
I-E sequence, and the oligopeptide sequence of the c-myc protein
recognized by the 9E10 mAb was made in Escherichia coli from
an ampicillin-resistant plasmid with isopropyl
-D-thiogalactoside-inducible expression
(data not shown) and was purified on a glutathione-Sepharose column.
All Ags were dialyzed extensively against PBS to remove small
degradation products, and the lack of processed fragments in all Ag
preparations was confirmed in presentation assays utilizing fixed
APCs.
Cell lines
The MHC class I (H-2Kb)-restricted
OVA-specific T cell transfectant line B3 was a gift from Dr. M. Bevan
(University of Washington, Seattle, WA). The MHC class II
(I-Ab)-restricted OVA-specific T cell hybridoma
13.8 was generated from OVA-immunized C57BL/6 mice and characterized as
I-Ab restricted (data not shown). The CA-specific
MHC class II (I-Ak)-restricted T cell line 002
which is a transfectoma derived from the CA-specific
I-Ak-restricted T cell clone D10.G4.1, and both
the I-Ab-restricted I-E peptide (Eap)-specific
T cell hybridomas, 1H3 recognizing both I-E peptide (aa 52–66)
(EapS) and EapL, and Ea6 recognizing EapS but not EapL
(17), were gifts from Dr. C. A. Janeway (Yale
University, New Haven, CT) as was the B cell line E4, which is a clone
of the A20 B cell line (H-2d) transfected with
I-Ab and I-Ab.
Cytosolic Ag delivery by osmotic lysis of pinosomes
Proteins were delivered into the cytosol using osmotic lysis of pinosomes as described previously (23). Briefly, APCs were incubated in hypertonic serum-free DMEM with 0.5 M sucrose, 10% polyethylene glycol 800, and 10 mM HEPES containing antigenic protein for 10 min at 37°C, followed by washing and incubation in isotonic serum-free DMEM for 5 min. Exogenous loading of Ags was done by following the same protocol using isotonic medium. Cells were again washed and, where indicated, incubated further for 3 h to allow processing before being fixed with 1% paraformaldehyde (Sigma) for 1 min, washed, and used as APCs in T cell stimulation assays.
Fluorescence microscopy
C57BL/6 macrophages adherent on coverslip-bottom dishes were incubated with 2 mg/ml fluorescein-conjugated dextran (F-Dex) dissolved in either hypertonic medium DMEM with 0.5 M sucrose and 10% polyethylene glycol 800) or in isotonic DMEM for 15 min at 37°C. Excess F-Dex was washed off with isotonic medium and the distribution of F-Dex in cells was immediately observed on an inverted fluorescence microscope (TE-300; Nikon, Tokyo, Japan) with a x60 oil objective with 1.4 numerical aperture. The images collected via a charge-coupled device camera (Princeton Instruments, Princeton, NJ) using Metamorph acquisition software (Universal Imaging, West Chester, PA) were processed for presentation in Adobe Photoshop (Adobe Systems, Mountain View, CA). Imaging of cells was done in the presence of 10 mM nigericin in a high potassium buffer (120 mM KCl, 5 mM NaCl, 1 mM CaCl2, 1 mM MgCl2, and 20 mM HEPES, pH 7.4) to neutralize endosomal pH.
Ag presentation assays
Plastic-adherent peritoneal macrophages from thioglycolate broth-primed mice were preincubated for 30 min with various drugs, such as the lysosomotropic agents chloroquine and NH4Cl (Sigma), the proteasome inhibitors lactacystin (Calbiochem, San Diego, CA), N-acetyl-leucinyl-leucinyl-norleucinal (LLnL) and N-acetyl-leucinyl-leucinyl-methioninal (LLM) (Sigma), or the tripeptidyl peptidase (TPP) inhibitors Ala-Ala-Phe-chloromethylketone (AAF-cmk) (Sigma) and butabindide (gift from Dr. C. Ganellin, University College, London, U.K.). Cells were then washed and either exogenously pulsed or cytosolically loaded with various Ags. They were then allowed to process Ags in the absence or presence of the inhibitors again for 3 h at 37°C and then fixed where required with 1% paraformaldehyde for 1 min, washed, and used as APCs in T cell stimulation assays.
T cell lines (1–10 x 104 cells/well) were stimulated with titrated concentrations of Ag-pulsed APCs in triplicate cultures in 200 µl of DMEM with 10% FCS, antibiotics, L-glutamine, and 0.5 mM 2-ME in 96-well flat-bottom plates (Falcon; Franklin Lakes, NJ) as indicated. Culture supernatants were collected 24–36 h later and used to estimate the IL-2 induced by stimulating the IL-2-dependent cell line CTLL-2 (1 x 104 cells/well) with them, incubating for 24 h, and pulsing the plates with 0.5 µCi of [3H]thymidine/well (NEN Life Science, Boston, MA) for 12–16 h to measure the proliferative responses. Plates were harvested onto glass fiber filters for scintillation counting (Betaplate; Wallac, Turku, Finland). Data are shown as proliferation (mean ± SE) observed in triplicate CTLL-2 cultures.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We introduced two proteins, CA or OVA, into the cytosol of mouse
peritoneal macrophages by osmotic lysis of hypertonic pinosomes
(23). Ag-loaded or mock-loaded H-2k
or H-2b cells were used for CA and OVA
presentation, respectively, to the appropriate T cell hybridomas, the
CA-specific I-Ak-restricted line 002, and the
OVA-specific I-Ab-restricted line 13.8. APCs
cytosolically loaded with either Ag were well recognized by the
relevant T cells (Fig. 1
, A
and B). Similar results were also seen with macrophage cell
lines such as BMC2, splenic B cells, and B cell lines such as LB27.4
(data not shown).
|
D) while it was in punctate
structures if endosomes were not lysed (Fig. 1E). We have
previously shown that exogenously delivered maleylated proteins bind to
scavenger receptors (SRs) on APCs and are presented to MHC class
II-restricted T cells 
1000-fold better than their nonmaleylated
counterparts (24). Therefore, to examine further whether
the presentation of cytosolically loaded proteins was due to very low
levels of residual unruptured endosomes, we determined whether maleyl
proteins were presented much better than their nonmaleylated
counterparts when delivered in this fashion.
Cytosolically delivered maleyl-CA or maleyl-OVA were not presented
better than their native counterparts; in fact, they were very poorly
presented (Fig. 1, A and B). In contrast,
exogenously added maleyl-OVA was presented 300- to 1000-fold more
efficiently than native OVA, and this enhanced presentation was due to
SR-mediated uptake, since it could be blocked by an excess of
maleyl-BSA (Fig. 1C). These data further confirm that the
MHC class II-restricted presentation of proteins delivered via osmotic
lysis of pinosomes is not due to the presence of residual unruptured
pinosomes. The cytosolic pathway for MHC class II-restricted
presentation observed is thus distinct from the classical pathway for
presenting exogenous proteins.
Efficiency of generation of peptide-MHC class II complexes via cytosolic or endocytic pathways
We next examined the efficiency of the MHC class II-restricted presentation of cytosolic proteins by comparing the relative levels of peptide-MHC complexes generated from varying concentrations of proteins loaded either into the cytosol or in endosomes. For this, either C57BL/6 or C3H/HeJ macrophages were pulsed with varying concentrations of native or maleylated OVA or CA in either hypertonic medium or isotonic medium for 15 min, followed by culture in isotonic medium along with the appropriate T cell lines.
The efficiency of presentation of cytosolic CA or OVA on MHC class II
was not very different from that of equivalent amounts of exogenous CA
or OVA (Fig. 2, B and
D). In contrast, cytosolic OVA was far better presented on
MHC class I than exogenously delivered OVA (Fig. 2F). As
expected, the maleyl proteins were very poorly processed and presented
on either MHC class I or MHC class II if they were loaded in the
cytosol, while their processing was very efficient for loading on both
classes of MHC if they were given exogenously (Fig. 2, A,
C, and E). To ensure that similar amounts of
extracellular fluid (and proteins) were taken up by APCs during
cytosolic and endocytic loading, we determined the amounts of HRP
included as a tracer associated with APCs by either mode of loading,
and found that 10 µl of fluid per million cells was taken up in
both cases (data not shown). These data suggest that the cytosolic Ag
processing pathway for peptide loading on MHC class II is efficient
even in comparison to the classical postpinocytotic endosomal
processing pathway, making the mechanisms involved in this pathway of
major interest.
|
The poor presentation of cytosolic maleyl proteins was surprising.
Maleylation blocks the free amino groups of lysine residues
(22), methylation of which marks proteins for
ubiquitination and proteasomal degradation (25). We have
already shown that protein maleylation inhibits their cytosolic MHC
class I-restricted presentation (26). It is therefore
possible that processing of cytosolic proteins for presentation via MHC
class II may also depend on the availability of free amino groups
in the protein. We confirmed this by demaleylating the maleyl protein,
which is known to maintain the SR-binding conformation conferred by
maleylation, although the lysine amino groups are reactivated
(22). Demaleylated maleyl-OVA showed regeneration of free
lysine amino groups (data not shown) as well as enhanced
SR-mediated MHC class II-restricted presentation exogenously (Fig. 1C). However, unlike maleyl-OVA, demaleylated maleyl-OVA was
well presented to T cells upon loading in the APC cytosol as well (Fig. 1B). These data show that MHC class II-restricted
presentation of cytosolically introduced proteins requires free
amino groups on lysine residues, suggesting the possibility that they
may be processed via ubiquitin-tagged proteasomal degradation similar
to that observed in the classical MHC class I-restricted presentation
pathway (25).
We next addressed the proteolytic mechanisms involved in the processing
of cytosolically loaded proteins for MHC class II-restricted
presentation. To explore the role of proteasomes, we used various
inhibitors of proteasomal degradation, lactacystin, LLnL, and LLM
(27). Of the three proteasome inhibitors used, LLnL and
LLM are reported to be equipotent inhibitors of cathepsin B and
calpains but with marked differences in their inhibition against the
proteasomes. The presence of lactacystin, LLnL, or LLM during
processing of cytosolic CA or OVA inhibited their MHC class
II-restricted presentation (Fig. 3, B and D). However, none of these inhibitors
affected the presentation of exogenous maleyl-CA or maleyl-OVA (Fig. 3, A and C) as well as of exogenous OVA or CA (data
not shown), ruling out any nonspecific inhibition mediated by their
ability to inhibit nonproteasomal activities. Lactacystin also
inhibited MHC class I-restricted presentation of cytosolic OVA (Fig. 3F), confirming that proteasome-dependent processes were
affected. MHC class I-restricted presentation of exogenous maleyl-OVA
was not inhibited by lactacystin (Fig. 3E), consistent with
the extra cytosolic processing of proteins delivered via SRs for
presentation on MHC class I (26). Thus, MHC class
II-restricted presentation of cytosolic proteins follows the MHC class
I-restricted presentation pathway in being dependent on proteasomal
activity, presumably for peptide generation. However, exogenous protein
presentation is proteasome independent for both classes of
MHC.
|
Proteasomal proteolysis has been implicated in the generation of
most MHC class I-restricted peptides (27). However, recent
data also point to the involvement of nonproteasomal proteolytic
mechanisms in the cytoplasm in generating peptides for MHC class
I-restricted presentation (28), perhaps in tandem with
proteasomal mechanisms (29). We therefore looked at the
role of nonproteasomal proteolysis in MHC class II-restricted
presentation of cytosolically introduced proteins by using two
inhibitors, AAF-cmk and butabindide, both of which inhibit TPPs,
particularly TPP-II (30, 31, 32). Butabindide appears to be
specific for TPP-II, whereas AAF-cmk inhibits other cytosolic proteases
implicated in the N-terminal processing of some MHC class I-binding
epitopes (33). Treatment of macrophages with as high a
concentration of AAF-cmk as 25 µM during presentation of cytosolic
OVA did affect the MHC class I-restricted presentation of OVA to B3 T
cells somewhat (Fig. 4B), but
inhibited its presentation to MHC class II-restricted 13.8 T cells
markedly (Fig. 4D). However, AAF-cmk had no effect on the
presentation of exogenous maleyl-OVA to either B3 or 13.8 T cells (Fig. 4, A and C), confirming the specificity of the
inhibition of the cytosolic pathway. Butabindide also caused some
inhibition of MHC class I-restricted presentation of cytosolic OVA,
although it had no effect on the presentation of exogenous maleyl-OVA
(Fig. 4). However, the inhibition of MHC class II-restricted
presentation of OVA by butabindide was not as marked as that caused by
AAF-cmk (Fig. 4). Taken together, these results indicate a key role
played by nonproteasomal TPP proteases in the MHC class II-restricted
presentation of cytosolic Ags, distinct from their role in MHC class
I-mediated presentation.
|
Ags in the endocytic compartments of the cells are hydrolyzed to
oligopeptides by acid-optimal proteases. Lysosomotropic agents such as
ammonium chloride and chloroquine raise the pH of the distal acidic
vesicles and inhibit proteolysis in these compartments
(8), inhibiting both Ag degradation and Ii proteolysis. To
examine whether the presentation of cytosolic OVA also required further
endolysosomal processing in addition to the cytosolic proteolysis
described above, its sensitivity to chloroquine and ammonium chloride
was next tested. APCs were preincubated with various concentrations of
either drug for 1 h before loading with Ags. They were then loaded
cytosolically with OVA in hypertonic medium or exogenously with either
OVA or maleyl-OVA in isotonic medium and incubated in isotonic medium
for processing for 3 h in the presence or absence of the drugs
before fixation and used for stimulating T cell lines. Neither 100 µM
chloroquine nor 200 mM ammonium chloride had any effect on the
presentation of cytosolic OVA (Fig. 5A), although the presentation
of exogenous fluid-phase OVA was consistently reduced by both
treatments (Fig. 5B). The presentation of exogenous
maleyl-OVA was blocked completely by chloroquine and 200 mM ammonium
chloride (Fig. 5C). In addition, 2 mM ammonium chloride also
effectively blocked presentation of maleyl-OVA on MHC class II (data
not shown), in keeping with the data we have previously reported for
its MHC class I-restricted presentation (26). These
results demonstrate that cytosolic OVA processing is independent of the
acidic environment of the endolysosomal compartments.
|
The existence of a separate and efficient pathway for processing
and presentation of cytosolic proteins on MHC class II suggests that it
may also be used for MHC class II-restricted presentation of endogenous
cellular proteins. To address this issue, we used T cell hybridomas
recognizing a well-characterized peptide epitope generated
constitutively from an endogenous protein source, I-E. The I-E
transmembrane MHC class II polypeptide yields the peptide epitope
52–68 (Eap), from the extracytoplasmic region of the molecule, which
binds to I-Ab (15). This
Eap-I-Ab complex is constitutively present at
very high frequencies (up to 15%) of surface
I-Ab in I-E-expressing cells, such as in the
mouse strains B10.A(3R) or B10.A(5R). Two kinds of Eap-specific T cell
hybridomas have been reported. One group (such as the cell line 1H3
used here) recognizes both APCs given exogenous Eap-containing protein
and APCs constitutively expressing I-Ab and
I-E (15), while the other (such as the cell line Ea6
used here) responds only to presentation of exogenous Eap-containing
protein and not to APCs expressing I-Ab and
I-E constitutively (17). This is because exogenous
processing of Eap-containing protein generates a short peptide (aa
52–66; EapS), while the constitutive endogenous Eap-containing protein
yields a long peptide (aa 52–68; EapL) for presentation on
I-Ab (17).
We used this difference to ask if presentation of Eap on
I-Ab via the cytosolic processing pathway
resembled its presentation from endogenously expressed I-E. To this
end, we used a chimeric fusion protein containing the EapL peptide
between GST and an 11-mer c-myc tag (GST-EapL-myc). We delivered
GST-EapL-myc to H-2b macrophages either
exogenously or cytosolically and tested the responses of the T cell
lines 1H3 (responsive to both EapS and EapL) and Ea6 (responsive only
to EapS). As previously reported (17), the
EapL-I-Ab complex on B10.A(5R) APCs was
recognized by 1H3 but not by Ea6 T cells (Fig. 6A). Similarly, cytosolically
delivered GST-EapL-myc was only recognized by 1H3 and not by Ea6 T
cells (Fig. 6B). However, exogenously processed GST-EapL-myc
was well recognized by both 1H3 and Ea6 T cells (Fig. 6C).
This result indicated that the peptide species generated from the
cytosolically delivered protein resembles that generated from the
processing of an endogenously expressed transmembrane protein.
Furthermore, lactacystin treatment inhibited the presentation of
cytosolically loaded GST-EapL-myc to 1H3, but did not affect the
presentation of exogenously delivered GST-EapL-myc to 1H3 or Ea6 T
cells (Fig. 6, B and C). Inhibitors of
endolysosomal acidification such as chloroquine or ammonium chloride
significantly inhibited the presentation of GST-EapL-myc as an
exogenous protein (Fig. 7A),
but they did not inhibit the presentation of cytosolic GST-EapL-myc
(Fig. 7B) to 1H3 T cells. Thus, the presentation pathways of
cytosolic Eap-containing protein and of the constitutive endogenous
transmembrane protein I-E are parallel in generating a similar fine
specificity of Eap.
|
|
by lactacystin and AAF-cmk
Since cytosolic GST-EapL-myc and endogenous I-E both appear to
yield the EapL epitope, I-Ab-restricted
presentation of constitutively expressed endogenous transmembrane
I-E is likely to be mediated by this novel pathway. If this were the
case, the constitutive expression of the
EapL-I-Ab complex should be sensitive to
inhibitors of the cytosolic pathway of MHC class II-restricted
presentation. We therefore incubated macrophages from B10.A(3R) mice,
which express both I-Ab and I-E, in
lactacystin for varying periods of time before fixing them and using
them as APCs for 1H3 T cells. The reactivity of 1H3 went down
progressively over time when the stimulating B10.A(3R) APCs were
incubated in lactacystin (Fig. 8A). This was a specific loss
of presentation rather than one due to nonspecific inhibitor toxicity,
since even after a 24-h incubation with or without lactacystin,
B10.A(3R) macrophages stimulated OVA-specific 13.8 T cells comparably
if exogenous maleyl-OVA was added during the last 3 h of the
incubation (Fig. 8B). Furthermore, the loss of 1H3
reactivity of B10.A(3R) macrophages incubated for 24 h with either
lactacystin, LLnL, or LLM (Fig. 8C) was specific, since
unchanged levels of EapS-I-Ab complexes
recognizable by Ea6 were generated by these same B10.A(3R) macrophages
from exogenous GST-EapL-myc protein added for the last 6 h of the
24-h incubation in proteasome inhibitors (Fig. 8D). Further
incubation of the APCs for 24 h after washing out lactacystin
allowed them to regain the ability to stimulate 1H3 T cells, confirming
that the effect of lactacystin was not due to nonspecific toxicity
(Fig. 8B).
|
A). When
B10.A(3R) macrophages were incubated in AAF-cmk for 24 h and
pulsed for the last 3 h of this period with maleyl-OVA, AAF-cmk
had no effect on their ability to present maleyl-OVA to 13.8 T cells,
despite abrogating 1H3-stimulatory capacity (Fig. 9B).
Furthermore, removal of AAF-cmk-containing medium followed by another
24 h of incubation allowed these cells to regain their ability to
stimulate 1H3 T cells (Fig. 9B). These data confirm that
nonproteasomal proteases are also specifically involved in the
processing of endogenous transmembrane protein for MHC class
II-restricted presentation.
|
. E4 is a B cell
line derived from the A20 B cell lymphoma line which expresses the
H-2d MHC haplotype and has been transfected with
I-Ab (34). Thus, E4 cells express
both I-Ab and I-E, and therefore constitutively
express the Eap-I-Ab complex. When E4 cells were
incubated in either lactacystin or AAF-cmk for 24 h, their ability
to stimulate 1H3 T cells was significantly lost (Fig. 10). There was no significant loss of
cell viability during this period (data not shown), and removal of the
inhibitors followed by further incubation for 24 h allowed the
cells to regain 1H3-stimulatory capacity (Fig. 10).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Most reported examples of MHC class II-restricted presentation of
endogenous cellular proteins involve either viral proteins (35, 36) or proteins encoded by gene constructs transfected into APC
lines (37). We have taken a different approach by loading
preformed proteins into the cytosol via the osmotic lysis of pinosomes.
The method allows us to make quantitative comparisons between the
presentation of similar quantities of the same protein delivered either
endogenously or exogenously. As against this, the disadvantage is the
possibility that some pinosomes may escape osmotic lysis and Ag
contained in them may be processed conventionally. This would have been
a problem of potential seriousness only if the method did not reveal
any major differences between the two pathways. Since our data clearly
demonstrate a number of such differences, this is not a significant
problem. In further support of this, we have demonstrated that the
delivery of a maleylated protein exogenously to macrophages leads to a
greatly improved efficiency of MHC class II-mediated presentation as
compared with the native form of the protein, and yet the same maleyl
protein is presented very poorly if at all by the cytoplasmic route.
This confirms that the MHC class II-mediated presentation seen after
cytosolic loading of APCs is not due to conventional processing of the
contents of residual unruptured pinosomes. Although the loading of
proteins via osmotic lysis of pinosomes is an unphysiological mode of
delivery, the relevance of the conclusions drawn from such an approach
is validated by the fact that a constitutive endogenous transmembrane
protein, I-E, is also proteasome dependent for its MHC class
II-restricted presentation.
There are conflicting reports about the efficiency of MHC class II-mediated presentation of cytosolic proteins. For example, cytosolically expressed H-2Ld (37) or viral nucleoprotein (36) are poorly presented on MHC class II, but cytosolic influenza virus matrix protein is well presented (10). We find that the introduction of various soluble proteins (OVA, CA, or GST-EapL-myc) into the cytosol by osmotic lysis of pinosomes allows their efficient MHC class II-restricted presentation. The evidence for generation of subtly different peptidic species (EapL vs EapS) from the same protein, depending on whether it is processed endogenously or exogenously, suggests a possible explanation for reported failures to detect MHC class II-mediated presentation of proteins from endogenous sources. In most of these studies, T cells generated by exogenously processed forms of the proteins were used to detect endogenous processing, which may generate either different fine specificities of the same peptide epitope as in the studies here or even completely different peptide epitopes from the same protein (38).
A number of findings suggest that this proteasome-dependent endogenous
pathway is far from being minor or inefficient. First, there is no
major difference in the efficiency of presentation of the same protein
as exogenous fluid-phase protein or as introduced cytosolic protein.
Second, the fine specificity of a peptide epitope generated via this
cytosolic pathway is the same as that generated at high density from a
constitutive endogenous protein and different from that seen upon
exogenous processing. Third, the constitutive expression of this
high-density epitope from its constitutive cellular source, I-E, is
also specifically inhibited by inhibitors of cytosolic proteolysis.
This pathway is thus likely to play a major role in the shaping of the
peptide repertoire of MHC class II on APCs in vivo.
It is increasingly evident that different APC lineages have significant differences in their pathways of Ag processing and presentation (37, 39, 40). We have therefore used normal ex vivo peritoneal macrophage APCs rather than transformed cell lines or mixed APC populations. However, we have also obtained similar results using freshly isolated B cells as well as transformed cell lines of other lineages (data not shown), suggesting that the pathway identified here is a general mechanism rather than a monocytic lineage-specific one.
Proteasomal proteolysis is crucial for generating peptide epitopes from
cytoplasmic proteins for presentation on MHC class I (reviewed in Ref.
1). The role of proteasomal activity in MHC class
II-restricted presentation of endogenous cellular proteins has also
been previously suggested. The lack of presentation of a variant Ag
with a short cytosolic half-life has been used to argue against the
involvement of proteasomes (10). However, inhibitors of
proteasomal activity such as LLnL and lactacystin have recently been
shown to inhibit MHC class II-restricted presentation of proteins such
as hen egg lysozyme (11) and glutamate decarboxylase
(41) from endogenous but not exogenous sources. Similarly,
our results here show that cytosolic presentation is inhibited by the
blocking of amino groups on the Ag by maleylation, indicating that
free amino groups are required for the process involved,
consistent, among other possibilities, with an effect mediated by
ubiquitination, which is a signal for proteasomal degradation
(1). Furthermore, lactacystin, LLnL, and LLM inhibit the
presentation of cytosolically loaded proteins, establishing that
proteasomal degradative activity is essential for MHC class II-mediated
presentation of cytoplasmic proteins.
Crucially, our results further show that the MHC class II-restricted
presentation of an endogenous transmembrane protein, I-E, is also
inhibited by proteasome inhibitors. Prolonged treatment with proteasome
inhibitors may have consequences in addition or further to the
inhibition of proteasomal activity, such as global inhibition of
protein synthesis. Although this formal possibility remains, global
changes such as inhibition of protein synthesis do not explain the
specific inhibition of MHC class II-restricted presentation of
cytosolic proteins, since during the same period, presentation of
exogenously added maleyl-OVA, which requires new protein synthesis
(being cycloheximide sensitive (data not shown)), is not affected at
all. Thus, these data suggest that MHC class II-restricted presentation
of endogenous transmembrane I-E protein is proteasome dependent.
This may be related to the retrotranslocation of membrane and secretory
proteins during their assembly and the role of proteasomes in the
degradation of their poorly folded forms (42, 43, 44, 45, 46). An
alternative explanation may lie in the generation of cytosolic forms of
I-E as defective ribosomal products due to inefficient translocation
or folding for a variety of reasons (47, 48, 49).
Although our results show that proteasomes are required for the generation of peptides for this pathway, they do not suggest that proteasomes are sufficient for peptide generation. It is quite possible that proteasomal degradation is one intermediate event in a chain of degradative steps. Although there have been previous suggestions that some of these steps may in fact be endosomal (18, 19, 20, 21), we find no evidence that this pathway is dependent on endosomal acidification. Nonetheless, it remains possible that translocation of large postproteasomal intermediates could be followed by further processing in the endolysosomal compartments, explaining why the presentation of some peptide-MHC class II epitopes from endogenous sources is sensitive to agents that disrupt lysosome function (30).
Furthermore, we find that nonproteasomal protease activities are likely to play a critical role in the presentation of endogenous proteins on MHC class II, and that this may be different from their role in processing of peptides for loading on MHC class I. There is controversy about the role of nonproteasomal cytosolic enzymatic systems in the generation of peptides loaded on MHC class I. Although cells selected in proteasome inhibitors have been shown to possess normal levels of surface MHC class I suggesting such a possibility (28), proteasomal inhibition in such situations has been shown to be incomplete (50). Since the presentation of cytosolic proteins on MHC class II is completely lactacystin inhibitable, our data do not suggest a proteasome-independent pathway of peptide generation in this event.
However, there have also been suggestions that amino-terminal trimming of proteasomally generated intermediates may be required before loading on MHC class I and that this trimming may be extraproteasomal (29, 33). A number of possibilities have been suggested for the identity and location of these nonproteasomal trimming peptidases. Some of these have been postulated as ER-based mechanisms (29), whereas others possibilities involve cytosolic enzymes such as cytosolic leucine amino peptidase (51) and the more recently identified puromycin-sensitive aminopeptidase and bleomycin hydrolase (33). Our data suggest that such an extraproteasomal degradatory component is also likely be involved in the processing of cytosolic proteins for loading on MHC class II.
It has been reported that the presentation of SIINFEKL, the peptide recognized by H-2Kb-restricted B3 T cells, on OVA-transfected APCs is inhibited by 50 µM AAF-cmk (33). AAF-cmk causes only marginal inhibition of MHC class I-mediated presentation of cytosolically loaded OVA at 25 µM in our experiments. The difference may be due either to the lower concentrations we have used and/or to the possibility that the newly synthesized nascent OVA polypeptide may be handled differently in cytosolic degradation from loaded, preformed mature OVA. However, MHC class II-mediated presentation is inhibited drastically, not only by 25 µM AAF-cmk, but also by 5 µM AAF-cmk (data not shown). This ability of AAF-cmk to inhibit the presentation of OVA on MHC class II even when it does not inhibit it for MHC class I suggests that extraproteasomal protease requirements may indeed be more stringent for peptides destined for MHC class II than for MHC class I. The identity and location of this essential TPP activity identified by AAF-cmk is still unclear. It is possible that TPP-II is not the only activity involved, since butabindide, an inhibitor of TPP-II (31, 32), causes less drastic inhibition of the presentation of cytosolic protein on MHC class II. Whether other related enzymatic activities such as TPP-I, puromycin-sensitive aminopeptidase, or bleomycin hydrolase are also involved remains to be investigated. Although our data do not address the possibility, calpain has also been implicated in the processing of cytosolic Ags for MHC class II-mediated presentation (41).These major differences between the cytosolic pathways for MHC class I and MHC class II may also depend on the fact that there may be multiple categories of proteasomes, perhaps in separate intracellular locations (52) or generating various kinds (perhaps differing lengths) of peptides differentially available for the transport pathways. The identity of these transport mechanisms carrying cytosolically generated peptide fragments into vesicular compartments for loading on MHC class II molecules is thus now an intriguing question, and we have data indicating that a novel cytosol-to-endosome pathway of peptide transport may be involved in the MHC class II-mediated presentation via this endogenous pathway (P. Mukherjee S. Bhatia, A. Dani, A. George, A. Y. Rudensky, S. Mayor, and S. Rath, manuscript in preparation).
There are major implications of such a distinct cytosolic proteolytic pathway for the presentation of endogenous proteins on MHC class II. The rules for generation of these peptides would be different in the two pathways even for the same protein, resulting in nonidentical peptide repertoires for the same self-protein depending on whether it is processed as a cellular protein or an extracellular protein during tissue damage. This in turn may be a factor modulating the education of the T cell repertoire in terms of self-nonself discrimination as well as its failure in autoimmune situations.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Satyajit Rath, National Institute of Immunology, Aruna Asaf Ali Road, New Delhi 110067, India. E-mail address: satyajit{at}nii.res.in
3 Abbreviations used in this paper: ER, endoplasmic reticulum; AAF-cmk, Ala-Ala-Phe-chloromethylketone; CA, conalbumin; Eap, I-E peptide; EapL, I-E peptide (aa 52–68); EapS, I-E peptide (aa 52–66); F-Dex, fluorescein-conjugated dextran; Ii, invariant chain; LLM, N-acetyl-leucinyl-leucinyl-methioninal; LLnL, N-acetyl-leucinyl-leucinyl-norleucinal; SR, scavenger receptor; TPP, tripeptidyl peptidase.
Received for publication January 25, 2001. Accepted for publication July 12, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
chain of the T cell antigen receptor is degraded in the cytosol. Immunity 7:113.[Medline]
chain in the endoplasmic reticulum requires proteasomes and is accelerated by BFA. J. Cell Sci. 111:2271.
can stimulate post-proteasomal trimming of the N-terminus of an antigenic peptide by inducing leucine aminopeptidase. J. Biol. Chem. 273:18734.This article has been cited by other articles:
|
|
 |
|
  J. L. Houlihan, J. J. Metzler, and J. S. Blum HSP90{alpha} and HSP90{beta} Isoforms Selectively Modulate MHC Class II Antigen Presentation in B Cells J. Immunol., June 15, 2009; 182(12): 7451 - 7458. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Solvason, Y.-P. Lou, W. Peters, E. Evans, J. Martinez, U. Ramirez, A. Ocampo, R. Yun, S. Ahmad, E. Liu, et al. Improved Efficacy of a Tolerizing DNA Vaccine for Reversal of Hyperglycemia through Enhancement of Gene Expression and Localization to Intracellular Sites J. Immunol., December 15, 2008; 181(12): 8298 - 8307. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Robila, M. Ostankovitch, M. L. Altrich-VanLith, A. C. Theos, S. Drover, M. S. Marks, N. Restifo, and V. H. Engelhard MHC Class II Presentation of gp100 Epitopes in Melanoma Cells Requires the Function of Conventional Endosomes and Is Influenced by Melanosomes J. Immunol., December 1, 2008; 181(11): 7843 - 7852. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Heemskerk, T. van Vreeswijk, L. A. Veltrop-Duits, C. C. Sombroek, K. Franken, R. M. Verhoosel, P. S. Hiemstra, D. van Leeuwen, M. E. Ressing, R. E. M. Toes, et al. Adenovirus-Specific CD4+ T Cell Clones Recognizing Endogenous Antigen Inhibit Viral Replication In Vitro through Cognate Interaction J. Immunol., December 15, 2006; 177(12): 8851 - 8859. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. S. Taylor, H. M. Long, T. A. Haigh, M. Larsen, J. Brooks, and A. B. Rickinson A Role for Intercellular Antigen Transfer in the Recognition of EBV-Transformed B Cell Lines by EBV Nuclear Antigen-Specific CD4+ T Cells J. Immunol., September 15, 2006; 177(6): 3746 - 3756. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. F. Shih, J. Racz, and P. M. Allen Differential MHC Class II Presentation of a Pathogenic Autoantigen during Health and Disease J. Immunol., March 15, 2006; 176(6): 3438 - 3448. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Delmas, L. Martin, M. Baron, J. A. Nelson, D. N. Streblow, and J.-L. Davignon Optimization of CD4+ T Lymphocyte Response to Human Cytomegalovirus Nuclear IE1 Protein through Modifications of Both Size and Cellular Localization J. Immunol., November 15, 2005; 175(10): 6812 - 6819. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Poea-Guyon, P. Christadoss, R. Le Panse, T. Guyon, M. De Baets, A. Wakkach, J. Bidault, S. Tzartos, and S. Berrih-Aknin Effects of Cytokines on Acetylcholine Receptor Expression: Implications for Myasthenia Gravis J. Immunol., May 15, 2005; 174(10): 5941 - 5949. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Dani, A. Chaudhry, P. Mukherjee, D. Rajagopal, S. Bhatia, A. George, V. Bal, S. Rath, and S. Mayor The pathway for MHCII-mediated presentation of endogenous proteins involves peptide transport to the endo-lysosomal compartment J. Cell Sci., August 15, 2004; 117(18): 4219 - 4230. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. R. Muller, G. Tsakou, F. Grunebach, S. M. Schmidt, and P. Brossart Induction of chronic lymphocytic leukemia (CLL)-specific CD4- and CD8-mediated T-cell responses using RNA-transfected dendritic cells Blood, March 1, 2004; 103(5): 1763 - 1769. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Sinnathamby and L. C. Eisenlohr Presentation by Recycling MHC Class II Molecules of an Influenza Hemagglutinin-Derived Epitope That Is Revealed in the Early Endosome by Acidification J. Immunol., April 1, 2003; 170(7): 3504 - 3513. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Powell Jr., L. C. Eisenlohr, and J. L. Rothstein A Thyroid Tumor-Specific Antigen Formed by the Fusion of Two Self Proteins J. Immunol., January 15, 2003; 170(2): 861 - 869. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
