| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Laboratories of
*
Viral Diseases and
Immunology, National Institute of Allergy and Infectious Diseases, Bethesda, MD 20892; and
Mount Sinai School of Medicine of City University of New York, Department of Pharmacology, New York, NY 10029
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It is current dogma that the vast majority of cytosolic peptides are
generated by proteasomes (4, 5). Proteasomes are abundant, ubiquitous
structures located in the nucleus and cytosol, and are essential for
cell viability (4, 6). The 20S proteasome
(Mr 
700,000) constitutes the
catalytic core of a larger 26S complex involved in ubiquitin
(Ub)-dependent and independent pathways of intracellular protein
degradation (4). 26S proteasomes are produced by the association of 19S
regulatory subunit with each end of a 20S proteasome. 11S regulatory
subunits can replace 19S regulators to form a distinct protease with
increased activity against some substrates (5).
The 20S proteasome is composed of a total of 28 subunits, arranged in 4
heptameric cylinders. Subunits within a cylinder are highly homologous,
and are grouped as 
and ß subunits, producing a structure
consisting of
7ß7ß77. Only
the ß subunits are believed to be catalytically active. The entrance
to the channel of the 20S proteasome from Thermoplasma has a
diameter of approximately 15 Å (7). This is thought to prevent access
to all but unfolded polypeptides. In the yeast 20S proteasome, the ends
are closed (8), leaving only narrow side windows into the catalytic
channel. Numerous catalytic activities have been described for the 20S
proteasome using small fluorogenic substrates (9). In the present
study, we use the tripeptide aldehyde zLLL, which inhibits all of the
defined activities of 20S proteasomes (10) and lactacystin (11),
produced by Streptomyces, which covalently binds to ß
subunits, preferentially inhibiting the chymotryptic-like and
tryptic-like activities. Unlike zLLL and other peptide aldehyde
inhibitors, which inhibit numerous other cellular proteases,
lactacystin is known only to interfere with proteasome activity.
Several lines of evidence implicate the proteasome in the production of class I-binding peptides. The initial evidence was the identification of genes in the class II region of MHC encoding the LMP2 and LMP7 proteasome ß subunits (12). Their expression is increased by cytokines that coordinately induce expression of defined components of the class I Ag-processing machinery (13). These subunits substitute for constitutively expressed alternative subunits, modifying the specificity of the proteasome, although the nature of the alterations is contentious (14, 15, 16, 17). Marginal effects on T cell responses have been described for mice lacking either LMP2 (18) or LMP7 (19). Three studies reported a correlation between the degradation rate of cytosolic proteins (presumably proteasome mediated), either biosynthesized by APCs (20, 21) or exogenously loaded into APCs (22), and the efficiency of antigenic peptide generation, while a fourth study reported no such correlation with biosynthesized Ags (23). Evidence has been presented showing a requirement for an intact Ub conjugation pathway in the presentation of some Ags (24), but we found no such requirement (25). More direct evidence came with the use of tripeptide aldehyde proteasome inhibitors that interfered with class I biogenesis and completely blocked the presentation of certain determinants (26, 27, 28, 29). However, class I biogenesis is blocked only partially by proteasome inhibitors, and we have found that there is a wide variation in the effect of the inhibitors in the presentation of determinants from exogenous and endogenous influenza virus Ags, ranging from total inhibition to enhancement (10). Another suggestive, although indirect, line of evidence implicating the proteasome comes from cell-free experiments. In these studies, the efficiency of Ag presentation in vivo of different determinants, in various contexts, was positively correlated with in vitro degradation of substrates by purified proteasomes (30, 31). In some instances, these digests yielded the antigenic peptide as a final product. Most recently, it was reported that the 11S regulator is induced by cytokines (32), and that Ag presentation can be enhanced by expression of 11S regulators (33, 34).
Altogether the evidence is compelling that proteasomes are involved in the production of class I-binding peptides, but the extent of proteasomal involvement is uncertain. In the present study, we use proteasome inhibitors to examine the role of proteasomes in generating peptides from metabolically stable and unstable proteins produced by recombinant vaccinia viruses (rVVs).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The peptide aldehyde inhibitors used, Cbz-Leu-Leu-Leucinal
(zLLL) and Cbz-Leu-Leucinal (zLL), were synthesized as described (35).
Lactacystin was obtained from Dr. E. J. Corey (Harvard University,
Cambridge, MA). Brefeldin A (BFA) was from Sigma (St. Louis, MO). Abs
to hsc70 (rat IgG1 mAb 1B5), heat shock protein 90 (hsp90) (rat IgM Ab
1R2D12p90), and grp94 (rat IgG2a Ab 9G10) were from StressGen
(Victoria, British Columbia, Canada), and a mouse IgG1 Ab specific for
hsp70 family of proteins (clone 3a3) was from Affinity Bioreagents
(Neshanic Station, NJ). GammaBind Plus Sepharose was from Pharmacia
(Piscataway, NJ), protein A-agarose from Pierce (Rockford, IL), and
mouse anti-rat 
Sepharose from Zymed (South San Francisco,
CA).
Cells, virus stocks, and Abs
The thymidine kinase-deficient human osteosarcoma cell line 143B and the fibroblastoid cell line L929 (H2k), transfected with the gene encoding for either Kd (L-Kd) (36) or Kb (L-Kb) (a kind gift from Dr. J. Sheil, West Virginia University, Morgantown, WV), were maintained in DMEM supplemented with 7.5% (v/v) of FCS and 2 mM L-glutamine, in air/CO2 (91%/9%). The P815 mastocytoma cell line was maintained in RPMI 1640, with the same supplements, and in air/CO2 (94%/6%). Long- and short-term TCD8+ cultures maintained in Iscove’s modified Dulbecco’s medium supplemented with 2-ME in air/CO2 (91%/9%). VV stocks were prepared from infected 143B cells and used as lysates after low speed centrifugation. Vaccinia virus recombinants (rVV) expressing the nucleoprotein from influenza virus A/Puerto Rico/8/34 (VV-NPPR8) (37) or A/NT/60/68 (VV-NPNT60)(20), as well as a secreted form of the former (VV-SNP) (38), and a chimeric form of the latter (VV-Ub-Arg-NP) (20) have been previously described. rVVs expressing N protein from the Indiana strain of vesicular stomatitis virus (VV-VSV-N) or a minigene containing the minimal Kb-restricted determinant of this protein (VV-NM52–59) has also been described (39, 40). rVVs expressing minigenes containing the minimal determinants from NP restricted by Kk (VV-NPM50–57) or Kd (VV-NPM147–155) were included, and have also been described (41). rVVs expressing full-length OVA (VV-OVA) or a minigene expressing the minimal determinant (VV-OVAM257–264) have been described (37). The mAb 25-D1.16, specific for the Kb-OVA257–264 complex (42), was used as a tissue culture supernatant.
Viral infection, metabolic labeling, and immunoprecipitation
Cells were incubated for 45 to 60 min with the different
inhibitors (or DMSO as a control) at 37°C before infection with rVV
at a multiplicity of infection of 10 plaque-forming units/cell. For the
analysis of stress protein induction, uninfected cells were incubated
with 5 µM zLLL for 5 h before labeling. For radiolabeling, cells
were incubated at 37°C for 30 min in methionine-free medium, and then
radiolabeled in the same medium supplemented with
[35S]methionine (Amersham, Arlington Heights, IL)
for 5 min. In all instances, media contained the appropriate
inhibitors. Cells were then lysed in extract buffer (50 mM Tris-HCl, pH
7.3, 100 mM NaCl, 1 mM EDTA, and 2% Triton X-100, with a mixture of
protease inhibitors (Complete, from Boehringer Mannheim, Indianapolis,
IN)). The nuclear fraction was discarded following centrifugation, and
the extracts were analyzed by SDS-PAGE (43), under reducing conditions,
either as lysates or after immunoprecipitation. For
immunoprecipitation, Abs were coupled to either GammaBind Plus
Sepharose (anti-hsc70, -hsp70, and -grp94), or anti-rat Ab
coupled to Sepharose (anti-hsp90). Samples were adjusted to contain
the same volume and amount of radioactivity before immunoprecipitation.
Samples were precleared with the appropriate beads coupled with serum
autologous to the Ab used. After immunoprecipitation, the beads were
washed as indicated (44), except for the anti-hsp90 Ab-coated
beads, which were washed twice in MENG buffer (25 mM MOPS, 1 mM EDTA,
0.02% NaN3, pH 7.5, and 10% glycerol) (45). The beads
were then boiled in sample buffer and analyzed by SDS-PAGE under
reducing conditions. For the study of the Ub-Arg-NP stability, cells
were infected for 2 h before labeling with
[35S]methionine for 5 min. Aliquots were incubated for
different times in medium supplemented with 10 mM cold methionine.
Cells were lysed in boiling sample buffer for electrophoresis,
containing 1% ß-mercaptoethanol, and Complete. The same
volume/sample was loaded onto the gels. Gels were quantitated by
analysis with a PhosphorImager (Molecular Dynamics, Sunnyvale, CA)
using the ImageQuant software provided. The counts obtained from the
band corresponding to NP were normalized using those of a stable VV
protein as a standard (this band is marked in Fig. 1
A with
an asterisk). The image was arranged and labeled with Adobe Photoshop
(Adobe Systems, Mountain View, CA), and printed with a digital printer
(Fujix 3000 Pictrography, Fuji Photofilm USA, Elmsford, NY).
|
L-Kb cells incubated in the presence or absence of proteasome inhibitors were infected with VV-NP, VV-OVA, or VV-OVAM257–264 for 5 h. Cells were then washed with ice-cold saline containing NaN3 and stained with 25-D1.16 Ab (42), biotin-conjugated goat F(ab')2 anti-mouse IgG, and FITC-conjugated streptavidin (Jackson ImmunoResearch, West Grove, PA). The amount of fluorescence associated with live cells was quantitated using a FACScan cytofluorograph (Becton Dickinson, Mountain View, CA).
Mice and generation of specific TCD8+
BALB/c (H2d) and C3H (H2k) mice, from Taconic Farms (Germantown, NY), were immunized with 5 x 107 plaque-forming units of VV-NP in 0.5 ml of PBS containing 0.2% BSA. At least 3 wk after immunization, spleens were harvested and restimulated in vitro for 1 wk with autologous splenocytes either infected with VV, or pulsed with synthetic peptides SDYEGRLI (Kk) or TYQRTRALV (Kd), provided by the Biologic Resources Branch (National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD). For quantitation of antigenic peptides in acid extracts from infected cells, TCD8+ lines were generated as described (46).
Cytotoxicity assays
Cytotoxicity assays were performed as previously described (47). Cells were preincubated with the inhibitors for 30 min before infection with rVVs. At different times after infections, the inhibitors were replaced by BFA at 5 µg/ml to block additional Ag presentation. Cells were kept in BFA-containing media for the remainder of the assay.
Peptide extraction and quantitation
This procedure has been described (46). Cells preincubated for 1 h with 10 µM lactacystin were infected for 6 h in 10 µM lactacystin. After acid extraction, low Mr material was fractionated by reverse-phase HPLC, as previously reported, collecting 0.25-ml fractions. The amount of peptide in these fractions was determined by their ability to sensitize target cells for recognition by specific TCD8+, extrapolating the concentration from titration curves obtained with measured amounts of synthetic peptides.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To gain a quantitative measure of the effectiveness of zLLL and
lactacystin on proteasomal degradation, we examined their ability to
inhibit degradation of a model substrate. To control for effects of
zLLL on nonproteasomal proteases, we used zLL. zLL is a more potent
calpain inhibitor than zLLL, similar in potency in inhibiting lysosomal
proteases, and at least 250 times less active on proteasomes (10, 48).
Townsend et al. (20) produced a rVV-expressing NP from NT60 influenza
virus fused to the COOH terminus of Ub with an intervening Arg. Ub is
cotranslationally removed from NP, leaving Arg as the
NH2-terminal residue of NP. Arg-NP obeys the N-end
rule defined by Varshavsky and colleagues (49) and is rapidly degraded
(20). As seen in Figure 1
(A and B), Ub-Arg-NP is unstable,
exhibiting a t1/2 of less than 20
min, whereas NP is stable throughout the chase period. zLL had no
significant effect on protein degradation. In contrast, 10 µM zLLL or
100 µM lactacystin completely blocked degradation, while lactacystin
at 10 µM had an intermediate effect on protein stability. Consistent
with these results, both zLLL and lactacystin (but not zLL) induced the
accumulation of polyubiquitinated cellular proteins in uninfected
L-Kd cells, as determined by Western blotting of detergent
lysates using anti-Ub Abs (results not shown). This is in agreement
with previous reports on the accumulation of ubiquitinylated proteins
in the presence of proteasome inhibitors (50, 51).
Proteasome inhibitors block presentation of OVA
The processing of the Kb-restricted determinant
from OVA (OVA257–264) has been reported to be proteasome
dependent, as determined by activation of a T cell hybridoma (26). To
confirm that the proteasome inhibitors were capable of blocking Ag
presentation in rVV-infected L929 cells, L-Kb cells
incubated with proteasome inhibitors were infected with rVVs producing
OVA or the cytosolic minigene product, and the expression of cell
surface OVA257–264-class I complexes was quantitated by
cytofluorography using a mAb highly specific for the complex (42). As
previously reported, this method easily detects quantitative
differences in the presentation of the OVA257–264 peptide
produced in different contexts: cells synthesizing cytosolic
OVAM257–264 express far more peptide-Kb
complexes on the cell surface than cells expressing OVA (Fig. 2). (All rVVs expressing cytosolic
minigene products synthesize peptides with an additional Met residue to
initiate protein synthesis: this residue is probably rapidly removed by
Met-aminopeptidase (52).) zLLL nearly completely blocked the generation
of OVA257–264-Kb complexes from OVA. The
control inhibitor, zLL, had little effect on peptide generation,
implicating proteasomes in this process. This is confirmed by the
inhibitory effect of lactacystin, which at 10 µM blocked complex
formation to the same extent as zLLL. In contrast to the results with
full-length OVA, the inhibitors had little effect on the generation of
complexes from the cytosolic minigene product. The most straightforward
explanation for these findings is that proteasome inhibitors block the
formation of OVA257–264-Kb complexes from OVA
by preventing proteasomes from contributing to the liberation of the
peptides.
|
Cells present two immunodominant NP peptides to NP-specific TCD8+: NP147–155 in association with Kd (53), and NP50–57 in association with Kk (54). We studied the effects of proteasome inhibitors on the presentation of these determinants in the following polypeptide contexts: full-length NP derived from PR8 or NT60 (termed NPPR8 and NPNT60, respectively), PR8 NP targeted to the ER by the IFN-ß leader sequence (termed SNP), Ub-Arg-NP, and cytosolic minigene products. Despite its efficient translocation to the ER, generation of peptides from SNP is strictly TAP dependent (38), and therefore requires the action of cytosolic proteases.
We previously showed that 6 h after infection with rVVs encoding the cytosolic minigene products, cells express approximately 50,000 Kd and Kk molecules with the respective peptides, as opposed to 30 copies/cell of NP147–155 and 1,800 copies of NP50–57 after infection with VV-NPPR8 (46). The extreme overexpression of the cytosolic minigenes makes them insensitive to even major effects of proteasome inhibitors on viral gene expression, at least as detected by TCD8+-based assays (this limitation applies to a much lesser extent to cytofluorographic analysis using mAbs specific for class I peptide complexes, as described above). In the experiments shown below, therefore, rVVs expressing cytosolic minigene products are included to demonstrate the maximal levels of lysis possible and to serve as controls for the specificity of the TCD8+ populations used.
rVV-infected L-Kd cells were used as target cells in a
51Cr release assay with TCD8+ specific for
either NP50–57 (Fig. 3A) or
NP147–155 (Fig. 3B). Cells were infected
without inhibitor present or in the presence of 10 µM zLLL, 10 µM
zLL, or lactacystin at 10 or 100 µM. BFA was added to cells 250 min
postinfection to block additional generation of cell surface class I
peptide complexes (55, 56). This allowed us to remove the proteasome
inhibitors from the TCD8+ assay, avoiding untoward effects
of the inhibitors on TCD8+ function or 51Cr
release from target cells. As we previously observed (38), targeting NP
to the ER decreased the efficiency of generating
NP147–155, while having little effect on the generation of
NP50–57. The presentation of NP147–155 from
NPNT60 was reduced relative to NPPR8. Since the
sequence of the determinant is identical between the two, this
represents an effect of flanking sequences or protein handling on
peptide liberation.
|
Lactacystin at 10 µM had little effect on presentation of NP50–57, and enhanced the presentation of NP147–155 from SNP and NPNT60, demonstrating the proteasome-specific nature of the latter effect. For NP50–57, 100 µM lactacystin partially blocked presentation from SNP, and did not affect presentation from NPPR8 or NPNT60. Lactacystin (100 µM) did not affect presentation of NP147–155 from either NPPR8 or NPNT60, and enhanced presentation from SNP.
In parallel experiments, we found that 100 µM lactacystin inhibits cell surface expression of rVV-encoded mouse CD54, human CD23, and human CD4, respectively, by 54, 52, and 38% after a 6-h infection, as determined using directly conjugated mAbs under saturating conditions. Thus, the inhibitory effects of 100 µM lactacystin on presentation of NP50–57 may in part be due to nonspecific effects on rVV gene expression or exocytosis of integral membrane proteins. The difference between 10 and 100 µM in enhancing NP147–155 presentation may reflect, to some extent, this nonspecific inhibition.
These findings were extended to another type of target cell (Fig. 4). P815 cells were coinfected with rVVs
expressing the target Ag of interest and a rVV expressing
Kb. This allowed us to examine the presentation of another
viral Ag, the nucleocapsid (N) of vesicular stomatitis virus, in
addition to NPPR8. P815 cells treated with 25 µM zLLL
demonstrated enhanced presentation of the NP147–155
peptide from NPPR8. In the same experiment, the
presentation of the immunodominant N peptide (N52–59) to
Kb-restricted TCD8+ was severely compromised by
zLLL. This shows that the inability of proteasome inhibitors to block
generation of NP147–155 is not a peculiarity of
L-Kd cells, and provides further evidence that proteasome
inhibitors have disparate effects on different determinants.
|
These findings indicated that blocking proteasome activity enhances presentation of the NP147–155 determinant. This could be due to enhanced generation of NP147–155-Kd complexes or secondary effects due to alterations in other peptide class I complexes or other molecules involved in T cell recognition and lysis of target cells. We therefore measured the amount of peptide class I complexes present in cells infected with either VV-NPPR8 or VV-SNP, in the presence or absence of 10 µM lactacystin. Acid-soluble peptides prepared from 5 x 108 cells were fractionated by HPLC, and the fractions containing antigenic peptides were titrated. Peptide concentrations were determined by comparative titration using a known quantity of synthetic peptide. We previously showed that virtually all peptides recovered by this method are derived from Kd molecules (46).
Peptide quantitation confirmed that the NP147–155 is
produced more efficiently from NPPR8 than from SNP:
targeting NP to the ER reduced peptide production fourfold (Table I). Incubation of cells with 10 µM
lactacystin enhanced peptide generation from both NPPR8 and
SNP by approximately fourfold. This provides a direct demonstration of
enhanced peptide generation following treatment with a proteasome
inhibitor.
|
) revealed that
lactacystin had little effect on the generation of
Kd-restricted antigenic peptides eluting at fractions
91–92 and 98. Interestingly, lactacystin inhibited the generation of a
VV peptide eluting at fraction 103, while enhancing the amount of
peptide eluting at fraction 107. These findings provide biochemical
evidence that supports the functional data indicating that proteasome
inhibition has disparate effects on peptide generation covering a
spectrum from inhibition (peptide in fraction 103) to enhancement
(peptide in fraction 107). We cannot rule out at this point that the
peptides in fractions 103 and 107 are related (e.g., the same core
antigenic peptide, but with different flanking residues, or two
overlapping epitopes, as has been described in a different system
(57)), and that lactacystin enhances the production of one at the
expense of the other.
|
The dissociation between cellular proteasome activity and peptide
generation prompted us to investigate the role of proteasomes in
generating peptides from Ub-Arg-NP. Despite the total inhibition of
Ub-Arg-NP degradation by zLLL at 10 µM, as shown in Figure 1, the
generation of NP50–57 and NP147–155 at 5
µM of the inhibitor was unimpeded (Fig. 6). Increasing the concentration of zLLL
to 15 µM still did not significantly block presentation. This
demonstrates that antigenic peptides can be produced in manner that is
independent of the measurable proteasome-specific degradation of a
full-length substrate.
|
In the course of characterizing the effects of proteasome
inhibitors on VV-gene expression, we noted a major alteration in the
expression of proteins in both VV-infected and uninfected cells. As
seen in Figure 7A, treatment
with zLLL strongly inhibited actin biosynthesis, and enhanced synthesis
of several other proteins. As two of the most prominent bands migrate
with molecular mass of 70 kDa, we conjectured that proteasome
inhibitors induced a stress response. This was confirmed using Abs
specific for different chaperones to collect specific gene products:
hsc70 (heat-shock cognate protein (hsc), a cytosolic chaperone
expressed at constitutively low levels and induced by stress), hsp70 (a
family of cytosolic chaperones including proteins not expressed
constitutively, as well as hsc70), hsp90 (another cytosolic chaperone),
and grp94 (an ER resident chaperone, also known as gp96) were all
induced by zLLL (Fig. 7B). Hsp90 was induced only
marginally in these experiments, although its induction in other
experiments was somewhat stronger. Similar results were obtained using
lactacystin (not shown).
|
C). This suggests that the enhanced presentation of
NP147–155 is not a consequence of the heat-shock response
induced by the proteasome inhibitors. The heat-shock response differs
from that induced by blocking the proteasome activity, since grp94 and
probably other stress proteins induced by proteasome inhibitors are not
induced by heat shock (60). It remains plausible that these proteins
are involved in the enhanced presentation of
NP147–155. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In the present study, we extend these findings using rVVs to express viral Ags. We found that for VV-produced cytosolic or ER-targeted NP, proteasome inhibitors have little effect on the generation of NP50–57, and enhance generation of NP147–155. A crucial question in analyzing these results is whether the inhibitors were used at sufficient concentrations to completely block proteasome activity. While this question cannot be definitively answered at present, our results argue toward the positive. We show that under the same conditions in which NP presentation is unaffected or enhanced, the generation of Kb-restricted determinants from OVA or N is severely compromised. Moreover, we also demonstrate a complete block of the degradation of Ub-Arg-NP. The latter finding is important as it shows that the inhibitors must be effective in 90% or more of the cells.
The resistance of NP147–155 generation to proteasome inhibitors is consistent with the recent findings of Yellen-Shaw et al., who found that the proteasome inhibitor N-acetyl-LLnL used at 50 µM enhances the generation of this determinant from rVV expressing different forms of NP (61). Under the same conditions, the generation of a determinant from another viral protein was inhibited (62). These findings are, however, apparently at odds with those of Cerundolo et al., who found that lactacystin inhibited presentation of VV-Ub-Arg-NP, VV-NP327–498, and a number of influenza virus-encoded proteins to TCD8+ (63).
To examine presentation of VV-encoded NP constructs, Cerundolo et al. used TCD8+ specific for a Db-restricted peptide (residues 366–374). Unlike the Kd- and Kk-restricted NP determinants we have studied, the generation of this peptide requires expression of LMP2 and LMP7 proteasome subunits (64). It is not surprising, therefore, that generation of this peptide is proteasome dependent.
Cerundolo et al. treated cells with lactacystin at concentrations of 10 or 100 µM overnight before infection with influenza virus. Given the pleiotropic effects of proteasome inhibition on other aspects of cellular metabolism, it is not possible to eliminate the possibility that presentation of viral Ags was not due to indirect effects (e.g., altering delivery of substrates to other proteases due to modifications in chaperones, or modifying peptide handling in the ER). Moreover, as mentioned above, we found that preincubating L-Kd cells even for as short as 45 min with 100 µM lactacystin reduces the cell surface expression of a number of VV-encoded reporter proteins by approximately 50%. Additional experiments revealed that overnight treatment of cells with lactacystin at 10 or 100 µM greatly reduced influenza virus gene expression, as measured by cell surface expression of viral neuraminidase (expression reduced more than 10-fold at 10 µM and to background values at 100 µM). Increasing the concentration of the peptide aldehydes above 10 µM also results in reduction of rVV gene expression.
These findings reinforce the general principle that inhibitors should
be used at the minimal concentration required to achieve inhibition of
the target process. Use of inhibitors at higher concentrations greatly
increases the chances of inhibiting unrelated, difficult to identify,
targets. Even in the case of lactacystin, the observation that only
proteasome ß subunits are covalently labeled by following incubation
of cells with [3H]lactacystin (65, 66) does not
constitute proof of its specificity, particularly when the abundance of
proteasomes (1% of cellular proteins) relative to other cytosolic
proteases is considered. Indeed, it was reported recently that
lactacystin inhibits the activity of a lysosomal protease (67). Despite
these problems with nonspecific inhibition of cellular processes
related to Ag presentation, it is remarkable that Cerundolo found that
the presentation of an HLA-A1-restricted NP peptide persisted in cells
treated overnight with 500 µM lactacystin. Thus, taken as a whole,
the results of Cerundolo et al. are basically in accord with our
findings that some determinants are presented in a
proteasome-independent manner.
This conclusion is fully supported by the work of Hughes et al., who found that assembly of human class I molecules was only partially compromised by the proteasome inhibitors Cbz-LLF and N-acetyl-LLnL (68). Analyzing radiolabeled peptides from purified class I molecules by HPLC, they noted that peptide yield was variably effected, and that the relative abundance of individual peptide species was greatly modified. This latter finding is the biochemical equivalent of our demonstration that lactacystin modifies the profile of VV-derived antigenic peptides recovered by HPLC.
An unexpected effect of the proteasome inhibitors was their enhanced generation of NP147–155. This could reflect a number of molecular processes. First, since the proteasome is a major source of class I-binding peptides, peptides produced in a proteasome-independent manner will have less competition for all of the events entailed in locating and binding to class I molecules. This may be particularly important for NP147–155, which does not display a particularly high affinity for Kd molecules relative to other peptides (38). Second, the proteasome may actively destroy NP147–155 or its precursors in the cytosol. In a collaborative study with Tevethia and colleagues, we recently found an rVV-encoded cytosolic peptide whose presentation could only be detected by incubating cells with zLLL (69). Ossendorp et al. reported that substitutions in an antigenic viral peptide that result in intradeterminant cleavage by purified 20S proteasomes also abrogate presentation of the peptide from the full-length protein (31). Third, the inhibitors may modify proteasome activity in a manner favoring the creation of some determinants. This would be the case if one of the proteasome activities cleaves within the epitope, as has been suggested by Yellen-Shaw et al. (61). It is possible that in vivo the different proteasome activities are not equally sensitive to the inhibitors. This is particularly relevant with lactacystin, which has been shown in vitro to preferentially block the chymotryptic- and tryptic-like activities of 20S proteasomes (11). Fourth, as discussed below in more detail, enhanced generation may result from secondary effects of proteasome inhibition.
Probably the most important finding in the present study is the persistence of peptide generation from Ub-Arg-NP in the presence of proteasome inhibitors at sufficient concentration to completely block its degradation. The discrepancy between protein degradation rates and peptide generation was originally noted by Goth et al., who studied the generation of antigenic peptides from N-end rule substrates biosynthesized by APCs (23). Together, these findings indicate that proteasomal digestion of ubiquitinated proteins does not always result in the production of antigenic peptides. Indeed, the relevance of Ub targeting of proteins to proteasomes in antigenic processing is uncertain; in a prior study, we failed to detect diminished generation of antigenic peptides in cells with a compromised Ub-conjugation pathway (25). The observation that the generation of antigenic peptides is unrelated to fate of full-length conformed proteins is consistent with our recent proposal that antigenic peptides can be derived from defective ribosomal products: truncated or misfolded proteins that are shunted to peptidogenic proteases (70).
The major question raised by our findings is the nature of the protease(s) responsible for the generation of antigenic peptides in the presence of proteasome inhibitors. As mentioned above, we cannot eliminate the possibilities that either a subset of proteasomes or activities present in proteasomes are resistant to the inhibitors. Also plausible is that sensitive determinants require multiple rounds of proteasomal cleavage, while resistant determinants require fewer cycles. Arguing against this are the findings of Yang et al. that zLLL can block presentation of a peptide consisting only of 21 residues (27).
If not proteasomes, then what liberates antigenic peptides from NP and other proteins resistant to the effects of proteasome inhibitors? The other major class of defined cytosolic proteases is that of the calpains, a family of cysteine proteinases that are activated by Ca2+. Some calpains are enzymatically active at µM concentrations of Ca2+, meaning that they could be constitutively active in the cytosol. None of the calpains examined has the broad specificity required to be a general source of antigenic peptides. Moreover, the inability of zLL, a potent calpain inhibitor, to block presentation of proteasome-resistant determinants also disfavors the involvement of calpains in Ag processing. Other cytosolic proteinases have been defined, but seem to be far too specific in their choice of substrates (e.g., proteases involved in programmed cell death), or limited in their abilities to degrade proteins (prolyl endopeptidase and zinc-metallopeptidases 24.15 and 24.16 are broadly active, but prefer oligopeptides to proteins), to be significant contributors of antigenic peptides. Basically then, we are left with the exciting, if troubling, possibility of an undefined major cytosolic proteinase of broad specificity.
We found that proteasome inhibitors induced a stress response in APCs. Zhou et al. recently reported that proteasome inhibitors induced hsp70 (71) by enhancing transcription, and noted that hsp90 and Ig heavy chain binding protein were not induced. However, Bush et al. (72) have shown recently that ER chaperones, including Ig heavy chain binding protein, can be induced by proteasome inhibitors. Our results are in agreement with these findings, as we found enhanced synthesis of the ER chaperone grp94, as well as of the cytosolic hsp70, hsc70, and hsp90. The induction of grp94 may be related to the proteasome-dependent degradation of misfolded ER proteins (73). It would be surprising if the induction of stress proteins by proteasome inhibitors had no effect on Ag processing. For example, molecular chaperones could be involved in targeting proteins to peptidogenic proteases. Two of the chaperones induced by proteasome inhibitors (hsp70 and grp94 (alternatively termed gp96)) have been reported to bind antigenic peptides (58, 74), and it has been suggested that they function to transport antigenic peptides from their sites of generation to class I molecules (75). Induction of stress proteins may enhance these functions, or could even have the opposite effect if the availability of the chaperones for peptides is reduced if induced proteins were overburdened with chaperoning defective cellular proteins.
In summary, our results question the simple model that proteasomes are involved in the generation of all antigenic peptides from cytosolic precursors. They also raise the issue of whether some of the effects of proteasome inhibitors on Ag presentation represent secondary effects that stem from the crucial role that proteasomes play in cellular metabolism.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Drs. Jack R. Bennink and Jonathan W. Yewdell, Room 213, Building 4, National Institutes of Health, Bethesda, MD 20892-0440. E-mail address:
3 Abbreviations used in this paper: TCD8+, CD8+ T lymphocytes; ER, endoplasmic reticulum; BFA, brefeldin A; N, nucleocapsid protein from vesicular stomatitis virus; NP, influenza virus nucleoprotein; SNP, nucleoprotein targeted to the secretory pathway; Ub, ubiquitin; VV, vaccinia virus; hsp, heat shock protein; grp, glucose-regulated protein; hsc, heat-shock cognate protein.
Received for publication August 5, 1997. Accepted for publication January 23, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
in antigen presentation. Nature 381:166.[Medline]
This article has been cited by other articles:
|
|
 |
|
  C. Smith, N. Wakisaka, T. Crough, J. Peet, T. Yoshizaki, L. Beagley, and R. Khanna Discerning regulation of cis- and trans-presentation of CD8+ T-cell epitopes by EBV-encoded oncogene LMP-1 through self-aggregation Blood, June 11, 2009; 113(24): 6148 - 6152. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Goldwich, S. S. C. Hahn, S. Schreiber, S. Meier, E. Kampgen, R. Wagner, M. B. Lutz, and U. Schubert Targeting HIV-1 Gag into the Defective Ribosomal Product Pathway Enhances MHC Class I Antigen Presentation and CD8+ T Cell Activation J. Immunol., January 1, 2008; 180(1): 372 - 382. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Marcilla, J. J. Cragnolini, and J. A. Lopez de Castro Proteasome-independent HLA-B27 Ligands Arise Mainly from Small Basic Proteins Mol. Cell. Proteomics, May 1, 2007; 6(5): 923 - 938. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Guil, M. Rodriguez-Castro, F. Aguilar, E. M. Villasevil, L. C. Anton, and M. Del Val Need for Tripeptidyl-peptidase II in Major Histocompatibility Complex Class I Viral Antigen Processing when Proteasomes are Detrimental J. Biol. Chem., December 29, 2006; 281(52): 39925 - 39934. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. J. Wherry, T. N. Golovina, S. E. Morrison, G. Sinnathamby, M. J. McElhaugh, D. C. Shockey, and L. C. Eisenlohr Re-evaluating the Generation of a "Proteasome-Independent" MHC Class I-Restricted CD8 T Cell Epitope J. Immunol., February 15, 2006; 176(4): 2249 - 2261. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. N. Golovina, S. E. Morrison, and L. C. Eisenlohr The Impact of Misfolding versus Targeted Degradation on the Efficiency of the MHC Class I-Restricted Antigen Processing J. Immunol., March 1, 2005; 174(5): 2763 - 2769. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. C. Reed, B. Berwin, J. P. Baker, and C. V. Nicchitta GRP94/gp96 Elicits ERK Activation in Murine Macrophages: A ROLE FOR ENDOTOXIN CONTAMINATION IN NF-{kappa}B ACTIVATION AND NITRIC OXIDE PRODUCTION J. Biol. Chem., August 22, 2003; 278(34): 31853 - 31860. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Lomas, E. Hanon, Y. Tanaka, C. R. M. Bangham, and K. G. Gould Presentation of a new H-2Dk-restricted epitope in the Tax protein of human T-lymphotropic virus type I is enhanced by the proteasome inhibitor lactacystin J. Gen. Virol., March 1, 2002; 83(3): 641 - 650. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Ota, T. Ono, A. Morita, A. Uenaka, M. Harada, and E. Nakayama Cellular Processing of a Multibranched Lysine Core with Tumor Antigen Peptides and Presentation of Peptide Epitopes Recognized by Cytotoxic T Lymphocytes on Antigen-presenting Cells Cancer Res., March 1, 2002; 62(5): 1471 - 1476. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. C. Norbury, M. F. Princiotta, I. Bacik, R. R. Brutkiewicz, P. Wood, T. Elliott, J. R. Bennink, and J. W. Yewdell Multiple Antigen-Specific Processing Pathways for Activating Naive CD8+ T Cells In Vivo J. Immunol., April 1, 2001; 166(7): 4355 - 4362. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Yewdell, U Schubert, and J. Bennink At the crossroads of cell biology and immunology: DRiPs and other sources of peptide ligands for MHC class I molecules J. Cell Sci., January 3, 2001; 114(5): 845 - 851. [Abstract] [PDF]
|
|
|
|
|
|
N. A. Marshall, J. G. Howe, R. Formica, D. Krause, J. E. Wagner, N. Berliner, J. Crouch, I. Pilip, D. Cooper, B. R. Blazar, et al. Rapid reconstitution of Epstein-Barr virus-specific T lymphocytes following allogeneic stem cell transplantation Blood, October 15, 2000; 96(8): 2814 - 2821. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Hudrisier, J. Riond, H. Mazarguil, M. B. A. Oldstone, and J. E. Gairin Genetically Encoded and Post-translationally Modified Forms of a Major Histocompatibility Complex Class I-restricted Antigen Bearing a Glycosylation Motif Are Independently Processed and Co-presented to Cytotoxic T Lymphocytes J. Biol. Chem., December 17, 1999; 274(51): 36274 - 36280. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Y. Mo, P. Cascio, K. Lemerise, A. L. Goldberg, and K. Rock Distinct Proteolytic Processes Generate the C and N Termini of MHC Class I-Binding Peptides J. Immunol., December 1, 1999; 163(11): 5851 - 5859. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. Gileadi, H. T. Moins-Teisserenc, I. Correa, B. L. Booth Jr., P. R. Dunbar, A. K. Sewell, J. Trowsdale, R. E. Phillips, and V. Cerundolo Generation of an Immunodominant CTL Epitope Is Affected by Proteasome Subunit Composition and Stability of the Antigenic Protein J. Immunol., December 1, 1999; 163(11): 6045 - 6052. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. P. Papadopoulos, C. S. Hesdorffer, N. Suciu-Foca, H. Hibshoosh, and P. E. Harris Wild-Type p53 Epitope Naturally Processed and Presented by an HLA-B Haplotype on Human Breast Carcinoma Cells Clin. Cancer Res., August 1, 1999; 5(8): 2089 - 2093. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. C. Anton, U. Schubert, I. Bacik, M. F. Princiotta, P. A. Wearsch, J. Gibbs, P. M. Day, C. Realini, M. C. Rechsteiner, J. R. Bennink, et al. Intracellular Localization of Proteasomal Degradation of a Viral Antigen J. Cell Biol., July 12, 1999; 146(1): 113 - 124. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. K. Sewell, D. A. Price, H. Teisserenc, B. L. Booth Jr., U. Gileadi, F. M. Flavin, J. Trowsdale, R. E. Phillips, and V. Cerundolo IFN-{gamma} Exposes a Cryptic Cytotoxic T Lymphocyte Epitope in HIV-1 Reverse Transcriptase J. Immunol., June 15, 1999; 162(12): 7075 - 7079. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Valmori, U. Gileadi, C. Servis, P. R. Dunbar, J.-C. Cerottini, P. Romero, V. Cerundolo, and F. Levy Modulation of Proteasomal Activity Required for the Generation of a Cytotoxic T Lymphocyte-defined Peptide Derived from the Tumor Antigen MAGE-3 J. Exp. Med., March 15, 1999; 189(6): 895 - 906. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Andre, M. Groettrup, P. Klenerman, R. de Giuli, B. L. Booth Jr., V. Cerundolo, M. Bonneville, F. Jotereau, R. M. Zinkernagel, and V. Lotteau An inhibitor of HIV-1 protease modulates proteasome activity, antigen presentation, and T cell responses PNAS, October 27, 1998; 95(22): 13120 - 13124. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. P. N. Emmerich, A. K. Nussbaum, S. Stevanovic, M. Priemer, R. E. M. Toes, H.-G. Rammensee, and H. Schild The Human 26 S and 20 S Proteasomes Generate Overlapping but Different Sets of Peptide Fragments from a Model Protein Substrate J. Biol. Chem., July 7, 2000; 275(28): 21140 - 21148. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Komlosh, F. Momburg, T. Weinschenk, N. Emmerich, H. Schild, E. Nadav, I. Shaked, and Y. Reiss A Role for a Novel Luminal Endoplasmic Reticulum Aminopeptidase in Final Trimming of 26 S Proteasome-generated Major Histocompatability Complex Class I Antigenic Peptides J. Biol. Chem., August 3, 2001; 276(32): 30050 - 30056. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
