| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Institute of Clinical and Molecular Virology and
Department of Dermatology, University Hospital of Erlangen, Germany;
Institute of Medical Microbiology, University of Regensburg, Germany; and
Institute of Virology and Immunology, University of Wuerzburg, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
ELISPOT after immunization of naive C57BL/6 mice with UbRGagSL/EL4 cells. These results further underline the role of the defective ribosomal product pathway in adaptive immunity. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
The nature of the protein substrates that enter the MHC-I processing pathway has been an issue of intense investigation (5, 6, 7, 8, 9). It has now become widely accepted that newly synthesized and rapidly degraded polypeptides represent a significant source of MHC-I-restricted epitopes. These so-called defective ribosomal products (DRiPs) have been shown to constitute a large fraction of newly synthesized proteins (9). DRiPs never attain their functional conformation due to errors in transcription, translation, or immediate posttranslational processes necessary for proper protein folding, subcellular targeting, or assembly into functional complexes (reviewed in Ref. 10). To prevent interference with cellular function, DRiPs are degraded irrespective of the half-life of their functional counterpart in a rapid process that occurs either immediately after translation or even cotranslationally by the ubiquitin proteasome system (UPS). The discovery of the DRiP pathway provides a reasonable explanation for the paradox that MHC-I determinants of stable viral structural proteins are presented on the cell surface for detection by CTLs instantaneously after infection of target cells (11). In addition, this concept gained support through the demonstration that protein de novo synthesis is essential for the efficient formation of MHC-I ligands as it was demonstrated by correlation of protein synthesis with TAP activity and export of MHC-I molecules from the ER (6, 7, 9). Furthermore, it has also been shown that the de novo synthesis of a stable viral Ag is essential for recognition of Ag-derived epitopes by specific CTLs (8).
Degradation of proteins via the UPS represents a tightly controlled process that is essential for the maintenance of cellular integrity and functionality (12). Specific amino acid sequence motives, so-called degron signals (13), have been shown to regulate degradation by the 26S proteasome. Some proteins with a metabolic half-life of 2 h or less contain regions rich in proline, glutamine, serine, and threonine, which have therefore been named PEST sequences (14). Another way to attract proteins to the UPS includes the N-end rule which links the stability of a protein to the identity of its N-terminal amino acid (15, 16). Certain ubiquitin (Ub) ligases selectively target the so-called N-degron, consisting of a destabilizing N-terminal amino acid and internal lysine residues. Proteins exposing destabilizing N-terminal residues, like arginine, can be engineered by N-terminal in-frame fusions with Ub that is cotranslationally removed by Ub hydrolases allowing the generation of different N termini of otherwise identical proteins (17). When the C-terminal diglycine motif of Ub is mutated, cleavage of the isopeptide bond by Ub hydrolases is abolished and lysine residues within the Ub-moiety itself serve as attachment sites for poly-Ub chains, targeting the fusion protein for proteasomal degradation. This system is referred to as the Ub fusion degradation (UFD) pathway (18).
The correlation between Ag stability and generation of MHC-I antigenic determinants has been extensively studied (5, 6, 19, 20, 21, 22, 23, 24, 25). First, it has been demonstrated in a recombinant vaccinia virus (rVV) system that expression of an instable variant of the influenza A virus (IAV) nucleoprotein (NP) increased the efficiency in presentation of one particular epitope of NP (19), a finding that has subsequently been extended by others (5, 6). Recombinant Ub-X-galactosidase fusion proteins were differentially degraded according to the N-end rule in an Ub- and proteasome-dependent manner when introduced exogenously into cells and the instable counterparts of these fusion proteins evoked enhanced T cell activation (20). A correlation between protein half-life and generation of class I ligands has also been reported for the Listeria Ag p60 (21, 22). In contrast, no such correlation has been found for generation of the SIINFEKL (SL) epitope derived from instable OVA and β-galactosidase (β-Gal) fusion proteins (25). Enhanced Ag presentation along with efficient stimulation of CTLs has been shown for a metabolically instable variant of the HIV-1 accessory protein Nef (24). However, instable variants of the HIV-1 Pr55 Gag polyprotein failed to elicit enhanced CTL responses, although increased numbers of Gag-derived epitopes were presented on the cell surface (23).
To investigate the correlation between efficiency of protein degradation and induction of a CTL response, we chose the HIV-1 Gag polyprotein as a model Ag for two reasons: First, it has already been demonstrated that Gag as a viral Ag can enter the DRiP pathway (9) and, second, Gag represents a highly attractive Ag for vaccine development. Using the OVA-derived peptide SL as a model epitope, the efficiency of entry of Gag variants carrying certain degron signals into the MHC-I pathway was analyzed. Although protein half-lives of all Gag variants were comparable, one particular variant, UbRGagSL, was exceedingly attracted to the DRiP pathway, resulting in better Ag presentation and activation of CD8+ T cells in vitro and in vivo.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
The generation of the expression plasmid psyngag has been reported elsewhere (26). For the constructs pUbMsyngag and pUbRsyngag, ubiquitin was amplified from human genomic DNA using oligonucleotides Ub-1f (5'-ATG CAG ATC TTC GTG AAG ACC C-3') and Ub-1r (5'-ACC CCC CCT CAA GCG CAG GAC C-3') and subcloned into pcR-Script. Using primers Ub_Af (5'-CGC TCA GAA TTC GCC GCC ACC ATG CAG ATC TTC GTG AAG ACC-3') and Ub_B1r (5'-GGC CCT GGC GCC CAT ACC CCC CCT CAA GCG CAG GAC-3') or Ub_B2r (5'-GGC CCT GGC GCC CCT ACC CCC CCT CAA GCG CAG GAC-3'), respectively, and primers Ub_C1f (5'-GGG GGT ATG GGC GCC AGG GCC AGC-3') or Ub_C2f (5'-GGG GGT AGG GGC GCC AGG GCC AGC-3') and Ub_Dr (5'-gct cgc aTC TAG ATC ATC AGT TGC AGT GGG CCT GC-3'); overlapping fragments of Ub and syngag were generated, fused by PCR and cloned into pcDNA3.1 using EcoRI and XbaI sites. A sequence coding for the SL epitope, corresponding to aa 257–264 of OVA, was introduced into syngag using a unique BstEII site at position 1107 and oligonucleotides BstEII-SL-fw (5'-GTG ACC TCG ATC ATC AAC TTC GAA AAG CTA-3') and BstEII-SL-rc (5'-GTC ACT AGC TTT TCG AAG TTG ATG ATC GAG-3'). The SL epitope was also expressed from a minigene construct generated by inserting the SL-coding sequence along with a start codon, a Kozak consensus sequence, and a double stop codon using oligonucleotides Hind-SL-Xho-fw (5'-AAG CTT GCC GCC ACC ATG AGT ATA ATC AAC TTT GAA AAA CTG TGA TAG C-3') and Hind-SL-Xho-rc (5'-CTC GAG CTA TCA CAG TTT TTC AAA GTT GAT TAT ACT CAT GGT GGC GGC A-3') into the expression vector pcDNA3.1zeo.
Cell culture, transfection procedure, and generation of cell lines
EL4 is a thymoma cell line derived from the C57BL/6 mouse (H2-Kb). The EL4-derived cell line E.G7 synthesizes and secretes OVA (27). Cells were maintained in RPMI 1640 medium supplemented with 10% (v/v) heat-inactivated FCS, 2 mM L-glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin, 0.01% sodium pyruvate, and 0.1% nonessential amino acids (all from Invitrogen Life Technologies). To generate stable gag-expressing cell lines, EL4 cells were transfected by electroporation using a GenePulserXcell facility (Bio-Rad,) and settings as follows: exponential decay, 140 V, 950 µF in a 0.2-cm cuvette. After at least 3 wk of maintenance in complete RPMI 1640 medium containing 1 mg/ml G418 (Invitrogen Life Technologies), single-cell cloning by limiting dilution was performed. After at least 3 wk, cell clones were analyzed by an indirect ELISA. Fifty microliters of cell suspension was transferred to a MaxiSorp plate (Nunc) and mixed with 50 µl of 100 mM Na2CO3 (pH 9.0) and 0.1% (v/v) Nonidet P-40. After overnight coating at 4°C, plates were washed three times with 0.1% (v/v) Tween 20 in PBS (PBST) and blocked with 5% (w/v) BSA in PBST for 2 h at room temperature. Gag was detected using a pooled serum of 20 HIV-1-positive patients (National Institutes of Health AIDS Research Reference Reagent Program, catalog no. 3975) diluted 1/5000 in FACS buffer (5% (v/v) FCS and 0.02% (v/v) NaN3 in PBS), HRP-conjugated anti-human IgG Abs (GE Healthcare) and tetramethylbenzidine substrate (1-Step Turbo TMB ELISA; Pierce). Positive clones were further analyzed by p24 Ag ELISA, Western blotting, and FACS double staining (see below). Generally, two rounds of single-cell cloning were sufficient to achieve stable clones.
p24 Ag ELISA
To quantify Gag expression, 1.0 x 106 cells were lysed in 100 µl of CHAPS/deoxycholate (Doc) buffer (100 mM NaCl, 50 mM Tris-HCl (pH 8.0), 0.5% (w/v) CHAPS, 0.3% (w/v) sodium deoxycholate, 1 mM PMSF, 5 mM N-ethylmaleimide, 20 µM carbobenzoxyl-leucine-leucine-leucinal (zLLL; Sigma-Aldrich) and complete protease inhibitor mixture (Boehringer Mannheim)), and whole protein content was measured using the bicinchoninic acid assay (Pierce). In parallel, the p24 content was determined using a modified sandwich ELISA consisting of a coating Ab, an alkaline phosphatase-conjugated detection Ab (Aalto Bio Reagents) and CDP star substrate (Tropix). Deviating from the standard protocol, to avoid unspecific binding, the recombinant p24 standard was diluted in CHAPS/Doc extract from parental EL4 cells.
Immunofluorescence microscopy
For immunofluorescence analysis, 10,000 cells were centrifuged onto a slide (Superfrost; Roth) and fixed with methanol. After two washing steps in PBS, the fixed cells were stained using a FITC-conjugated anti-capsid (CA) Ab (KC57; Beckman Coulter) and 1 µg/ml 4',6-diamidino-2-phenylindole. Intracellular p24 was visualized using a fluorescence microscope (Axiovert 135; Zeiss) equipped with an Apochromat x63/1.40 oil objective.
Western blotting
Stably transfected EL4 cells were lysed in CHAPS/Doc buffer as described above and 10 µg of whole protein was separated by SDS-PAGE on a gradient gel (6–15%). Proteins were transferred onto a polyvinylidene difluoride membrane (GE Healthcare) and Gag was detected using anti-p6 (28) and anti-p24 sera (Seramun) followed by standard ECL detection.
Metabolic labeling
For detection of Gag-DRiPs, short-term pulse-chase experiments were performed, whereas long-term pulse-chase experiments were used to analyze protein stability. Briefly, for short-term pulse-chase experiments, transgenic EL4 cells were washed in PBS and treated with 25 µM each of zLLL and lactacystin (LC; Sigma-Aldrich) or DMSO as solvent control during the last 10 min of a 30-min starvation period in methionine (Met)-free, serum-free RPMI 1640 (Invitrogen Life Technologies). Cells were radiolabeled for 15 min with 3 mCi/ml [35S]Met (Amersham Life Sciences) and chased for up to 120 min while shaking at 37°C in DMEM supplemented with 10% FCS and 10 mM Met in the presence or absence of proteasome inhibitors. Cells were harvested and lysed in 200 µl of CHAPS/Doc buffer for 5 min on ice and separated from the insoluble fraction by centrifugation at 20,000 x g for 10 min. Gag was recovered by immunoprecipitation using a mixture of polyclonal rabbit anti-p6 and anti-p24 Abs prebound to protein G-Sepharose (GE Healthcare). Samples were separated by SDS-PAGE on a 10% (w/v) acryl amide ProSieve gel (Cambrex Bioscience) backed with Gel Bond film (FMC Bioproducts). Following fixation for 1 h in 40% methanol and 10% acetic acid, gels were rinsed with water, soaked in 1 M sodium salicylic acid solution for 5 h, and dried. Radioactivity in gels was analyzed using a phosphor imaging instrumentation (Fuji film BAS-2000;) or fluorography using BioMax MR films (Kodak) and quantified by AIDA imaging software (Raytest). For long-term pulse-chase experiments, after a 30-min starvation period, cells were radiolabeled with 3 mCi/ml [35S]Met for 30 min and then chased for up to 48 h. Analyses of radioactivity in cell lysates were performed as described above.
Flow cytometry
For detection of intracellular p24 and H2-Kb-bound SL epitope, double staining using KC57-FITC and mAb 25D1.16 (29) was performed. The latter was purified from hybridoma cell supernatant using fast protein liquid chromatography and a 5-ml G protein column (both GE Healthcare). One milligram of the purified Ab was labeled with Alexa Fluor 647 (Molecular Probes). After surface staining with 25D1.16, cells were fixed and intracellular staining using KC57-FITC was performed in FACS buffer containing 1% saponin. Cells were fixed in 1% paraformaldehyde and flow cytometry was performed on a FACSCalibur using CellQuest software (BD Biosciences).
Acid wash
SL-H2-Kbcomplex formation at the cell surface was followed by flow cytometry after an acid wash procedure as follows: cells were incubated for 2 min at pH 3 and 4°C in buffer containing 131 mM sodium citrate and 66 mM NaH2PO4. After neutralization in PBS, cells were incubated in complete RPMI 1640 medium in the presence or absence of 20 µM zLLL or protein synthesis inhibitors (6.3 µM emetine, 8.3 µM cycloheximide, and 1.7 µM puromycin; all purchased from Sigma-Aldrich) for desired periods at 37°C. At least 500,000 cells were incubated with supernatants of hybridomas 25D1.16, B8-24-3 (30), or medium alone as a control for 1 h on ice, followed by two washing steps in PBS and staining with Cy2-conjugated anti-mouse IgG (Rockland) diluted 1/200 in FACS buffer for 30 min on ice. Cells were fixed and analyzed by flow cytometry as described above.
T cell activation assay
The SL-H2-Kb-specific, murine CD8+ T cell hybridoma B3Z express the lacZ reporter gene under the control of the NFAT enhancer (31, 32). B3Z cells were provided by Prof. Nilabh Shastri (University of California, Berkeley, CA). In 96-well plates, 50,000 B3Z cells were cocultured with parental or transgenic EL4 cells in an E:T ratio ranging from 1:10 to 1:0.013 in a total volume of 200 µl/well at 37°C for 16 h. Cells were washed once with PBS and lysed by addition of 100 µl 0.15 mM chlorophenyl red β-galactopyranoside/0,5% (v/v) Nonidet P-40 in PBS. After incubation for 4–6 h at 37°C, the absorbance was determined at 595 nm.
Adoptive transfer of OT-1 T cells
OT-1 mice (33), which express a transgenic TCR (V
2/Vβ5) specific for SL, were received with permission of Dr. F. Carbone (University of Melbourne, Australia). Transgene status was confirmed by flow cytometry using a mAb specific for Vβ5 (BD Biosciences). Congenic C57BL/6 mice (Charles River Laboratories) were used as recipients. Spleen and lymph nodes (LN; cervical, axillary, brachial, inguinal, and mesenteric) from OT-1 mice were harvested, homogenized, and washed in PBS. After lysis of erythrocytes using 0.8% (w/v) NH4Cl for 5 min at 37°C, cells were passed over a 70-µm cell strainer (BD Biosciences). Cells were labeled with CFSE (Vybrant CFDA SE Kit; Molecular Probes) and transferred i.v. into the tail vein of naive C57BL/6 recipient mice (10 x 106 cells/mouse). Parental or transgenic EL4 cells (0.2 x 106 cells/mouse) were transferred 5 min later into the contralateral tail vein. Animals were sacrificed 2 or 3 days after adoptive transfer, spleen cells were harvested and analyzed by flow cytometry. To analyze the OT-1 cell proliferation, OT-1 cells were identified within the live lymphocyte population by staining for the transgenic TCR -chain V2 and CFSE levels were quantified.
IFN- ELISPOT
Naive C57BL/6 mice were injected into the tail vein with parental EL4 or UbGagSL/EL4 cells. As a positive control, 50 µg of synthetic SL peptide dissolved in 50 µl of IFA were s.c. injected near the tail root. After 9 days, the mice were sacrificed, the spleen was removed, and an ELISPOT assay was conducted to monitor the SL-specific T cell response. Briefly, splenocytes were collected under sterile conditions and erythrocytes were lysed using 0.8% (w/v) NH4Cl. Cells were plated in 96-well ELISPOT plates precoated with rat anti-mouse IFN- Ab (Mabtech). Following washing, splenocytes (10 x 106 cells/ml) were added and cells were pulsed with SL peptide (100 ng/ml) or PHA and schistosome egg Ag mitogen (5 µg/ml, each) in 100 µl of RPMI 1640 medium supplemented with 10% (v/v) FCS. To control for unspecific reaction, cells were treated in parallel without peptide. All experiments were performed in triplicates. After incubation at 37°C in a 5% CO2 atmosphere for 20 h, cytokine production was detected by using biotinylated Ab against IFN- (5 µg/ml; Mabtech) and alkaline phosphatase-streptavidin (0.2 U/ml). The IFN- spots were developed by addition of 100 µl of 5-bromo-4-chloro-3-indolyl phosphate/NBT solution. Spots in dried plates were counted using computer-assisted image analysis with a Zeiss Axioplan 2 and VisionKS ELISPOT version 4.9.15. For analysis, the number of spots without peptide (unspecific reaction) were subtracted from the number of spots with peptide (specific reaction). Statistical analysis was conducted using a t test, assuming unequal variance. To control for variations in the frequencies of CD8+ cells in the spleens of individual mice, FACS analyses using anti-CD8a-FITC (Beckman Coulter) were performed.
Statistical analysis
All statistical analyses were performed using the InStat 2.01 software and Student’s t test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Assuming the importance of DRiPs as the major source of antigenic peptides, we set out to enhance the generation of Gag-derived epitopes for MHC-I presentation by targeting the Gag polyprotein Pr55 originating from the isolate HIV-1HX10 for rapid degradation by the 26S proteasome. In the background of HIV-1, the coexpression of the regulatory HIV-1 Rev protein is strictly required for efficient nuclear export of Gag-specific mRNA (34). For analysis of Gag independently of other HIV-1 proteins, a codon-optimized, synthetic gag gene was expressed under the control of the CMV promoter, yielding high and Rev-independent expression levels of Gag in various cell lines (26). Unfortunately, there have been no Abs available which recognize MHC-I-bound Gag-derived epitopes. As an indicator for Ag processing, a sequence coding for the SL epitope of OVA was introduced into the p2 spacer region of the gag gene (Fig. 1A). In previous studies, it was shown that introduction of degron signals caused up to 3-fold reduction of the half-life of HIV-1 Gag (23). In an attempt to extrapolate these results to our Gag variant, a variety of degron signals, including a C-terminal PEST sequence, an N-end rule motif, as well as an UFD signal were inserted at various positions into Gag. Most astonishingly, among all of the degron signals tested, none affected the metabolic half-life of Gag significantly as measured by Western blotting and metabolic pulse-chase kinetics (data not shown).
|
To determine and compare the intracellular steady-state level of Gag, a p24 ELISA was performed on cell extracts standardized for protein content. The level of Gag in the UbMGagSL/EL4 clone (4.14 ± 2.36 ng of p24/Mio cells) was almost twice the amount of Gag determined in the UbRGagSL/EL4 clone (1.59 ± 1.14 ng of p24/Mio cells; Fig. 1A). This slight difference in Gag expression was routinely observed when transgenic EL4 cell lines expressing UbMGagSL and UbRGagSL were compared. Thus, it was conceivable to assume that the lower steady-state level of Gag in the UbRGagSL/EL4 clone represents an inherent characteristic of the N-end-ruled Gag substrate. Similar results were obtained by Western blot analysis using both, anti-p24 and anti-p6 Abs to visualize Gag proteins (Fig. 1B). Gag and GagSL proteins migrated at the appropriate molecular mass of 
55 kDa, confirming the complete cleavage of the Ub moiety. A slight difference in the molecular mass of Gag proteins reflects the presence or absence of the SL epitope. Consistent with the ELISA data, less Gag was observed for the R-Gag proteins irrespective of the presence of the SL epitope. In addition to the 55-kDa Gag signal, a second band migrating at 40 kDa was observed that presumably originates from an internal translation event as was described for another Gag expression system (35).
To control whether functional Gag proteins were expressed and not targeted to the centrosome, as it was observed for similar Ub fusion variants of the HIV-1 accessory protein Nef (36), the intracellular distribution of Gag was analyzed by immunofluorescence. In both, the UbMGagSL/EL4 and the UbRGagSL/EL4 clones, Gag proteins were equally distributed throughout the cytoplasm (Fig. 1C). Transport of Gag to the cell membrane and budding of virus-like particles further support the notion that, after cleavage of the Ub moiety, a functional Gag protein was expressed similar to its viral counterpart (data not shown).
Half-life and DRiP rate of UbRGagSL and UbMGagSL proteins
Protein turnover can be characterized by two different parameters: first, the overall metabolic half-life, which describes the degradation rate of a functional protein in its standard life cycle following folding and assembly into operating complexes, and, second, the DRiP rate, which describes the fraction of de novo synthesized proteins that do not follow the quality criteria and are degraded instantaneously in concert with or shortly after translation (6).
By performing distinctly designed short- and long-term pulse-chase experiments in the presence or absence of proteasome inhibitors, we addressed the question in which of these characteristics the UbRGagSL and UbMGagSL proteins differ. The overall protein half-life was determined by pulse radiolabeling of EL4 cells with [35S]Met for 20 min followed by a 48-h long chase. [35S]Met Gag proteins that were recovered by precipitation using p24- and p6-specific Abs, subjected to SDS-PAGE, and visualized by fluorography displayed in both cases a half-life of 6 h (Fig. 2A). In parallel, total cell lysates were resolved by SDS-PAGE and the major band corresponding to β-actin served as an internal control (Fig. 2A, lower panel). Nevertheless, the decay of Gag expressed in the UbRGagSL/EL4 clone was somewhat faster during the first 24 h of chase when compared with the UbMGagSL/EL4 clone. This was further supported by phosphor imager analysis of three independently conducted pulse-chase experiments (Fig. 2B), demonstrating that the overall half-life of Gag is not governed by the N-end rule, although in the case of UbRGagSL, the decay was to some extent enhanced in the first hours of chase, indicating that the slightly shorter half-life cycle of UbRGagSL might be affected by increased turnover occurring in concert with de novo synthesis.
|
60–250 kDa were measured using a phosphor imager (Fig. 2D). Similar to the long-term pulse-chase (Fig. 2A), the decay of Gag was slightly faster in UbRGagSL/EL4 cells. Although data obtained in both cell lines with active proteasomes were comparable, the situation after proteasome shutdown was quite different: although addition of zLLL/LC rescued Gag proteins in both cultures from degradation, the amount of Pr55 recovered from UbRGagSL/EL4 cells increased by >50% immediately after the pulse, reaching maximum values of up to 80% increase within 15 min (Fig. 2D, upper right panel) compared with only a 30% increase of Pr55 recovered from UbMGagSL/EL4 cell extracts (Fig. 2D, upper left panel). The same effect was observed for the Gag fragment migrating at 40 kDa: the rate of recovery of which was again more pronounced in UbRGagSL/EL4 cells when compared with UbMGagSL/EL4 cells. In addition, an increase in the smear of proteins migrating in the molecular mass range of 60–250 kDa was detected in cells treated with zLLL/LC. This smear recovered by anti-Gag Abs was not recovered under the same conditions from nontransfected EL4 cells (data not shown), indicating that the smear represents polyubiquitinated Gag-DRiPs and not cellular proteins that bind nonspecifically to Gag-anti-Gag immune complexes. Similar to previous calculations in HIV-1-expressing cells (9), the radioactivity in this molecular mass range was taken into account for the estimation of the DRiP rates of both UbGagSL fusion proteins. Although treatment of UbMGagSL/EL4 cells with zLLL/LC resulted in modestly enhanced recovery of proteins in the molecular mass range of 60–250 kDa, the magnitude of increase was clearly higher in UbRGagSL/EL4 cells (Fig. 2D, lower panels), confirming results observed for Pr55. Similar to the kinetics of DRiP formation reported previously (9), the accumulation of labeled Gag proteins during the pulse and the first 15 min of chase is of transient nature (Fig. 2D). After reaching a certain plateau, processes like deubiquitination, proteolysis, or aggregation, as well as association with membranes (39), altogether continuously remove DRiPs from the pool of total Gag proteins accessible for immunoprecipitation. Thus, the results of the pulse-chase analyses demonstrate that the DRiP rate of the Gag proteins clearly followed the N-end rule, although the N-terminal fusion of UbR to Gag does not result in a significant reduction of the overall metabolic half-life. Correlation of DRiP rate with the efficiency of MHC-I presentation of Gag proteins
Although the stability of both Gag proteins was not governed by the N-end rule, an observation which is consistent with previous reports by others using different Gag expression systems (23), we observed an elevated DRiP rate of the UbRGagSL variant compared with UbMGagSL. To analyze whether UbRGagSL enters the MHC-I processing pathway more efficiently, the presentation of SL in complex with MHC-I molecules by flow cytometry was assessed using the fluorescently labeled mAb 25D1.16, which is specific for H2-Kb-bound SL (29). In parallel, the expression of Gag was monitored by intracellular staining with a CA-specific Ab. Parental EL4 cells and a cell clone expressing relatively high levels of the SL epitope from a minigene construct, called SL/EL4, served as negative and positive controls, respectively. Similar to previous studies using rVV expression systems (5, 40, 41), a CMV promoter construct that directs high-level expression of the sequence MSIINFEKL was used to generate the stably transfected SL/EL4 clone. FACS analyses revealed that the cell surface presentation of the SL epitope generated from UbRGagSL clearly surpasses the level of presentation observed for the UbMGagSL counterpart (Fig. 3A). However, consistent with the ELISA and Western blot data, the level of Gag in UbRGagSL-expressing cells is about half of that observed in the UbMGagSL-transgenic cells as shown by double staining with mAb 25D1.16 and a CA-specific Ab (Fig. 3B). To further support the notion that the increased presentation of H2-Kb-SL complexes on the surface of UbRGagSL/EL4 cells is not simply dependent on variations in expression levels of Gag, the presentation of H2-Kb-SL complexes was correlated with the intracellular amount of Gag in five independently performed double-staining experiments (Fig. 3C). The ratio of the mean fluorescence intensity (MFI) at 647 nm (staining with 25D1.16-Alexa Fluor 647) and the MFI at 488 nm (staining with anti-CA-FITC) was calculated demonstrating 2-fold increase in H2-Kb-SL presentation (p = 0.019; Student’s t test).
|
2-fold higher in UbRGagSL-processing cells (Fig. 4B, time point –10). Second, after acid wash that causes a drop to almost identical baseline levels (Fig. 4B, time point 0), the kinetics of recovery of SL-H2-Kb complexes on the cell surface was clearly faster in the UbRGagSL/EL4 clone. According to previous studies that demonstrated the generation of epitopes from de novo-synthesized Ag of different origin in a proteasome-dependent manner (1, 5, 7, 8, 9, 43, 44, 45), control experiments using protein synthesis inhibitors or proteasome inhibitors (Fig. 4, C and D) revealed that both protein biosynthesis and the activity of the proteasome are necessary in both cell clones for the efficient presentation of H2-Kb molecules on the cell surface. This was shown for total H2-Kb (Fig. 4C) and SL-loaded H2-Kb molecules (Fig. 4D).
|
So far, we have demonstrated more efficient SL presentation in UbRGagSL-processing cells when compared with UbMGagSL-processing cells, although both clones differ only in the DRiP rate but not in the overall protein stability of Gag. To analyze whether the augmented SL presentation led to an enhanced T cell activation, we used the SL-H2-Kb-specific T cell line B3Z, which had been generated by fusion of the NFAT-lacZ T cell clone with a SL-specific V2Vβ5 T cell clone (46). After overnight cocultivation of B3Z cells with corresponding EL4 target cells, the specific T cell activation was detected by a colorimetric assay detecting TCR signaling. First, we established that the experimental conditions were adequate to detect specifically differences in the presentation of H2-Kb-SL complexes. Incubation of B3Z cells with increasing concentrations of the synthetic peptide SIINFEKL or a control peptide SIIKFEKL demonstrated that exogenously added epitopes are bound to cell surface H2-Kb molecules and presented to B3Z cells (Fig. 5A). A linear correlation between peptide concentration and activation of B3Z cells was observed in the range of 0.3–30 ng/ml, which corresponds to 0.3–30 nM of available peptides. Complete absence of T cell activation in the case of the control peptide confirms the high specificity of the B3Z line. In a further experiment, parental EL4 cells were incubated for 1 h with increasing concentrations of exogenously added peptides. After intensive washing, the cells were mixed with B3Z cells in an E:T ratio of 1:1 (Fig. 5B), revealing that even a short peptide pulse can induce T cell activation that started at 10 ng/ml added SL. Thus, the B3Z cell line represents a very sensitive and specific indicator for SL presentation which is able to detect subtle differences in epitope density at the level of the signaling pathway. When Gag-expressing EL4 clones were cocultivated with B3Z cells at different E:T ratios ranging from 1:10 down to 1:0.013, a significant stronger T cell activation was observed for the UbRGagSL/EL4 clone compared with the UbMGagSL/EL4 counterpart, as shown in a representative experiment in Fig. 5C.
|
To test whether the enhanced MHC-I presentation of SL epitopes generated from the UbRGagSL variant also augments T cell stimulation in vivo, UbMGagSL/EL4 and UbRGagSL/EL4 cells were adoptively transferred together with OT-1 splenocytes into naive C57BL/6 mice. The OT-1 mouse is transgenic for the SL-specific TCR (V2/Vβ5), with approximately one-third of all splenocytes and LN cells representing naive, SL-specific CD8+ T cells (47). As positive controls both, the OVA-expressing EL4-derived cell line E.G7 (27) as well as the minigene-expressing SL/EL4 cell clone (Fig. 3) were used, while the parental EL4 cells served as negative control. To quantify proliferation of OT-1 T cells, splenocytes and LN cells derived from OT-1 mice were labeled with CFSE before injection. This method is meant to track the number of cell divisions in vivo (48). To prevent target and effector cell encounter before injection and, in addition, to enhance the chance of interaction of both cell types in the recipient mouse, OT-1 cells were first injected into one tail vein and, after a delay of 5 min, the SL-presenting EL4 cells were injected into the contralateral vein. Two days after injection, the spleens of the recipient C57BL/6 mice were removed and the splenocytes were analyzed by flow cytometry. Among splenocytes the OT-1 cells were identified as V2- and CFSE-positive cells (Fig. 6A, gate R3). In contrast to V2-negative cells (Fig. 6A, gate R4), a certain fraction of the V2-positive cell population showed clear evidence for proliferation, which was evident from a decaying CFSE fluorescence intensity. Analyzing V2- and CFSE-positive cells derived from the UbRGagSL/EL4 recipient revealed a relatively higher number of OT-1 cells that were primed and had started to proliferate compared with OT-1 cells derived from the UbMGagSL/EL4 recipient mice (Fig. 6A, lower panels). This notion was further supported by the down-regulation of CD62L expression on the surface of V2- and CFSE-positive cells, which is indicative of an activated T cell phenotype (data not shown). The statistical analysis of 10 independent experiments revealed that SL-expressing cells stimulate the V2- and CFSE-positive population of the transferred spleen and LN cells to proliferate at different levels (Fig. 6B). To compare differences in OT-1 T cell stimulation, we calculated the proliferation index which takes into account the number of proliferating cells relative to the number of cell divisions. The comparison of the proliferation index induced by UbRGagSL/EL4 cells (0.85 ± 0,31) and UbMGagSL/EL4 cells (0.53 ± 0.25) revealed that UbRGagSL induces a significantly stronger proliferation of adoptively transferred OT-1 T cells. Thus, the efficiency of SL presentation in UbRGagSL/EL4 cells correlates with a better T cell activation in vivo as quantified on the level of T cell proliferation.
|
ELISPOT. Parental EL4 cells served as negative control, whereas for positive control the SL peptide was injected s.c. Results shown in Fig. 7 reveal that higher frequencies of SL-specific T cells were induced by immunization with UbRGagSL/EL4 cells when compared to immunization with UbMGagSL/EL4 cells, indicating that, in naive mice, UbRGagSL/EL4 cells are more potent in activating not only transgenic OT-1 T cells, but also naive SL-specific T cells.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
6 h, the formation of DRiPs derived from the R-Gag variant was about twice the DRiP rate of wild-type (wt) M-Gag. Using the OVA-derived SL as an indicator epitope, it was shown that the intracellular steady-state level of Gag and the efficiency of SL presentation were inversely proportional: although UbRGagSL/EL4 cells expressed less Gag, the number of H2-Kb-SL complexes on the cell surface and the kinetics of loading of empty H2-Kb molecules were elevated when compared with UbMGagSL/EL4 cells. In consistency with the enhanced Ag presentation, the UbRGagSL/EL4 cell line also induced a better CD8+ T cell activation in vitro and in vivo. Regardless of all of the efforts spent, there is a certain need to improve current strategies for protective or therapeutic vaccinations against HIV-1. Although high-frequency T cell responses can be observed in HIV-1-infected individuals, even in those with progressing disease (49), there is evidence that virus-specific CD8+ T cells display a key function in the immune control of virus spread (50). Even more, there has been cumulating evidence reported that indicate differences in the capability of virus-specific T cell clones to interfere with HIV-1 infection. Therefore, vaccine strategies aimed at the generation of potent virus-specific CTL response are now generally considered as a therapeutic option to interfere with disease progression subsequent to HIV infection (51), particularly as stimulation of sterilizing immunity by induction of neutralizing Abs has not been achieved so far (52). Most interestingly, it has been shown recently that an effective CTL response acting specifically against HIV-1 Gag, but not against other viral Ags, can be correlated with a significant reduction in viral load in HIV-1-infected patients (53). In addition, HIV-1 Gag can be considered as an interesting model Ag because it is exclusively present in the cytosolic compartment and thus fully accessible to the UPS and subsequent entry into the Ag processing pathway. Furthermore, principal biochemical procedures to measure the DRiP rate of Gag have been established (9). However, although a number of Gag-derived MHC-I epitopes are characterized, there is no specific Ab available to analyze the quantity of Gag-derived epitopes in complex with MHC-I molecules at the cell surface. To circumvent this problem we found that insertion of the standard model epitope SL into the polymorphic p2 spacer region of Gag (54) resulted in efficient processing and presentation of SL on the cell surface that did not interfere with virus budding and release of the respective Gag proteins (data not shown).
The validity of the N-end rule has been successfully demonstrated for a number of Ags derived from viruses and bacteria, like the IAV NP, the E. coli β-Gal, the Sindbis virus RNA polymerase, the Listeria p60 protein, or the HIV-1 accessory protein Nef (5, 19, 20, 21, 25, 55, 56). However, the N-end rule does not seem to apply ubiquitously for all proteins tested so far (23, 57). In agreement with others, we could not detect significant changes in protein half-life for the UbRGagSL fusion protein (23). Even addition of a short lysine-rich leader sequence to a similar UbRGag fusion protein resulted only in moderate destabilization of Gag, as shown by others (23). A possible explanation for the resistance of Gag to the N-end rule might be its inherent capacity to self-assemble after transport to virus-budding sites at the cell membrane. This might obscure the recognition of the N-end degron in the context of completely folded Gag during the formation of multi-Gag protein complexes of budding virus particles. In contrast, the destabilizing R-residue at the N terminus of the nascent protein should be better accessible to cognate Ub ligases during translation. As part of the N-end degron, accessible internal lysine residues are essential for Ub attachment and, thus, for targeting of the substrate for degradation by the 26S proteasome. Moreover, inconsistent reports about the validity of the N-end rule might be consequent to differences of proteasome subunit composition of the cell type used for expression of the model Ag driven by various expression systems. Although we can merely speculate about the underlying reason for the variability of the N-end rule, it seems reasonable that damaged proteins or polypeptides tagged for destruction by degron signals are removed as quickly as possible, preferentially during synthesis and thus affected by destabilizing N-end residues preferentially during the accessibility to the DRiP pathway.
Accumulation of polyubiquitinated Gag-DRiPs derived from UbRGagSL occurred after proteasome inhibition, as it was also observed for HIV-1-expressing cells (9). Nevertheless, Ub-independent degradation by the proteasome has been described (58) and might also be regulated by destabilizing N-terminal residues, for instance by affecting the folding state of the nascent peptide chain.
Distinct to these previous observations of inaccessibility of HIV-1 Gag to the N-end rule in terms of overall protein half-life (23), we now demonstrate that the N-end rule regulates the DRiP rate and thus the Ag-processing rate of Gag. Wong et al. (23) already reported that rVV-mediated expression of subgenomic, altered, and highly turned over fragments of Gag resulted in increased numbers of SL-H2-Kb complexes presented at the cell surface. However, when both primary and memory CTL responses were compared in vaccination studies in mice using Gag-encoding plasmid DNA or rVVs, the stimulation of T cell responses in vivo did not correlate with the processing rate of those Gag variants. Mechanisms that regulate the entry into the DRiP pathway have been intensively investigated by others, and it is becoming increasingly clear that the efficiency in translation and protein folding strongly affect the DRiP rate of a given protein inasmuch the efficiency of protein folding is also critical for entry of a given Ag into the MHC-I processing pathway (45, 59, 60). Variations between full-length wt Gag studied here and fragments of Gag studied by others (23) might contribute to those differences. Impaired protein folding might be directly related to enhanced DRiP rate of a given Ag, and induction of misfolding by introduction of mutations into structured regions of the protein (59, 60) or chaperone inhibition (45) should support entry into the MHC-I pathway.
Rapid cotranslational degradation of a certain proportion of total Gag should slightly reduce the steady-state level of R-Gag that otherwise exhibits the same turnover rate during the posttranslational "second" half-life as wt M-Gag. This might explain why less Gag was observed in R-Gag-expressing cells on the steady-state level when compared with the M-Gag-expressing cells. If aged proteins that enter the MHC-I pathway during their second half-life were the main source of antigenic peptides, this reduced availability in the second half-life should lead to less MHC-I presentation of Gag-derived epitopes. Nonetheless, we detected clearly more H2-Kb-SL complexes presented at the cell surface of the UbRGagSL/EL4 cell line.
Importantly, the increase in MHC-I Ag presentation was accompanied by enhanced T cell activation and clearly less UbRGagSL/EL4 cells compared with UbMGagSL/EL4 cells were necessary to achieve the same level of T cell activation, as demonstrated in vitro using the B3Z T cell hybridoma. Similarly, in the in vivo model of adoptive transfer, 2 x 106 of both target cell lines UbRGagSL/EL4 and UbMGagSL/EL4 were sufficient to induce maximal stimulation of 1 x 107 cotransferred SL-specific OT-1 T cells. However, when the number of target cells was limited to 2 x 105, a clear difference in OT-1 T cell proliferation was observed. This indicates that for efficient stimulation of CD8+ T cells in vivo, limited availability of target cells can be compensated by a higher presentation rate of specific MHC-I complexes per target cell. A correlation between the amount of TCR ligands at the surface of professional APCs and the induction of a CTL response has been shown in other models (61, 62, 63, 64).
The MHC-I pathway is considered to be a complex and multistep process. We analyzed the initial step, the synthesis and degradation of an Ag, and the final step, the amount of peptide-MHC-I complexes on the cell surface. Because the amount of total H2-Kb-molecules and the rate of reappearance after the acid wash procedure were similar in both EL4 cell lines, we assume that the H2-Kb expression levels as well as the intermediate steps in the MHC-I pathway do not markedly differ between the two cell lines, whereas the differences in the proportion of H2-Kb molecules loaded with SL epitope were consequent to disparities in the DRiP rate. Similar to previous studies, the generation of the SL epitope was dependent on proteasome activity and ongoing protein biosynthesis (1, 5, 7, 8, 9, 43, 44, 45), further supporting the notion that this phenomenon studied in our system is related to the DRiP pathway.
During the peptide loading of MHC-I molecules in the ER, a great number of peptides, derived from turnover of an estimated number of 2–3 x 109 cellular proteins, compete for the available number of binding grooves (65). Thus, an enhanced DRiP rate should result in an increased generation and steady-state level of antigenic peptides. DRiPs as a source for antigenic peptides may become especially crucial for efficient MHC-I Ag presentation in early stages of viral infection during the onset of synthesis of viral proteins, especially of stable, structural proteins like the retroviral Gag polyproteins (65, 66).
Several DNA vaccination studies, using rVV vectors or plasmids expressing an instable variant of HIV-1 Nef, UbRNef, and an HIV-1 envelope (Env) protein mistargeted for expression in the cytoplasm (55) as well as lymphocytic choriomeningitis virus NP targeted for degradation using an UFD signal (67) or N-end-ruled variants of β-Gal (68) challenged altogether the hypothesis that immune responses can be improved by enhancing proteasomal degradation of a given Ag used for vaccination. Consistently, in these reports, highly efficient induction of a protective CTL response was observed in the absence of a humoral immune response. Furthermore, a DNA vaccine encoding an UbR fusion to IAV NP that exhibited wt half-life induced lower Ab responses while CTL responses directed against one specific epitope were comparable to wt NP (57). The reduced humoral immune response was thought to be a consequence of reduced availability of the full-length Ag, presumably due to increased continuous turnover and a reduced steady-state level of NP. Rapidly degraded forms of Ag not only fail to elicit Ab responses, but might also represent poor substrates for cross-presentation, which is thought to be an important mechanism for priming of CD8+ T cells by dendritic cells (DCs) (69). Although priming with high levels of epitope seems to be more efficient to induce CTL responses, memory CTLs of lower avidity (61) were generated or the number of Ag-specific memory T cells progressively declined (64).
T cell response against viruses that do not preferentially infect APCs is now considered to be mediated by mechanisms linked to the potent pathway of cross-priming (for review, see Refs. 70 and 71). Virus-derived Ags are introduced into professional APCs by phagocytosis either in the form of bits and pieces of infected cells or even by particulate protein aggregates or Ags stabilized by certain chaperones (72). Intriguingly, there is evidence that proteasome shutdown in Ag-contributing cells increases the capability of the donor cell to deliver cross-presented proteins to APCs (73). Furthermore, ubiquitinylated Gag associates with membranes (39) and proteasome inhibitors cause the accumulation of Gag-DRiPs in insoluble protein aggregates (S. Hahn and S. Schreiber, unpublished observation). Thus, driving Gag into the DRiP pathway might also enhance its attraction for cross-presentation, particularly as the entry of incoming Gag into the proteasome after HIV infection has already been demonstrated (74). Finally, the transmission of Gag-derived Ag through gap junctions (75) might also represent an efficient pathway for T cell priming by APCs.
In conclusion, we propose that targeting of Ag to the DRiP pathway represents a suitable strategy for increasing the number of epitope/MHC complexes at the cell surface. Although CD8+ T cells generally exhibit an exceptionally high sensitivity for specific epitope/MHC-I complexes (76), enhancing Ag processing and thereby MHC-I Ag presentation might be advantageous in situations where low-avidity T cells encounter target cells with insufficient MHC-I presentation, e.g., virus-infected cells at an early stage of virus replication.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported by a grant from the Bavarian Research Network for Infectogenomics, BioMedTec Internationational Graduate School of Science, "Lead Structures of Cell Functions," funded by the State of Bavaria, Germany, by Grant IE-S08T06 from the German Human Genome Research Project, and by Grant SFB 643-A1 from the German Research Council. 
2 Current address: Institute of Clinical Microbiology, Immunology, and Hygiene, University Hospital of Erlangen, Erlangen, Germany.
3 Address correspondence and reprint requests to Dr. Ulrich Schubert, Institute of Clinical and Molecular Virology, University Hospital of Erlangen, Schlossgarten 4, 91054 Erlangen, Germany. E-mail address: ulrich.schubert{at}viro.med.uni-erlangen.de
4 Abbreviations used in this paper: MHC-I, MHC class I; ER, endoplasmic reticulum; DRiP, defective ribosomal product; UPS, ubiquitin proteasome system; Ub, ubiquitin; UFD, Ub fusion degradation; rVV, recombinant vaccinia virus; IAV, influenza virus A; NP, nucleoprotein; β-Gal, β-galactosidase; SL, SIINFEKL; Doc, deoxycholate; LN, lymph node; MFI, mean fluorescence intensity; LC, lactacystin; wt, wild type; DC, dendritic cell; zLLL, carbobenzoxyl-leucine-leucine-leucinal; CA, capsid.
Received for publication June 8, 2007. Accepted for publication October 28, 2007.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
