MHC Class I Antigen Processing of an Adenovirus CTL Epitope Is Linked to the Levels of Immunoproteasomes in Infected Cells

DNA constructs

The expression vectors pTET-Splice and pTET-tTAk, containing the tetracycline (TET)-regulated tTAk transcription activator, were described by Shockett et al. (30). cDNAs encoding LMP2 (H-2^b haplotype), LMP7, and MECL-1 were cloned into pTET-Splice using standard methods.

Cell lines and transfections

The MEC-18 cell line was derived from a primary culture of embryonal cells (MEC) of a C57BL/6 p53^−/− mouse and cultured in IMEM (Biochrom, Berlin, Germany) with 10% FCS, 2 mM l-glutamine, 20 μM 2-ME, penicillin, and streptomycin. Human WEHI-164 clone 13 cells were maintained as described (31). The CTL clones 100B6 (28) and 100 PKOB4 (31) are specific for the Ad5 CTL epitope E1B_192–200 in the context of H-2D^b MHC class I molecules.

Transfectant cell lines were established using calcium-phosphate precipitation. MEC-18 cells were transfected with pTET-tTAk and pLXSH, and positive clones were selected with hygromycin B (200 μg/ml) in medium containing 400 ng/ml TET. The presence of tTAk was confirmed by PCR analysis using the oligonucleotides 5′-CCTGGAGACGCCATCCACGC-3′ and 5′-CGTTTAAGGGCACCAATAACTGC-3′, which are specific for pTET-Splice. TET-inducibility of tTAk expression was verified by Northern analyses. MEC/tTAk clone 29 (MEC-29) was further transfected with pTETsplice-LMP2, -LMP7, and -MECL-1 and pLXSP. Transfectant cells were selected with puromycin (50 μg/ml) in medium with 200 μg/ml hygromycin and 400 ng/ml TET. Inducible expression of the proteasome subunits was examined by immunoblot analysis.

Western blot analysis

Western blot analyses were performed as described (28). Expression of the proteasome immunosubunits was detected with polyclonal rabbit antisera raised against recombinant mouse LMP2 and against the synthetic peptides SDVSDLLYKYGEAAL and TAGGAKLQRALSTPTEPVQ corresponding to residues 265–276 of mouse LMP7 and 218–236 of mouse MECL-1, respectively.

Metabolic labeling and immunoprecipitation

A total of 2 × 10⁶ MEC217 cells were seeded in flasks in Iscove’s medium with 400 ng/ml TET, washed after adherance, and placed in medium without or with TET (400 ng/ml or 10 ng/ml) for 10 h. Cells were labeled overnight with translabel [³⁵S]methionine (50 μCi/ml), washed twice, chased for 4 h in medium with 400 ng/ml TET, and lysed in lysis buffer (1% Nonidet P-40, 20 mM Tris.Cl (pH 7.5), 10 mM EDTA, 100 mM NaCl) with 1 mM PMSF, 2 μg/ml pepstatin, 2 μg/ml leupeptin, and 2 μg/ml aproptinin as described (3). Centrifuged detergent lysates were precleared with 25 μl protein A-Sepharose (50% slurry) and 25 μl rabbit preimmune serum o/n at 4°C. 20S proteasomes were immunoprecipitated with 25 μl protein A-Sepharose and 25 μl polyclonal 20S-proteasome-specific rabbit antiserum for 1 h at 4°C. Sepharose beads were washed with NET buffer (50 mM Tris.Cl (pH 8), 150 mM NaCl, 5 mM EDTA) with 0.5% Nonidet P-40, resuspended in nonequilibrium pH-gradient gel electrophoresis (NEPHGE) sample buffer, and subjected to NEPHGE two-dimensional gel electrophoresis. Gels were exposed for autoradiography. Radioactive signals were quantified using a Molecular Dynamics (Sunnyvale, CA) phosphor imager.

Proteasome isolations

20S proteasomes used in this study were freshly prepared as described (14) and kept on ice. The purity of the proteasome preparates, checked by Coomassie-stained SDS-PAGE, was >95%.

NEPHGE two-dimensional gel electrophoresis

Purified proteasomes were subjected to NEPHGE-PAGE as described (14). Gels were Coomassie-stained according to standard procedures.

Ad5 and infections

Ad5 dl7001 (33) lacking the E3 region was propagated in 911 cells as described (34). Viral stocks were aliquoted and stored at −80°C in 10% glycerol. Adhered MEC217 cells were infected at a multiplicity of infection of 100 in PBS with 0.5% BSA for 35 min at room temperature. This procedure results in infection of 100% of cells. Infected cells were cultured in Iscove’s medium at 37°C until use in a CTL assay.

Eu³⁺-release assay

MEC217 cells were cultured for 3 days in the absence or presence of TET and then infected with Ad5 dl7001 or left uninfected. After 24 h of infection, the cells were harvested and used as targets in a 16-h Eu³⁺-release assay with CTL clone 100B6 as described (35).

TNF-α production assay

MEC217 cells cultured for 2 days without or with 10 ng/ml or 400 ng/ml TET were infected with Ad5 dl7001 or were left uninfected. After 12 h of infection, the cells were harvested and used as targets in a TNF-α production assay. Target cells were placed in the wells of 96-well plates (1 × 10⁴ cells/well) in a volume of 50 μl. CTL clone 100 PKOB4 cells (5000 cells/well) were added in a volume of 50 μl, and the plates were incubated at 37°C for 2.5, 10, and 22 h. Sixty microliters of supernatant of independent triplicate wells were transferred into the wells of fresh 96-well plates and were assayed for the presence of TNF-α by measuring the cytotoxic effect on WEHI cells as described (31). Conversion of MTT substrate by WEHI cells incubated in CTL supernatant was determined by measuring the OD570s of the microcultures with a Dynatech (Chantilly, VA) MR5000 ELISA reader. The OD570s of wells with WEHI cells cultured with mouse recombinant TNF-α (35 pg/ml) were used as a reference for total WEHI cell death.

Peptides

Z-LLE-βNA was purchased from Bachem (Heidelberg, Germany). Synthetic AKVALRPDCKYKISKLVNIRNCCYISGNGAEVEIDTEDRV (Ad5 E1B E1B_176–215) and VNIRNCCYI (E1B_192–200) were kindly provided by Dr. P. Henklein (Institute of Biochemistry).

Peptide digestion assays and quantitation of CTL epitope generation

Digestion assays using Z-LLE-βNA as a substrate were performed as described (14). To determine proteasome-mediated cleavage of Ad5 E1B 40-mer, 15 μg of polypeptide and 1 μg of purified 20S proteasomes were incubated in 100 μl assay buffer (20 mM HEPES/KOH (pH 7.8), 2 mM MgAc₂, 5 mM DTT) at 37°C for the time periods specified in the figure legends, and then they were frozen. Samples were analyzed by reversed phase HPLC (RP-HPLC) (HPLC system HP1100 (Hewlett-Packard, Waldbron, Germany) system equipped with a μRPC C2/C18 SC 2.1/10 column (Pharmacia, Freiburg, Germany); eluent A, 0.05% trifluoroacetic acid ; eluent B in 30 min, 63–95% eluent B in 4 min; flow rate, 50 μl/min). Analysis was performed on line with an ion trap mass spectrometer (ThermoQuest, Egelsbach, Germany) equipped with an electronspray ion source. Each scan was acquired over the range m/z = 300-1300 in 3 s. The peptides were identified by their molecular masses calculated from the m/z peaks of the single or multiple charged ions and were confirmed by mass spectrometric sequencing analyses. To measure E1B_192–200 generation, the digests were diluted in PBS with 0.5% BSA and 80 μM 2-ME and were titrated on IFN-γ-induced uninfected MEC217 cells in 96-well plates (2 × 10⁴ cells/well). After 30 min at 37°C, cells were washed with PBS used as targets in a TNF-α production assay. A standard curve generated in the same experiment by diluting synthetic E1B_192–200 on MEC217 cells was used to quantify E1B_192–200 in the digests.

TET-regulated expression of proteasome immunosubunits

To examine the effect of LMP2, LMP7, and MECL-1 on proteasome-mediated Ag degradation and CTL epitope generation, transfectant mouse embryonal cell (MEC) lines expressing these subunits under a TET-regulatable inducible promoter were generated as described in Materials and Methods. A total of 54% of the established MEC lines showed inducible expression of the three introduced subunits in immunoblot analyses, as demonstrated for a representative clone (MEC217) in Fig. 1⇓. This clone was selected for complete processing of the tranfected subunits into the mature form, lacking the prosequence, which indicates efficient incorporation into 20S proteasomes. MEC217 cells cultured in the presence of TET, which inactivates tTAk, did not express the LMPs or MECL-1 at detectable levels, and thus lack any significant expression of the inducible proteasome subunits. In the absence of TET, both LMPs and MECL-1 were clearly detectable (Fig. 1⇓). Kinetic experiments established that the levels of induced LMP2 increased from the first up to the third day of TET removal and then stabilized. In contrast, TET-mediated repression of gene transcription resulted in a substantial drop of LMP2 levels within 2 days, probably due to cell division (not shown). The expression of the transfected subunits did not influence the levels of endogenous PA28α or β (not shown), implicating the absence of stabilizing effects of immunoproteasomes on PA28, which could serve as a mechanism to optimize MHC class I Ag processing. Thus, in MEC217 cells the levels of immunosubunits are tightly regulated by TET, and immunoproteasomes are expressed in the absence of elevated levels of PA28.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

TET-regulated expression of the proteasome immunosubunits in MEC217 tranfectant cells. Cells were cultured in the absence or presence of 400 ng/ml TET for 3 days and then harvested. Cellular lysates were separated by SDS-PAGE and probed with LMP2-, LMP7-, or MECL-1-specific polyclonal rabbit antiserum. The immunosubunits and their precursor forms (pLMP2, pLMP7, and pMECL-1) are indicated. The higher molecular mass bands in the LMP2 lanes are unspecific reactions. Molecular mass markers in kDa are shown on the left.

Dose-dependent exchange of constitutive for induced proteasome subunits in MEC217 cells

To further study the efficiencies of subunit exchange, MEC217 cells were cultured in the presence of graded concentrations of TET and then subjected to pulse-chase analysis (Fig. 2⇓). As expected, high concentrations of TET (400 ng/ml) fully suppressed expression of the transfected subunits, resulting in a 100% constitutive proteasome population. In contrast, culture in 10 ng/ml TET effected a roughly 55% replacement of constitutive for induced subunits among newly assembled proteasomes, as measured by quantifying the radioactive signals for δ and LMP2 (Fig. 2⇓). In the absence of TET, a nearly complete exchange of δ for LMP2 and MB1 for LMP7 was reached (Fig. 2⇓). Thus, LMP2, LMP7, and MECL-1 efficiently replace their constitutive homologues δ, MB1, and MC14 in induced MEC217 cells.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Immunosubunit incorporation in 20S proteasomes of MEC217 cells in response to the TET concentration in the culture medium. Cells were grown in the absence or in the presence of 10 or 400 ng/ml TET for 10 h, labeled o/n with [³⁵S]methionine, and then chased for 4 h. 20S proteasomes were precipitated from the detergent lysates with a 20S-specific polyclonal rabbit antiserum and separated by two-dimensional NEPHGE/PAGE. The gels were exposed for autoradiography and phosphor imager analysis. Proteasome subunits were assigned on the basis of their migrational behavior as indicated (36 ).

To estimate the effect of induced immunosubunit expression on the total cellular proteasome population, MEC217 cells were grown in bulk culture in the absence and in the presence of 400 ng/ml and 10 ng/ml TET for 3 days, and 20S proteasomes were isolated. The exchange rates of constitutive for induced subunits LMP2 and LMP7, estimated from Coomassie-stained two-dimensional gels, were ∼0, 40, and 60% for cultures with 400 ng/ml TET, 10 ng/ml TET, and without TET, respectively (not shown). Because replacement of δ reduces the proteasomal peptidyl glutamyl hydrolyzing acivity (14), the subunit exchange efficiencies were further established by measuring the enzymatic activity of the isolated proteasomes against Z-LLE-βNA. Whereas constitutive proteasomes cleaved this substrate efficiently, immunoproteasomes of cells cultured with 10 ng/ml TET or without TET hydrolyzed Z-LLE-βNA to relative levels of roughly 50 and 20% only (Fig. 3⇓), reflecting the gradual substitution of δ for LMP2 in the proteasome preparates. Thus, these experiments (Figs. 2⇑ and 3⇓) reveal an efficient assembly of immunoproteasomes in induced MEC217 cells that can be leveled by titration of the TET concentration in the medium.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

20S proteasomes of uninduced and induced MEC217 cells differ in peptidyl glutamyl hydrolyzing activity. MEC217 cells were grown at the indicated TET concentrations for 3 days, and 20S proteasomes were isolated. Hydrolytic activity was assayed by incubating 200 ng purified 20S proteasomes with Z-LLE-βNA substrate (300 μM) in a volume of 100 μl for 60 min at 37°C. Release of the fluorescent βNA group was measured with a spectrofluorometer at 355 nm excitation/405 nm emission (y-axis). Mean values of duplicate wells are plotted.

Immunosubunits enhance MHC class I Ag processing in infected cells

To examine the effects of LMP2, LMP7, and MECL-1 on MHC class I Ag processing, we used Ad5, against which specific CTL clones exist as a model system. MEC217 cells were cultured in the presence or absence of TET and infected with Ad5 mutant dl7001. Dl7001 lacks the viral E3 region and thus the E3-19K protein, which retains MHC class I molecules in the ER. Viral infection by itself did not affect proteasome subunit composition in MEC217 cells (data not shown). Presentation of the viral E1B_192–200 epitope by H-2D^b MHC class I molecules was tested in an Eu³⁺-release assay with CTL clone 100B6 (Fig. 4⇓A). Whereas uninduced infected MEC 217 cells (+TET) were not lysed above background levels, induced infected cells (−TET) expressing the proteasome immunosubunits appeared highly susceptible to lysis by E1B_192–200-specific CTL. Infected MEC-29 cells containing tTAk only and cultured with and without TET were not lysed (not shown), indicating that tTAk expression or the presence of TET did not influence the results. Observed effects were reproduced in a second independent experiment (not shown). Thus, the presence of immunoproteasomes enhances the generation of E1B_192–200 in infected cells.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Immunosubunit expression enhances E1B_192–200 presentation by Ad5 dl7001-infected MEC217 cells. MEC217 cells were cultured for 3 days with or without TET and were then infected with Ad5 dl7001 or were left uninfected. After 24 h of infection, the cells were harvested and used as targets in an Eu³⁺-release cytotoxicity assay with CTL clone 100B6. Mean percentages of specific lysis of triplicate wells (y-axis) for different E:T ratios (x-axis) are plotted (A). MEC217 cells were cultured for 2 days with or without TET and were then infected with Ad5 dl7001 or were left uninfected. After 12 h of infection, the cells were harvested and tested for recognition by CTL clone 100 PKOB4 in a TNF-α production assay. Target cells and CTL were coincubated for the indicated time periods, and TNF-α release into the supernatants was assayed by measuring the cytotoxic effect on WEHI cells. Remaining metabolic activity in WEHI cells was quantified by the conversion of MTT substrate into a blue salt, measured with an ELISA reader at 570 nm. The OD570s, subtracted with the background values of wells with total WEHI cell death, are depicted on the y-axis (B). It is of note that these values are reciprocal to the quantity of TNF-α release by activated CTL.

To confirm and expand these results, we performed a more sensitive assay that measures TNF-α production by activated CTL. MEC217 cells cultured for 2 days with different concentrations of TET (400 ng/ml and 10 ng/ml) or without TET were infected with Ad5 dl7001 and then cocultured for different time periods with E1B-specific CTL clone 100 PKOB4. TNF-α secretion was measured (Fig. 4⇑B). Because 100 PKOB4 CTL lack perforin expression and thus do not lyse their target cells efficiently and because Ad5 infection is not lytic in mouse cells, decreases in target cell numbers during the assay were only minimal. Both uninduced and induced infected MEC217 cells were recognized by E1B_192–200-specific CTL (Fig. 4⇑B). However, cells expressing LMP2, LMP7, and MECL-1 induced significantly higher levels of TNF-α release than did cells cultured with 400 ng/ml TET, already at early time points after infection. Thus, the presence of immunoproteasomes enhanced the processing efficiency of E1B_192–200, resulting in a faster recognition of infected cells by virus-specific CTL. Moreover, the kinetics of CTL activation by cells cultured in the presence of 10 ng/ml TET and in the absence of TET appeared identical (Fig. 4⇑B), indicating that the lower levels of immunosubunits sufficed to effect efficient E1B_192–200 presentation.

Enhanced antigenic peptide generation by immunoproteasomes

To explain the observed enhancement of MHC class I Ag processing in induced MEC217 cells, the effects of the inducible proteasome subunits on E1B degradation and E1B_192–200 generation were further analyzed in vitro. A 40-mer E1B polypeptide (E1B_176–215) encompassing the E1B CTL epitope was synthesized and incubated with purified 20S proteasomes from induced and uninduced MEC217 cells. Generated digestion products were analyzed by RP-HPLC and on-line analyzed with a mass spectrometer. The identity of detected peptide fragments was confirmed by ms/ms analysis. Remarkably and consistent with earlier findings with murine CMV pp89 25-mer (U. Kuckelhorn and P. M. Kloetzel, manuscript in preparation), we found that over the course of digestion (as measured at 3 and 6 h), immunoproteasomes degraded the E1B 40-mer faster than constitutive proteasomes did (not shown). After 3 h of digestion, when turnover of the E1B 40-mer was still incomplete, the E1B_192–200 flanking fragments E1B_176–191 and E1B_201–215 were among the predominant peptide products (data not shown). The E1B epitope VNIRNCCYI was just discernable from background in the HPLC-mass spectrometry of digests of constitutive proteasomes (Fig. 5⇓, upper left panel; at 21.63 min). In contrast, immunoproteasomes produced this peptide efficiently, as is evidenced by a clearly distinguishable peak at 21.74 min. (Fig. 5⇓, upper right panel). The same relative differences in E1B_192–200 abundancies were found in the 6-h-digestion products (Fig. 5⇓, lower panels). Because not only E1B_192–200 but also the N- and C-terminal flanking fragments (176–191 and 201–215, respectively) were most prevalent in immunoproteasome digests, we infer that the observed better liberation the E1B epitope is caused by a more efficient usage of the Leu¹⁹¹-Val¹⁹² and Ile²¹⁵-Ser²¹⁶ cleavage sites. Peptides with the correct COOH terminus of the epitope but prolonged at the N terminus were not detected, implying that both types of proteasomes generated the E1B CTL epitope at exact size.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Proteasome-mediated generation of E1B_192–200 from E1B 40-mer in vitro. Synthetic E1B_176–215 was digested at 37°C with 20S proteasomes isolated from MEC217 cells cultured in the presence or in the absence of TET. Cleavage products were identified by mass spectrometry after HPLC separation as described in Materials and Methods. Shown are the relative abundances of E1B_192–200 (retention time of 21.63–21.74 min) after 3 h of digestion (upper panel) and of E1B_192–200, E1B_176–191 (retention time of 19.10–19.18 min), and E1B_201–215 (retention time of 18.06–18.07 min) after 6 h of digestion (lower panel) as deduced from the HPLC UV profiles. The identity of each of these fragments was confirmed by ms/ms.

The relative efficiencies of E1B_192–200 production by different proteasome populations were further analyzed using E1B_192–200-specific CTL as read out. First, peptide fragments generated during a 90-min digestion of E1B 40-mer with constitutive and with immunoproteasomes were separated by RP-HPLC, and targeting activity of the individual fractions was assayed with E1B_192–200-specific T cells. As expected, only one single fraction which corresponded to the fraction in which synthetic E1B_192–200 eluted sensitized uninfected target cells for T cell recognition (data not shown), supporting our notion that both constitutive and immunoproteasomes generate E1B_192–200 precisely.

To minimize potential losses during HPLC purification, E1B_192–200 generation was quantified from the unfractionated peptide digests. Dilutions of digests generated with 20S proteasomes of MEC217 cultured in the presence of 400 or 10 ng/ml TET or in the absence of TET were loaded on uninfected MEC217 cells and cocultured with CTL clone 100 PKOB4, and TNF-α release into the culture supernatant was determined using WEHI cells. Relative antigenic peptide quantities (Fig. 6⇓) were calculated from a standard curve generated with synthetic E1B_192–200 in the same experiment. Two independent measurements showed that proteasomes of cells grown without TET produced the E1B epitope 3-fold more efficiently than proteasomes of cells grown with 10 ng/ml TET and 10-fold more efficiently than proteasomes of uninduced cells. Thus, the incorporation of LMP2, LMP7, and MECL-1 into the proteasome enhanced E1B_192–200 generation (Fig. 6⇓) in a dose-responsive fashion.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Immunosubunit incorporation enhances E1B_192–200 generation in vitro. E1B_176–215 was digested for 60 min with purified 20S proteasomes of MEC217 cells cultured at the indicated TET concentrations. To quantify E1B_192–200 generation, the cleavage products were loaded on uninfected cells and tested for targeting activity in a TNF-α production assay with CTL clone 100 PKOB4. The molar amounts of antigenic peptide in the E1B_176–215 digestion reactions (y-axis) were calculated from a standard curve generated with synthetic E1B_192–200.