The synthesis of truncated polypeptides for immune surveillance and viral evasion

Abstract

Background: Cytotoxic T cells detect intracellular pathogens by surveying peptide loaded MHC class I molecules (pMHC I) on the cell surface. Effective immune surveillance also requires infected cells to present pMHC I promptly before viral progeny can escape. Rapid pMHC I presentation apparently occurs because infected cells can synthesize and present peptides from antigenic precursors called defective ribosomal products (DRiPs). The molecular characteristics of DRiPs are not known.

Methodology/principal findings: Here, using a novel method for detecting antigenic precursors and proteolytic intermediates, we tracked the synthesis and processing of Epstein-Barr Virus encoded nuclear antigen 1 (EBNA1). We find that ribosomes initiated translation appropriately, but rapidly produced DRiPs representing approximately 120 amino acid truncated EBNA1 polypeptides by premature termination. Moreover, specific sequences in EBNA1 mRNA strongly inhibited the generation of truncated DRiPs and pMHC I presentation.

Significance: Our results reveal the first characterization of virus DRiPs as truncated translation products. Furthermore, production of EBNA1-derived DRiPs is down-regulated in cells, possibly limiting the antigenicity of EBNA1.

Figure 1. Schematic representation of cDNA constructs encoding EBNA1 and its EBNA1ΔGA derivative lacking the glycine-alanine repeat (GAr).

To track the translated products, the OVA-derived epitope (SIINFEHL or SHL8) was inserted in either the N-terminal (positionα) or C-terminal region (positionγ) of EBNA-GFP or its ΔGA derivative. The SHL8 antigenic peptide was flanked at its N- and C-termini by lysines (QLK- and –KEW) to allow detection of potential antigenic intermediates containing the SHL8 peptide by enzymatic methods. The number of amino acids and the predicted molecular weights of putative polypeptides are shown. (B) In vitro expression EBNA1 and EBNA1ΔGA full-length proteins. The mRNAs encoding EBNA1 (α or γ) and EBNA1ΔGA (α or γ) were transcribed in vitro, purified and translated in a rabbit reticulocyte lysate in the presence of radiolabeled ³⁵S-Met or ³⁵S-Cys as tracers. The translated products were fractionated on 7.5% SDS-PAGE gels and detected by PhosphoImager. As controls, translation was carried out in parallel without mRNA (no mRNA) or with mRNAs encoding the 27kD KOVAK protein. Arrows indicate the predominant products from translation of ΔGAα, ΔGAγ and KOVAK mRNA (73 and 27 kD respectively). The expected location of the 86 kD product of EBNAα and EBNAγ translation is indicated by an asterisk. Autoradiographs on right show the KOVAK translated products fractionated on higher resolution 16.5% SDS-PAGE. Data shown is representative of 3 different experiments.

Figure 2. GAr inhibits overall translation of EBNA1 mRNA without affecting ribosomal initiation.

(A) Schematic of the primer extension inhibition (toeprint) assay illustrating how presence of ribosomes bound to the mRNA initiation codon is detected by analysis of reverse transcriptase (RT) products of varying length. (B) The mRNA encoding EBNA1 (α or γ) and EBNA1ΔGA (α or γ) were used as template for the toe-printing assay. Reactions were carried out with either no ribosomes, or in presence of elongation inhibitors, cycloheximide (CHX), CHX plus sparsomycin (SPR), initiation inhibitor, edeine or magnesium chelator, EDTA. Dried gels were analyzed using a PhosphoImager. The full-length and shorter RT (toeprint) products are indicated by arrows. The toeprint corresponds to 17 nucleotides downstream of the AUG codon as judged by the size of fragments on the sequencing gel shown on the right. (C) The intensity of the ³²P radiolabeled toeprint bands observed under the indicated conditions was quantitated and is shown as arbitrary units (AU.) Data are representative of 3 different experiments.

Figure 3. Ribosomes prematurely terminate translation within the 5′ region of EBNA mRNA.

(A) Schematic model to account for relative differences in the translational efficiency of EBNA or ΔGA mRNAs. The ribosomes initiate translation at the AUG codon (5′ blue dot) of either EBNA or ΔGA mRNAs. However, translation is terminated within or close to the GAr region shown as [GAGA]n in the EBNA mRNA. The methionine and leucine codons are indicated as blue or green dots respectively. The inserted SHL8 nucleotides, at either the N-terminal “α” or the C-terminal “γ” positions are shown as a red rectangle. All constructs included the indicated in-frame green fluorescent protein (GFP) sequence. (B) The mRNA coding for EBNAα, ΔGAα, EBNAγ or ΔGAγ were translated in vitro in the presence of ³⁵S-Met or ³H-Leu for 1 h. Translation products were precipitated with trichloroacetic acid and collected by filtration. Radioactivity measured by liquid scintillation is shown as relative counts per minute compared to no mRNA control. Bars are average of 5 different experiments (± standard deviation). Standard t-test was used to calculate statistically significant differences shown as asterisks (***: p<0.001).

Figure 4. Truncated translation products are antigenic precursors.

(A) Schematic illustration of the KOVAK method for detection and quantitation of proteolytic intermediates. The SHL8 (SIINFEHL) codons (red circle) are flanked at the N- and C-termini by lysine (K) residues. The putative products of translation are collected in 10% acetic acid before HPLC fractionation. Each fraction is treated with trypsin and carboxypeptidase B (CPB) to release the optimally active SHL8 peptide. (B) The amount of embedded SHL8 in the proteolytic intermediates (“SHL8” activity) is shown in two representative HPLC fractions (#9 and #18). The B3Z stimulating activity in dilutions of fraction #9 and #18 was compared with a standard curve generated with synthetic SHL8 peptide. (C) The indicated mRNA coding for EBNAα, EBNAγ, ΔGAα, ΔGAγ and luciferase as a negative control, were translated in vitro for varying time periods. Polypeptides in aliquots of the translation reactions were extracted assayed without HPLC fractionation as in (A). (D) Products translated in 60 min. in vitro from the indicated mRNAs were passed through a 30kD molecular weight filter. The filtrate (<30kD) was fractionated by HPLC and the amount of “SHL8” activity in each fraction was determined. Data shown represent 3 different experiments. (E) Polypeptides translated from EBNAα mRNA and eluting in “7–11” HPLC fractions in (C) were fractionated on high resolution tricine 16.5% SDS-PAGE gels. “No mRNA” samples were analyzed in parallel as negative controls. The silver-stained gel shown is representative of 3 different experiments. (F) Indicated slices (S1–S4) were excised from the gel, treated with trypsin and CPB and dilutions were tested for “SHL8” activity with SHL8/K^b-specific B3Z lacZ inducible T cell hybridoma and K^b-L cells as APC. The conversion of the lacZ substrate chlorophenol β-D-pyrannoside was measured as absorbance at 595 nm with 655 nm as reference wave length. Data are representative of 3 different experiments.

Figure 5. Generation of DRiPs in living cells.

(A) HEK293 cells were induced to express EBNA or ΔGA with SHL8 in position “α” (upper panel) or “γ” (lower panel) with doxycycline for 24 h. The cell extracts were passed through 30kD molecular weight filters and the <30kD filtrates were fractionated by HPLC. Each HPLC fraction was treated with trypsin and CPB and the amount peptides containing “SHL8” was determined. (B) The total amount of “SHL8” activity recovered in the polypeptides eluting in all HPLC fractions versus fractions “7–11” and “15–23”. Data represent 3 different experiments. (C) HEK293 cells expressing vector alone, EBNAα or ΔGAα were cultured for 24h with doxycycline. The cell extracts were passed through 30 kD molecular weight filters and fractionated by HPLC. The polypeptides in fractions “7–11” were pooled and separated on tricine 16.5% SDS-PAGE. The silver-stained gels shown is representative of three different experiments. (D) The indicated slices (S1–S4) were cut from the gel and digested for 8 h by trypsin. Each digested slice was split into two aliquots. Aliquots from the indicated vector, EBNAα or ΔGAα samples were further digested with CPB for 4 h. The “SHL8” activity was measured in the indicated dilutions using B3Z T cells and K^b-L cells as APC. The B3Z response is shown as products of the lacZ substrate CPRG as in Fig. 4. (E) The presence of the GAr motif was detected by an ELISA assay using a mouse anti-GAr antibody and anti-mouse conjugated to horse-radish peroxidase. Data shows the absorbance (A450) of the peroxidase substrate converted in 30 min.

Figure 6. Newly synthesized precursors are processed rapidly to the final SHL8 peptide.

(A) HEK293 cells expressing K^b MHC together with EBNAα (red) or ΔGAα (green) were treated with doxycycline. At the indicated time points, peptides were extracted from the cells and fractionated by reverse phase HPLC. The fractions were assayed without any further treatment using B3Z T cells and K^b-L cells as APC and quantitated using synthetic SHL8 as a standard (upper panel). The total SHL8 recovered at different time points is indicated in lower panel. Data are representative of 3 different experiments. (B) Protein expression was induced in the same cells with doxycycline for 3 h. The cells were treated (+) with the protein synthesis inhibitor cycloheximide or used as such (−). The total amount of SHL8 eluted in fractions 17 to 22 was measured at 0 h, 3 h and 6 h. Arrow indicates addition of cycloheximide. Data shown are typical of three different experiments. (C) HEK293 cells expressing EBNAα (red) or ΔGAα (green) were treated with doxycycline for the indicated times. The polypeptides were extracted and “SHL8” activity was quantitated as described in Fig. 4. Total “SHL8” activity in fractions “7–11” are presented in lower panel. Data is from three different experiments. (D) Same cells were stimulated 3 h with doxycycline following same treatments than in B. The “SHL8 activity” in fractions “7–11” was measured at 0 h, 3 h and 6 h, in the presence or not of cycloheximide. Arrow indicates addition of drug. Data shown are typical of three different experiments.