Fitness costs limit viral escape from cytotoxic T lymphocytes at a structurally constrained epitope

Abstract

The intense selection pressure exerted by virus-specific cytotoxic T lymphocytes (CTL) on replicating human immunodeficiency virus and simian immunodeficiency virus results in the accumulation of CTL epitope mutations. It has been assumed that fitness costs can limit the evolution of CTL epitope mutations. However, only a limited number of studies have carefully examined this possibility. To explore the fitness costs associated with viral escape from p11C, C-M-specific CTL, we constructed a panel of viruses encoding point mutations at each position of the entire p11C, C-M epitope. Amino acid substitutions at positions 3, 4, 5, 6, 7, and 9 of the epitope significantly impaired virus replication by altering virus production and Gag protein expression as well as by destabilizing mature cores. Amino acid substitutions at position 2 of the epitope were tolerated but required reversion or additional compensatory mutations to generate replication-competent viruses. Finally, while amino acid substitutions at positions 1 and 8 of the p11C, C-M epitope were functionally tolerated, these substitutions were recognized by p11C, C-M-specific CTL and therefore provided no selection advantage for the virus. Together, these data suggest that limited sequence variation is tolerated by the region of the capsid encoding the p11C, C-M epitope and therefore that only a very limited number of mutations can allow successful viral escape from the p11C, C-M-specific CTL response.

FIG. 1.

Sequence alignment of the p11C, C-M region from various retrovirus capsid proteins. Consensus SIV and HIV sequences were obtained from the Los Alamos HIV Sequence Database (http://www.hiv.lanl.gov), and the additional protein sequences of other retroviruses were obtained from the National Center for Biotechnology Information Entrez database. Sequences were aligned with DNAStar MegAlign by the ClustalW method and were visualized with BOXSHADE (http://www.ch.embnet.org/software/BOX_form.html). Black shading indicates sequence identity; gray shading indicates sequence similarity. The following viruses were used: FIV, feline immunodeficiency virus (accession no. P16087); EIAV, equine infectious anemia virus (accession no. AAK21105); ovine lentivirus (accession no. NP_041249); Visna virus (accession no. NP_040839); BIV, bovine immunodeficiency virus (accession no. NP_040562); Jembrana disease virus (accession no. NP_042684); HTLV-1, human T-lymphotropic virus type 1 (accession no. NP_057862); HTLV-2 (accession no. P03346); STLV-3, simian T-lymphotropic virus type 3 (accession no. AAO62100); MLV, murine leukemia virus (accession no. P29167).

FIG. 2.

Mutations in the p11C, C-M epitope decrease protein expression but do not affect virion release. (A and B) 293T cells were transfected with 10 μg of full-length proviral plasmid DNA containing the indicated Gag point mutations. Forty-eight hours later, supernatants (A) and cell lysates (B) were analyzed by Western blotting. (C) Virion release efficiencies were calculated as the total amounts of Gag in the supernatants divided by the total amounts of Gag in both cell lysates and supernatants, as determined for the gels shown in panels A and B. The values used for these calculations were arbitrary intensity units per square millimeter that were derived by using Quantity One image analysis software.

FIG. 3.

Mutations in the p11C, C-M epitope decrease virus production. 293T cells were transfected with 10 μg of full-length proviral SIVmac239 plasmid DNA (A and B) or single-round SIV.GFP plasmid DNA (C) containing the indicated Gag point mutations. Forty-eight hours later, the supernatants were harvested and filtered through a 0.45-μm-pore-size filter. Cell-free supernatants were analyzed for virus production by a colorimetric RT assay (A) and SIV p27 ELISA (B and C). The values illustrated are the means ± standard errors of duplicates.

FIG. 4.

p11C, C-M epitope mutations alter SIV core stability. 293T cells were transfected with 10 μg of full-length proviral plasmid DNA containing the indicated Gag point mutations. Forty-eight hours later, the viruses were harvested and pelleted through a 20% sucrose cushion. The wild-type virus was layered onto 30 to 70% linear sucrose gradients with (B) and without (A) a layer of 0.5% NP-40. After ultracentrifugation at 100,000 × g for 20 h at 16°C, 1-ml fractions were collected from the top of the gradient and analyzed by p27 ELISA (filled columns). The density of each fraction was determined by refractometry (open diamonds). (C) Wild-type and mutant viruses were concentrated and ultracentrifuged through a 0.5% NP-40 layer into a linear 30 to 70% sucrose gradient. Percent yield of cores, percentage of p27 detected in the core fraction, as determined in panel B. The values illustrated in panel C are means ± standard deviations of three experiments. WT, wild type.

FIG. 5.

Mutations in the p11C, C-M epitope decrease the SIV replication rate. Five micrograms of p239SpSp5′ containing the noted gag point mutations and 5 μg of p239Sp3′ were digested, extracted with phenol-chloroform, ethanol precipitated, ligated together, and transfected into CEMx174 cells by the DEAE-dextran method. Cell-free supernatants were monitored for the SIV p27 antigen (A and B) and for RT activity (C and D). The values illustrated are means ± standard deviations of triplicates. (E) Sequence analysis of five to eight clones determined at the peak of viral replication for viruses that replicated (WT [wild type], C46A, T47A, N52A, Q53A, and Q53T). The deduced amino acid sequences for 60 amino acids are shown. The Mamu-A*01-restricted p11C, C-M epitope is bracketed by vertical lines. Amino acids that differ from the wild type are shown in bold and underlined. Numbers in parentheses indicate the number of mutant clones divided by the total number of clones analyzed for each mutant.

FIG. 6.

The p11C, C-M epitope of replication-competent mutant SIVs is recognized by epitope-specific CTL. PBL from a SIVmac251-infected, Mamu-A*01⁺ rhesus monkey were stimulated in vitro with the SIV Gag p11C, C-M peptide. On day 17 of culture, the lymphocytes were assessed as effector cells in a standard 4-h ⁵¹Cr-release CTL assay. The target cells were 721.221 cells stably expressing the Mamu-A*01 molecule that had been incubated for 1.5 h with the indicated peptides at concentrations of 1, 10, 100, and 1,000 pM.