• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of jvirolPermissionsJournals.ASM.orgJournalJV ArticleJournal InfoAuthorsReviewers
J Virol. Apr 2011; 85(8): 3746–3757.
PMCID: PMC3126128

Epitope-Specific CD8+ T Lymphocytes Cross-Recognize Mutant Simian Immunodeficiency Virus (SIV) Sequences but Fail To Contain Very Early Evolution and Eventual Fixation of Epitope Escape Mutations during SIV Infection [down-pointing small open triangle]


Human immunodeficiency virus (HIV) and simian immunodeficiency virus (SIV) evade containment by CD8+ T lymphocytes through focused epitope mutations. However, because of limitations in the numbers of viral sequences that can be sampled, traditional sequencing technologies have not provided a true representation of the plasticity of these viruses or the intensity of CD8+ T lymphocyte-mediated selection pressure. Moreover, the strategy by which CD8+ T lymphocytes contain evolving viral quasispecies has not been characterized fully. In the present study we have employed ultradeep 454 pyrosequencing of virus and simultaneous staining of CD8+ T lymphocytes with multiple tetramers in the SIV/rhesus monkey model to explore the coevolution of virus and the cellular immune response during primary infection. We demonstrated that cytotoxic T lymphocyte (CTL)-mediated selection pressure on the infecting virus was manifested by epitope mutations as early as 21 days following infection. We also showed that CD8+ T lymphocytes cross-recognized wild-type and mutant epitopes and that these cross-reactive cell populations were present at a time when mutant forms of virus were present at frequencies of as low as 1 in 22,000 sequenced clones. Surprisingly, these cross-reactive cells became enriched in the epitope-specific CD8+ T lymphocyte population as viruses with mutant epitope sequences largely replaced those with epitope sequences of the transmitted virus. These studies demonstrate that mutant epitope-specific CD8+ T lymphocytes that are present at a time when viral mutant epitope sequences are detected at extremely low frequencies fail to contain the later accumulation and fixation of the mutant epitope sequences in the viral quasispecies.


CD8+ T lymphocytes play a critical role in controlling viral replication following human immunodeficiency virus (HIV) and simian immunodeficiency virus (SIV) infection (14, 22, 27). It is well documented that CD8+ cytotoxic T lymphocyte (CTL) responses impose a strong selection pressure on the replicating virus during early infection. Virus mutations accumulate at major histocompatibility complex (MHC) class I-restricted epitopes, and these virus sequence changes reduce the capacity of CD8+ T lymphocytes to recognize epitope peptides on the surface of virally infected cells (2, 9). This can result in the loss of CD8+ T lymphocyte-mediated control of viral replication and eventual clinical deterioration (4, 10).

Viral epitope evolution in HIV-infected humans and SIV-infected monkeys has been studied by sequencing virus clones sampled from plasma. While traditional sequencing technologies have demonstrated virus escape from epitope-specific CD8+ T lymphocytes in HIV/SIV-infected individuals, the completeness of that escape remains unclear. The assessment of viral sequence diversity using traditional methods is, in the end, limited by the relatively small number of virus clones that can be sequenced. These selected clones may not fully represent the breadth of mutations present in a viral population at a given time following infection. Emerging sequencing technologies, such as 454 pyrosequencing, make it possible now to sample a vastly greater number of clones using high-throughput platforms and therefore to better characterize the diversity of viral epitope sequences that emerge during infection (6, 12).

The SIV Nef p199RY CTL epitope can mutate under pressure from CTLs in rhesus monkeys infected with SIVmac239 or SIVmac251 and therefore provides a useful model for studying viral escape at MHC class I-restricted epitopes (24, 33). The 9-amino-acid epitope sequence is restricted by the rhesus monkey MHC class I allele Mamu-A*02, which is expressed in approximately 20% of Indian-origin rhesus monkeys (25, 33). p199RY-specific CD8+ T lymphocyte responses are dominant in monkeys that express Mamu-A*02 (24, 33) but not to the extent seen in HLA-B57-restricted HIV type 1 (HIV-1) Gag-specific or Mamu-A*01-restricted Tat- or Gag-specific responses. In contrast to other reported instances of virus epitope escape from CTLs, studies have demonstrated that the affinity of p199RY epitope peptides for Mamu-A*02 is not affected by amino acid substitutions in the epitope that typically evolve in infected monkeys (33).

The present studies were undertaken to examine how the cellular immune response deals with viral escape variants as they emerge in the viral population over the course of an AIDS virus infection. We did this by exploring the evolution of SIVmac251 and p199RY emergence of epitope-specific CD8+ T cell populations in Mamu-A*02+ SIVmac251-infected rhesus monkeys. To define comprehensively the breadth of p199RY epitope variation during the course of SIVmac251 infection, viral RNA was sequenced using 454 pyrosequencing technology. These studies showed that the p199RY sequence in the virus inoculum was largely replaced by epitope sequences containing mutations by 84 days after infection. Surprisingly, some of these mutations were selected in spite of preexisting CD8+ T lymphocyte populations that recognize the emerging epitope mutations.


Animals and SIVmac251 challenge.

All animals used in this study were Indian-origin rhesus monkeys (Macaca mulatta). They were housed in accordance with the guidelines of the Institutional Animal Care and Use Committee for Harvard Medical School and the Guide for the Care and Use of Laboratory Animals (22a). Monkeys were challenged intravenously (i.v.) with a 1/3,000 dilution of an uncloned SIVmac251 inoculum (19).

Sanger sequencing of the SIV Nef p199RY epitope in plasma viral RNA.

Viral RNA Sanger sequencing was carried out as described previously (24). Approximately 10 to 12 clones were selected from each plasma sample for sequencing.

Peptides and oligonucleotides.

Synthetic peptides were obtained at >95% purity from Quality Control Biochemicals or New England Peptide, LLC. These peptides included the wild-type (WT) SIVmac251 Nef p199RY (YTSGPGIRY) and two p199RY epitope sequence variants, T2S (YSSGPGIRY) and Y9F (YTSGPGIRF). High-performance liquid chromatography-purified oligonucleotides were obtained from Biosource International or Invitrogen.

Peptide-MHC class I tetramer construction.

To construct Mamu-A*02 tetrameric complexes with variant p199RY epitope peptides, a Mamu-A*02 fusion gene plasmid containing a small portion of the α3 domain of Mamu-A*01 (24) was modified to increase biotinylation efficiency. A 3-amino-acid linker between the BirA substrate peptide and the Mamu-A*02/A*01 fusion sequence was introduced by PCR mutagenesis using the Stratagene QuikChange II kit (Agilent Technologies) and the oligonucleotide primers pMN1insertF (5′-CCTGAAATGGGAGCCGTCTTCCCAGGGATCCCTGCATCATATTCTGG-3′) and pMN1insertR (5′-CCAGAATATGATGCAGGGATCCCTGGGAAGACGGCTCCCATTTCAGG-3′) according to the manufacturer's directions. The expressed protein was refolded in vitro with human β2-microglobulin in the presence of WT or mutant p199RY epitope peptides, and the resulting complexes were biotinylated as previously described (22). Monomers were mixed with fluorophore-labeled streptavidin (SA) at a 4:1 molar ratio and titrated dropwise to maximize tetramer formation. All tetramers were stored and used for lymphocyte staining as described previously (24).

Tetramers and antibodies.

CD8+ T lymphocytes were stained with a cocktail consisting of the WT p199RY tetramer and tetramers bearing two prototypic mutant epitope peptides, T2S and Y9F. Biotinylated WT p199RY-Mamu-A*02 monomers were mixed with either phycoerythrin (PE)-labeled SA (ProZyme) or PE-cyanine 7 (PE-Cy7)-labeled SA (BD Biosciences), T2S p199RY-Mamu-A*02 tetramers were mixed with allophycocyanin (APC)-labeled SA (ProZyme), and Y9F p199RY-Mamu-A*02 tetramers were mixed with PE-labeled SA. The Nef p56-Mamu-A*02 monomer was refolded previously (24) and mixed with APC-labeled streptavidin.

PE-Cy7-labeled WT p199RY-, APC-labeled T2S p199RY-, and PE-labeled Y9F p199RY-Mamu-A*02 tetrameric complexes were used with anti-CD8α-PE-Texas Red (in-house production of clone 7pt3F9 custom conjugated by Beckman Coulter) and anti-CD3-Alexa Fluor 700 (SP34-2; BD Biosciences) to stain WT and mutant p199RY-specific CD8+ T lymphocytes. PE-labeled WT p199RY and APC-labeled p56 tetrameric complexes were used with anti-CD8α-fluorescein isothiocyanate (FITC) (SK1; BD Biosciences) and anti-CD3-APC-Cy7 (SP34-2; BD Biosciences) to stain WT p199RY-specific and Nef p56-specific CD8+ T lymphocytes.

Tetramer staining of epitope specific CD8+ T lymphocytes and flow cytometric analysis.

Peripheral blood mononuclear cells (PBMCs) were separated from EDTA-preserved whole blood by Ficoll density gradient centrifugation. Cells were resuspended in 100 μl of phosphate-buffered saline (PBS) supplemented with 2% fetal calf serum (FCS) and then incubated first with WT/T2S/Y9F p199RY tetramer cocktail or WT p199RY/p56 tetramer cocktail for 15 min at room temperature and then with the appropriate anti-CD3/anti-CD8α antibody cocktail for 15 min at room temperature. Cells were washed once with PBS supplemented with 2% FCS and fixed in PBS with 1% formaldehyde. Samples were collected on an LSRII instrument (BD Biosciences) and analyzed using FlowJo software (Tree Star).

PCR preamplification of plasma viral RNA for 454 pyrosequencing.

Plasma was separated from EDTA-preserved whole blood by Ficoll density gradient centrifugation, and viral RNA was extracted from 140 μl of plasma using the viral RNA minikit (Qiagen) according to the manufacturer's directions and eluted in a 60-μl volume. cDNA was synthesized using Superscript III reverse transcriptase (Invitrogen) under the following conditions: 60 μl of purified viral RNA, 1× reverse transcriptase buffer, 0.5 μM antisense primer 9945R (5′-ATCAAGAAAGTGGGCGTT-3′), 0.5 mM each deoxynucleoside triphosphate (dNTP) (Invitrogen), 2.5 U/μl RNase OUT (Invitrogen), 6 mM dithiothreitol (DTT), and 9 U/μl Superscript III reverse transcriptase. The viral RNA and primers were first combined and denatured at 65°C for 3 min and then immediately placed on ice until they were added to the reaction mixture. The entire reaction mixture was incubated at 50°C for 60 min, followed by an increase in temperature to 55°C for 60 min. The reaction mixture was heat inactivated at 70°C for 10 min and then treated with 2 units of RNase H (Invitrogen) at 37°C for 20 min.

The cDNA was immediately subjected to nested PCR amplification using the high-fidelity DNA polymerase Phusion (New England BioLabs). For each sample, 20 replicate PCRs (40 μl each) were performed with 5 μl cDNA, 1× Phusion HF buffer, 0.2 mM each dNTP, 2.5% dimethyl sulfoxide (DMSO), 0.6 unit of Phusion DNA polymerase, 0.5 μM antisense primer 9949R, and 0.5 μM sense primer 9278F (5′-TCCATGGAGAAACCCAGCT-3′). The first-round PCR was performed using the following parameters: 1 cycle of 99°C for 30 s; 20 cycles of 99°C for 8 s, 59°C for 20 s, 72°C for 20 s; and 1 cycle of 72°C for 10 min. All first-round products of each cDNA sample were pooled and used as the template for inner PCRs. Inner PCRs were performed in 32 replicates (40 μl each) with 5 μl of pooled first-round template, 1× Phusion HF buffer, 0.2 mM each dNTP, 2.5% DMSO, 0.6 unit of Phusion DNA polymerase, 0.5 μM sense primer 5′-GCCTCCCTCGCGCCATCAGXXXXXXAGAATCTTAGARATGTACTTAG-3′, and 0.5 μM antisense primer 5′-GCCTTGCCAGCCCGCTCAGXXXXXXGAGTTGGATCAAACTTCCACGC-3′, where “XXXXXX” represents the appropriate 454 identification tag in each sense and antisense primer pair used to amplify each sample. The following 15 unique 6-nucleotide identification tags were used to preamplify each viral cDNA sample (five monkeys and three time points): ACGTCT, AGATGC, ATCACT, ACACTC, AGTATC, AGTAGC, ACAGCT, CGCGAT, CTATGC, TATCAT, TCTGAC, TGCACT, TGACGC, CATCTC, and AGCTGT. PCR parameters used for the second-round amplification were identical to those used in the first-round cDNA amplification. PCR amplicons were pooled and visualized with crystal violet-stained agarose gels illuminated on a white light box. Amplicons of the correct size were excised and purified using the QIAquick gel extraction kit (Qiagen) according to the manufacturer's directions.

454 pyrosequencing of preamplified cDNA.

We initially used a Bioanalyzer (Agilent Technologies) to determine sequence quality and quantity for preamplified samples. All samples were diluted to 2 × 105 molecules/μl and pooled. For emulsion PCR, 72 μl of the pooled sample was added to a reaction mix plus 600,000 capture beads per reaction; this ratio should favor a single DNA molecule per capture bead in the emulsified reaction mixture. We recovered the resulting products by breaking the emulsions and enriching for beads containing amplified products. Approximately 750,000 enriched DNA beads from each reaction were deposited into a 70- by 75-mm picotiter plate and run on the Genome Sequencer FLX (Roche). Samples from day 21 were run at one DNA molecule per capture bead, and 48 μl of the pooled sample was added to the reaction mix. Pyrosequencing errors were corrected as described elsewhere (32). In addition, insertions and deletions (indels) found in homopolymer tracts were corrected by eliminating or adding bases of the homopolymer that matched the transmitted virus to keep the sequence codons in frame (8a).

PBMC expansion and cytotoxicity assay.

PBMCs from monkeys chronically infected with SIVmac251 were separated from EDTA-preserved whole blood by Ficoll density gradient centrifugation. PBMCs were incubated at 37°C in a 5% CO2 environment in the presence of RPMI 1640-10% FCS and 1 μg/ml of WT p199RY peptide for 10 days, receiving fresh interleukin-2 (IL-2) (Hoffman-LaRoche) at 20 U/ml on days 3, 5, 7, and 9 of culture. Live cells were separated from dead cells and debris by Ficoll density gradient centrifugation, and CD8+ T lymphocytes were enriched using the MACS nonhuman primate CD8+ T cell isolation kit (Miltenyi Biotec GmbH) according to the manufacturer's directions. The frequency of p199RY-specific CD8+ T lymphocytes in this CD8 T lymphocyte-enriched population was determined as described below. These p199RY-specific CD8+ T lymphocytes were assessed as effector cells in a standard 4-h 51Cr release CTL assay. Target cells were Mamu-A*02 transfectants of the MHC class I-deficient human cell line 721.221 that had been incubated overnight with WT, T2S, or Y9F p199RY epitope peptides at various concentrations, as well as with 100 μCi of sodium [51Cr]chromate. p199RY-specific CD8+ T lymphocyte effector cells were incubated with target cells at various WT p199RY-specific CD8+ effector-to-target ratios and assessed for epitope-specific lysis as described previously (25).

Determination of p199RY-specific effector cell number for chromium release assays.

Cells in the MACS CD8 T lymphocyte-enriched population were quantified using the Guava EasyCyte Plus instrument (Millipore). An aliquot of approximately 2.5 × 105 lymphocytes was stained with PE-conjugated WT p199RY-Mamu-A*02 tetramer for 15 min at room temperature, followed by anti-CD8α-FITC (SK1; BD Biosciences) and anti-CD3-APC (in-house production of SP34 custom conjugated by BD Biosciences) for 15 min at room temperature. Cells were washed, fixed, and analyzed by flow cytometry as described previously. The concentration of WT p199RY-specific CD8 T lymphocyte-enriched lymphocytes was calculated by multiplying the Guava EasyCyte lymphocyte count by the percentage of p199RY-specific CD8+ T lymphocytes determined by flow cytometry.

Statistical analyses of selection within the epitope.

The very large sample sizes that resulted from 454 pyrosequencing were computationally prohibitive for most statistical procedures that test for positive selection, so we used a simple procedure called SNAP (18, 23). SNAP was implemented as a stand-alone routine available from the HIV database at Los Alamos National Laboratory (http://www.hiv.lanl.gov). This method corrects for multiple mutational pathways possible in a codon. We divided sequences from day 21 samples into two groups to represent regions inside and outside the epitope. We computed the distance-corrected synonymous and nonsynonymous substitution rates (dS and dN, respectively) with SNAP. When dS is greater than dN, we see support for negative selection; conversely, a dS value less than the dN value indicates positive selection. We summarized SNAP results as two-by-two contingency tables per animal where columns indicate dS > dN or dS < dN and rows indicate sites inside or outside the epitope, populated the table with counts for each sequence relative to the transmitted sequence (excluding cases where dS = dN), and used the one-sided Fisher exact test function in R (version 2.11.0; http://www.R-project.org) to evaluate whether the epitope is enriched for nonsynonymous substitutions.


Mutations in the SIV Nef p199RY epitope were positively selected as early as day 21 of SIVmac251 infection.

We and others have previously reported that p199RY epitope mutations accumulate in the viral population in Mamu-A*02+ rhesus monkeys during infection with SIVmac239 and SIVmac251 (24, 33). To define further the sequence evolution patterns in the epitope region during the course of SIV infection, we initially investigated the evolution of the Mamu-A*02-restricted SIV Nef p199RY epitope by sequencing plasma viral RNA in a cohort of 5 Mamu-A*02+ rhesus monkeys infected with SIVmac251. Plasma was sampled from each animal at four time points following infection: days 21, 35, 84, and 399 (Fig. 1A, B, C, and D, respectively). Because a day 399 specimen from monkey AJ82 was not available, plasma viral RNA sampled on day 217 from this animal was sequenced. The region spanning the p199RY epitope was reverse transcribed, amplified by PCR, cloned into a vector, and transformed into Escherichia coli. Approximately 10 to 12 clones were sequenced for each animal at each time point and compared to 12 sequenced clones of the SIVmac251 challenge stock, with the predominant sequence being defined as wild type (WT).

Fig. 1.
Sanger sequencing of the immunodominant, Mamu-A*02-restricted Nef epitope p199RY in SIVmac251-infected Mamu-A*02+ and Mamu-A*02 monkeys. (A to D) Five Mamu-A*02+ rhesus monkeys were infected with SIVmac251, and viral RNA was extracted from frozen ...

Given that this traditional sequencing strategy relied on randomly selecting only 10 to 12 clones for epitope sequencing, it could provide only a limited view of the evolution of sequence diversity in the viruses replicating in these monkeys. To explore the viral sequence diversity in greater depth, we complemented the sequencing data generated using Sanger technology with sequence data generated using high-throughput 454 pyrosequencing technology. Using 454 technology, we obtained an average of 55,500 sequences for each sample from each monkey (median, 48,719; range, 24,870 to 110,200).

Pyrosequencing of samples obtained at 21 days postinfection (dpi) indicated that each infection was heterogeneous, as expected following an i.v. infection with an uncloned SIVmac251 quasispecies. There was a predominance of WT epitope sequences (Fig. 2), although rare variants that were undetected by Sanger sequencing were observed (Fig. 1A). Each day 21 sample was dominated by greater than 99.8% of the transmitted epitope sequence (Fig. 2; see Fig. S1 in the supplemental material). Further, rare epitope variants found by Sanger sequencing in later samples were detected by 454 pyrosequencing at 21 dpi. Two variants, a position 3 serine-to-leucine mutation (S3L) and a position 7 isoleucine-to-threonine mutation (I7T), were found in all five monkeys, although neither was present at a frequency of greater than 0.04% (Fig. 2). The rare variants were present in multiple copies and were distributed over the sequencing plate, indicating that these reads indeed reflected different molecules. Two additional variants, a position 2 threonine-to-serine substitution (T2S) and a position 9 tyrosine-to-phenylalanine substitution (Y9F), which were undetected before day 84 by Sanger sequencing, were each detected in one animal on day 21 following infection (Fig. 2). Peptides bearing each of these position 2, 3, 7, and 9 p199RY epitope mutations demonstrated binding to Mamu-A*02 that was equal to the affinity of the wild-type peptide binding to Mamu-A*02 (reference 33 and data not shown).

Fig. 2.
454 sequencing of the p199RY epitope in SIVmac251-infected Mamu-A*02+ rhesus monkeys. The absolute counts (n) and frequencies (%) of unique p199RY sequences, each denoted by a different color, detected by 454 sequencing are compared to the absolute counts ...

Inferential tests supported the possibility that rare epitope variants at 21 dpi were a consequence of immune selection. In four of five animals, positive selection tests indicated that nonsynonymous mutations were enriched within the epitope region relative to the region outside the epitope region (P < 0.05) (Table 1), suggesting that immune selection had begun in p199RY by day 21 following infection. Surprisingly, in monkey CA53, the pattern was reversed, and sites outside the epitope were enriched for nonsynonymous substitutions (P = 0.04 by the one-sided Fisher exact test). This finding might be accounted for by another epitope in the Nef region, perhaps restricted by another MHC class I allele in this monkey, that was under selective pressure.

Table 1.
Nonsynonymous substitutions are enriched in the p199RY epitope at 21 dpi in four monkeys

A number of mutations flanking the p199RY epitope region were also detected in viruses from all five monkeys on day 21 postinfection (see Fig. S2 in the supplemental material). Since mutations in the flanking residues may alter epitope processing, the three amino acid residues immediately upstream and downstream of the epitope region were analyzed to determine the relative contribution of flanking mutations to the viral sequence diversity of the p199RY epitope region of the Nef protein. Positive selection tests indicated that nonsynonymous mutations were enriched within these flanking regions relative to the region outside the epitope and flanking regions in the viruses from all five monkeys (P < 0.05) (data not shown), consistent with immune selection for mutations that may affect epitope processing as early as day 21 following infection.

Both Sanger and 454 sequencing technologies detected a nearly complete replacement of the WT p199RY epitope in SIVmac251-infected monkeys between days 21 and 84 following infection.

By 35 dpi, Sanger sequencing showed that epitope sequence mutations had emerged in all monkeys, with the exception of monkey BR32 (Fig. 1B). The most frequent substitutions detected were the S3L mutation, found in two monkeys, and the I7T mutation, found in three monkeys. Despite the emergence of these p199RY epitope mutations, the wild-type sequence remained predominant in the viral population in monkeys AJ82, BR32, and CA53.

Pyrosequencing indicated that the epitope sequence of the transmitted virus was the predominant form of the virus at 35 dpi in all animals except BE86 (Fig. 2; see Fig. S1 in the supplemental material). Greater than 99% of sequences from 35 dpi in three animals (AJ82, BR32, and CA53) were of the epitope of the transmitted virus, consistent with the Sanger sequencing results. However, 454 sequencing also detected multiple epitope variants. Epitope mutations present at greater than a 1% frequency were observed at 35 dpi in monkeys BE86 and BH25, consistent with the Sanger sequencing data that demonstrated significant p199RY epitope sequence variation in the plasma viruses of these two monkeys. The S3L and I7T mutations were detected at a greater frequency by 35 dpi in all five animals. Frequencies of S3L ranged from 0.02% in BR32 to 74% in BE86. A double mutant form of the epitope, containing both S3L and I7T mutations, appeared in three of the five animals. The T2S and Y9F forms appeared together in two animals (BE86 and BH25) and separately in two animals (BR32 and CA53); both, however, were rare, being present at less than a 1% frequency at this time following infection.

By day 84 after infection, the transmitted form of the epitope was greatly diminished in the virus populations in this cohort of monkeys, with the epitope sequence of the transmitted virus detected by Sanger sequencing in only two animals, AJ82 and BH25 (Fig. 1C). The S3L mutation was detected in all five monkeys. Two more mutations, T2S in monkeys BE86, BH25, and BR32 and Y9F in monkey BH25, appeared at this time. Interestingly, we no longer detected the previously observed I7T mutation in the virus population. However, a new position 7 isoleucine-to-methionine mutation (I7M) was detected in monkey BH25.

In contrast to the data generated using Sanger sequencing, 454 pyrosequencing at 84 dpi detected the WT transmitted p199RY epitope in all five monkeys (median frequency, 2%; range, 0.05% to 44%) (Fig. 2). The 454-generated sequences indicated extensive epitope evolution in all five monkeys, with a substantial proportion of WT p199RY epitope sequence remaining in the viral population only in monkey AJ82. These findings were consistent with the Sanger sequencing data (Fig. 1C). The most common epitope mutation detected by 454 pyrosequencing in the plasma samples from all five monkeys on 84 dpi was S3L, a mutation detected in the viruses from all five monkeys by Sanger sequencing (Fig. 1C). The 454 technology also detected additional mutations, such as isoleucine-to-alanine and isoleucine-to-asparagine mutations at position 7 of the epitope, that were not observed by conventional methods. The detection of additional epitope mutations, as well as the observation of mutations at frequencies different from those determined using Sanger sequencing methods, was likely a consequence of the much greater sampling (by 3.4 to 4 orders of magnitude) of viral clones for epitope sequencing using the 454 pyrosequencing technology.

Strikingly, ultradeep sequencing demonstrated that the selection pressure exerted on SIVmac251 by the p199RY-specific CD8+ T lymphocytes resulted in a near-total replacement of the WT p199RY epitope in these monkeys (Fig. 2; see Fig. S1 in the supplemental material). Nonetheless, the WT epitope sequence persisted in the virus populations in these monkeys at low frequencies at least through day 84 following infection. The epitope sequence of the transmitted virus was first replaced by mutated epitope sequences (S3L and I7T) that were largely undetected by Sanger sequencing. These mutant forms were then replaced by a wave of different mutated epitope sequences (T2S and Y9F).

Beyond 200 days of infection with SIVmac251, the wild-type p199RY epitope sequence was no longer detected in the cohort by Sanger sequencing (Fig. 1D). Two of the epitope mutations detected earlier following infection, S3L and Y9F, predominated in the viral populations in these animals, with an additional epitope variant, S3F, detected in monkey AJ82. These data indicate that the S3L and Y9F p199RY epitope mutations became fixed in the viral population in the SIVmac251-infected Mamu-A*02+ rhesus monkeys.

Epitope frequencies determined by Sanger and 454 sequencing technologies identified similar common epitope sequences in each evaluated sample (Fig. 3). The abundance of each unique epitope sequence per milliliter of plasma was calculated by multiplying the viral load reported by Newberg et al. (24) by the frequency of the epitope sequence observed. A binomial model of sampling uncertainty for epitope frequencies by Sanger sequencing gave confidence intervals that overlapped the epitope frequencies determined by 454 sequencing. The binomial model quantifies detection limits from a given number of samples and degree of confidence. Sampling 12 sequences (as with Sanger sequencing) can detect variants present with at least 22% frequency, while 50,000 reads (as with 454 pyrosequencing) can detect variants represented by at least 0.006% frequency, both with 95% confidence (Fig. 3).

Fig. 3.
Dynamics of p199RY epitope sequence evolution as detected by Sanger and 454 sequencing. Sanger (left) and 454 (right) sequencing results are indicated. Epitope frequencies, in copies per milliliter of plasma, are scaled to the viral RNA load in the sample ...

Finally, to determine if p199RY epitope sequence variation was restricted to infected monkeys that express the MHC class I Mamu-A*02 allele, we used Sanger technology to sequence plasma viral RNA sampled late in chronic SIVmac251 infection in two monkeys that did not express Mamu-A*02 (Fig. 1E). In both animals, no sequence evolution was detected in the p199RY epitope region. Therefore, sequence evolution in the p199RY epitope was selected only in the presence of Mamu-A*02-restricted p199RY epitope-specific CD8+ T cell responses during SIV replication.

p199RY epitope-specific CD8+ T lymphocytes from SIVmac251-infected Mamu-A*02+ monkeys were capable of cross-recognizing the WT and mutant forms of the p199RY epitope.

We hypothesized that the p199RY epitope sequence evolution observed in Mamu-A*02+ rhesus monkeys was a consequence of evolving immune selection pressure first exerted on the virus by CD8+ T lymphocytes that were primed by the inoculated WT virus during acute infection and subsequently exerted by emergent CD8+ T lymphocyte populations that were primed de novo by viruses harboring emerging epitope mutations later during infection. To test this hypothesis, we examined the coevolution of the p199RY epitope sequences and the p199RY epitope-specific CD8+ T lymphocyte populations in SIVmac251-infected rhesus monkeys by tetramer staining of peripheral blood mononuclear cells (PBMCs) sampled at three time points following infection with SIVmac251. We constructed three tetramers with Mamu-A*02 and representative selected p199RY epitope variants, including the transmitted epitope (the form that dominates each infection at day 21, denoted WT), a common variant at 84 dpi (T2S), and the variant dominant at a year postinfection in two animals (Y9F). Neither of these mutations was significantly associated with mutations flanking the p199RY epitope region that may affect intracellular processing (see Fig. S2 in the supplemental material). We hypothesized that the tetramers constructed with the T2S and Y9F epitope peptides would bind to newly emergent CD8+ T lymphocyte populations generated in response to mutant viruses expressing the T2S and Y9F Nef sequences. Sampled PBMCs were stained with a cocktail of all three tetramers, each complexed with a different fluorophore, and gated on CD3+ CD8+ T lymphocytes. On day 21 following infection, most CD8+ T lymphocytes stained with the WT tetramer only. Interestingly, however, a subset of CD8+ T lymphocytes that stained with the WT tetramer also stained with the T2S tetramer and the Y9F tetramer (Fig. 4A, left and center panels, respectively). Moreover, the majority of cells that stained with the T2S tetramer also stained with the Y9F tetramer (Fig. 4A, right panel).

Fig. 4.
CD8+ T lymphocytes specific for the WT Nef p199RY epitope cross-recognized mutant p199RY epitope sequences. PBMCs from 5 Mamu-A*02+ SIVmac251-infected monkeys were stained with a cocktail of all three tetramers to evaluate the cross-reactivity of p199RY-specific ...

Cells that cross-stained with the WT tetramer and the T2S and/or Y9F tetramer were maintained in the p199RY epitope-specific CD8+ T cell populations in these monkeys on day 84 (Fig. 4B, left and center panels) and on day 196 (Fig. 4C, left and center panels) following SIV infection. As seen on day 21 of infection, the majority of cells staining with the T2S tetramer also stained with the Y9F tetramer (Fig. 4B and C, right panels). Notably, the subset of cells that stained with WT and T2S/Y9F tetramers became enriched in the epitope-specific CD8+ T cell populations at these time points.

The apparent cross-staining of cells with WT, T2S, and Y9F tetramers raised the possibility that the tetramers were staining CD8+ T lymphocytes nonspecifically, possibly due to epitope-independent binding of the Mamu-A*02 protein used in constructing the tetramers. To assess this possibility, we constructed a tetramer with the Mamu-A*02 protein and another Mamu-A*02-binding epitope peptide, SIV Nef p56 (YTYEAYVRY). PBMCs sampled from three SIVmac251-infected Mamu-A*02+ rhesus monkeys were stained with a cocktail of the WT p199RY tetramer and the p56 tetramer, each complexed with different fluorophores (Fig. 4D). Epitope-specific cells from each of the three monkeys stained with either the p199RY or the p56 tetramer, with no cross-staining detected. Taken together, these data suggest that a subset of p199RY-specific CD8+ T lymphocytes had the potential of cross-recognizing the T2S and/or the Y9F variant of the epitope from the earliest time point of sampling and that this apparent cross-recognition was not a consequence of nonspecific binding of Mamu-A*02 tetramers to CD8+ T lymphocytes. Therefore, the p199RY epitope sequence evolved in the presence of a CD8+ T lymphocyte population that recognized the mutant forms of the epitope. This CD8+ T lymphocyte population was already present in the monkeys before viral mutations began to accrue.

Evolution of p199RY-specific CD8+ T lymphocytes in SIVmac251-infected Mamu-A*02+ monkeys.

We next performed a quantitative analysis to determine the percentage of p199RY-specific CD8+ T lymphocytes that stained with each combination of one, two, and all three of the WT, T2S, and Y9F p199RY peptide tetramers. CD3+ CD8+ T lymphocytes were initially gated on cells that stained positively or negatively with the WT p199RY tetramer, and each gated population was analyzed to determine which cells also stained with the T2S and/or the Y9F p199RY tetramer (Fig. 5A). The percentage of cells that stained with each combination of one, two, or all three tetramers was calculated.

Fig. 5.
p199RY-specific CD8+ T lymphocytes cross-recognized mutant forms of the epitope prior to viral evolution. (A) The gating strategy used to enumerate cells that stained with one, two, or all three tetramers is shown for a representative monkey, BR32, on ...

The majority of cells sampled from each of the five monkeys at 21 dpi stained with the WT p199RY tetramer only (Fig. 5B, maroon wedges), consistent with the detection of the vast majority of p199RY viral RNA sequences being the transmitted form at this time point. Interestingly, however, at least 23% of tetramer+ CD8+ T lymphocytes stained with various combinations of tetramers constructed with T2S and Y9F p199RY epitope peptides in four of the five monkeys. In these same four animals, at least 15% of tetramer-binding CD8+ T lymphocytes stained with all three tetramers (Fig. 5B, yellow wedges).

Cell populations isolated from monkeys at 84 dpi, after sequence variation in the p199RY epitope was detected by Sanger sequencing in all five monkeys, exhibited a shift in the tetramer-binding specificity relative to that observed on day 21 following infection. The proportion of all tetramer-positive lymphocytes that stained with only the WT p199RY tetramer diminished, while the cells binding to all three tetramers became enriched in the p199RY-specific CD8+ T lymphocyte population. This staining pattern was also observed in cells sampled at 196 dpi. These tetramer staining data suggest that a significant proportion of p199RY-specific CD8+ T lymphocytes are capable of recognizing multiple sequence variants of the p199RY epitope before such mutations are abundant in the viral population and that these polyspecific cells appear to be responsive to the evolving virus and become enriched in the epitope-specific CD8+ T cell population.

p199RY epitope-specific CD8+ T lymphocytes isolated from Mamu-A*02+ SIVmac251-infected rhesus monkeys lysed with comparable efficiency Mamu-A*02-expressing target cells pulsed with the WT and T2S p199RY epitope peptides.

While staining CD8+ T lymphocytes from Mamu-A*02+ SIVmac251-infected monkeys with WT, T2S, and Y9F p199RY tetramers indicates that p199RY-specific cells had the potential for recognizing multiple sequence variants of the epitope, these data did not directly show that the cells had comparable functional activity upon T cell receptor (TCR) engagement of the variant p199RY peptide-Mamu-A*02 epitope complexes. Indeed, the emergence and fixation of mutant epitope sequences in the presence of cross-reactive CD8+ T lymphocyte populations may have been a consequence of the inability of the mutant epitopes to stimulate the function of the CD8+ T lymphocytes. To evaluate p199RY-specific CD8+ T lymphocyte function upon recognition of WT and mutant p199RY ligands, we tested the cytotoxic capacity of p199RY-specific CD8+ T lymphocytes upon interaction with WT, T2S, and Y9F peptide-pulsed target cells. PBMCs collected from four Mamu-A*02+ rhesus monkeys chronically infected with SIVmac251 were expanded in the presence of WT p199RY epitope peptide for 10 days. These effector cells were then cocultured for 4 h with a Mamu-A*02-expressing cell line pulsed with serial log dilutions of WT, T2S, and Y9F p199RY epitope peptides. p199RY-stimulated CD8+ T lymphocytes from all four monkeys were capable of lysing target cells pulsed with the WT, T2S, and Y9F epitope peptides (Fig. 6). At high concentrations of peptide, comparable lysis of WT, T2S, and Y9F peptide-pulsed target cells by PBMCs from all four monkeys was observed. At limiting peptide concentrations, however, T2S-specific lysis was only modestly less efficient in PBMCs of three of the four monkeys, while Y9F-specific lysis was more inefficient than WT epitope-specific lysis. These data show that effector cells primed with the WT p199RY epitope were capable of cross-recognizing and lysing cells that present mutant sequences of the epitope, consistent with the demonstration of CD8+ T lymphocytes that stain with multiple p199RY variant peptide-MHC class I tetramers. Lysis by PBMCs isolated from SIVmac251-infected monkeys was modestly less for targets pulsed with the two prototype mutant epitope peptides. The same was observed in experiments in which the PBMCs isolated from infected monkeys were expanded in the presence of either the WT, the T2S, or the Y9F epitope peptide and then cocultured with target cells pulsed with the peptide matching that used for the expansion of effector cells (data not shown). This modestly lower efficiency might account for successful escape of mutant viruses from CD8+ T lymphocyte-mediated elimination. Importantly, however, very little expansion of CD8+ T lymphocyte populations that preferentially recognize only mutant epitope sequences was observed in the SIVmac251-infected, Mamu-A*02+ monkeys (Fig. 5B). The inability of the immune system to generate new populations of CD8+ T lymphocytes that preferentially recognize mutant epitope sequences may contribute to the success of this virus in evading immune control.

Fig. 6.
p199RY-specific CD8+ T lymphocytes killed Mamu-A*02+ target cells pulsed with different p199RY epitope peptides. WT p199RY epitope peptide-stimulated effector cells were cocultured at the indicated tetramer-positive effector-to-target ratios with a Mamu-A*02 ...


The conventional viral RNA sequencing method initially used in this study had two disadvantages for assessing virus epitope sequence variation during the course of SIVmac251 infections. In order to preamplify viral RNA templates for sequencing, the RNA was initially subjected to reverse transcription-PCR (RT-PCR) under conditions where multiple, nonidentical viral clones were amplified in a single bulk reaction. This introduced the possibility that some p199RY sequences, such as the S3L I7M double mutation detected in monkey BH25 on day 84 of infection, may have been a consequence of RT-PCR-mediated recombination rather than immune selection pressure. While it is highly unlikely to occur within a 15-nucleotide region spanning positions 3 and 7 of the p199RY epitope, this phenomenon has been observed in similar settings (15) and therefore cannot be ruled out.

The other major caveat associated with conventional viral RNA sequencing in this setting is the limited numbers of viral clones that can be subjected to sequencing. Indeed, only 9 to 12 clones were selected for sequencing from each monkey at each time point in this study. The probability of seeing a fraction f in a sample of size n is 1 (1 − f)n (17). For example, with a sample size of 10, there is a 95% probability of detecting mutations present at a frequency of 0.26; however, less common sequences would likely be missed. Therefore, most p199RY mutations present at low frequencies in the viral population will not be represented in sequence data based on conventional sequencing. One could simply sequence many more clones to facilitate detection of rare epitope mutations. However, detection of rare epitope variants present at frequencies of less than 1% in the viral population could require the sequencing of hundreds or thousands of cDNA clones, which would be costly and labor-intensive. Large numbers of viral RNA clones have also been sampled using WT and mutant sequence-specific probes in quantitative RT-PCRs to detect epitope escape mutations present at low frequencies in the viral population (20). This technique, however, requires prior knowledge of how an epitope evolves over the course of an infection, and it can be used to monitor the emergence of only a limited number of mutations.

The 454 pyrosequencing technology allows sequencing of 25,000 to 100,000 viral RNA clones from individual, preamplified viral cDNA samples. Increased sampling by 3 to 4 orders of magnitude enabled us to detect epitope mutations that were present in the viral population at frequencies of less than 0.01% and represented by multiple copies. Moreover, the preamplification of viral cDNA prior to 454 pyrosequencing was carried out under conditions favoring the amplification of single cDNA clones in individual reactions (32), which minimized the possibility of RT-PCR-mediated recombination of mixed, nonidentical viral clones. The presence of S3L and I7T epitope mutations from day 21 and their increased frequencies in later samples indicate an early wave of mutations undetected by Sanger sequencing. Furthermore, using these single-clone preamplification conditions for 454 pyrosequencing, we detected p199RY variants with more than one mutation. The S3L I7M double mutation was detected at low frequencies in monkeys BE86, BH25, and BR32 using 454 pyrosequencing, while it was found in only monkey BH25 using Sanger sequencing. Another double mutant, S3L I7T, was detected in four monkeys using 454 pyrosequencing, while it was never detected using conventional sequencing methods. Interestingly, this mutation was represented in 6% of the viral clones in monkey BH25 on day 84 following infection. Finally, despite the displacement of the original epitope form by mutated variants, the original form persisted through 84 days after infection, consistent with other findings from ultradeep sequencing and suggesting that there are reservoirs established by the earliest founder virus (21). These findings highlight the limitations of conventional sequencing to detect epitope mutations that are present in a significant fraction in a viral population.

The application of 454 pyrosequencing allowed us to determine definitively the extent of evolution of the virus during the course of SIVmac251 infection. Though over 99.8% of the more than 65,000 viral clones sequenced in all five monkeys at 21 dpi represented the WT p199RY sequence, we found evidence for immune selection in the epitope region in four of five animals at this time. The T2S and Y9F mutations were present at extremely low levels (one or three copies sampled) and in single infected animals at 21 dpi. Therefore, nearly all the p199RY-specific CD8+ T lymphocytes that were capable of cross-recognizing the T2S and Y9F epitope mutations, including those detected on day 21 following infection, were most likely being primed by the WT epitope sequence. Interestingly, we detected viruses with T2S and/or Y9F epitope mutations in over 5% of the viral population on day 84 after infection in four of the five monkeys, at a time when the majority of p199RY-specific CD8+ T cells present were capable of recognizing these mutations. This finding indicates that the polyspecific CD8+ T lymphocytes in these monkeys failed to contain the replication of the evolving mutant viruses.

The relative timing of the emergence of p199RY epitope mutations observed in this cohort of SIVmac251-infected Mamu-A*02+ monkeys corresponded with known processes of molecular evolution (16). In general, the S3L and I7T mutations were detected earlier and at higher frequencies than the T2S and Y9F mutations. The S3L and I7T mutations resulted from single pyrimidine transitions in the second codon position (TCA → TTA for S3L and ATT → ACT for I7T), while both the T2S and Y9F mutations resulted from transversions (ACC → AGC for T2S and TAC → TTC for Y9F). Given that transitions occur more frequently than transversions, it is possible that the I7T and S3L mutations were repeatedly selected first in the viral population because they were present at a higher frequency prior to immune selection. The T2S and Y9F mutations would then be favored later because of a stronger selective advantage under pressure from the immune response.

Numerous reports indicate that mutations in the regions of a virus immediately flanking epitope sequences are also selected during infection, and these flanking mutations impart a selective advantage for viruses harboring them by disrupting intracellular proteasomal processing and presentation of these antigens on the surface of infected cells (8, 26). We therefore examined our 454 pyrosequencing data to determine the amino acid sequence diversity in the three residue positions immediately upstream and downstream of the p199RY epitope. Not surprisingly, a significant enrichment of flanking mutations was detected as early as week 21 following infection. However, there appeared to be no significant association between the appearance of altered residues in these flanking regions and the presence of the T2S and Y9F epitope mutations. Moreover, our laboratory and others have shown that the T2S and Y9F mutations do not disrupt binding to Mamu-A*02 (33), indicating that the mutant epitope peptides, once processed, are capable of being stably presented on the surface of SIV-infected cells.

A number of studies have demonstrated that epitope-specific CD8+ T lymphocyte responses coevolve in infected individuals with emerging viruses displaying altered epitope sequences. As viruses with mutated epitopes can come to outnumber those with WT epitope sequences, a decline in the WT-specific CD8+ T lymphocyte responses is often observed (35). Some groups have reported that emerging viruses with epitope mutations can prime de novo CD8+ T lymphocyte populations capable of recognizing the emerging mutant viruses (3, 11). Moreover, these mutant-specific, de novo-primed cells do not cross-recognize the WT epitope sequence (3). Given these observations, we hypothesized that the emergence of the T2S p199RY mutation by day 84 and its disappearance in favor of the fixation of the Y9F p199RY mutation by day 399 following infection in the SIVmac251-infected Mamu-A*02+ rhesus monkeys would be associated with a coevolution of the p199RY-specific CD8+ T lymphocyte response. For example, the initial WT p199RY epitope-specific CTL response would favor the selection of the T2S mutation, which in turn would prime a de novo T2S epitope-specific response that would exert pressure on the virus to mutate further to the Y9F epitope sequence. Our findings, however, did not support this hypothesis. Rather, a significant proportion of the WT p199RY-specific CD8+ T cell population cross-recognized both the T2S and Y9F epitope mutants as early as day 21 following infection, before these mutations were even detected in appreciable numbers of viruses in the infected monkeys by 454 pyrosequencing. Furthermore, these polyspecific lymphocytes became enriched in p199RY-specific CD8+ T lymphocyte populations following the emergence and fixation of these virus epitope mutations, with an associated loss of CD8+ T lymphocytes that recognize only the WT p199RY sequence. These observations suggest that epitope evolution did not prime mutant epitope-specific CD8+ T lymphocytes de novo but rather continued to stimulate cross-reactive cell populations that were initially primed by the WT virus.

It is puzzling that p199RY-specific CD8+ T lymphocyte populations able to recognize epitope variants failed to eliminate mutated viruses expressing those variant epitopes from the virus population. While staining CD8+ T lymphocytes with MHC class I-peptide tetramers is a useful tool for evaluating the specificity of epitope-specific cells, the technique does not assess the functional consequences of TCR engagement of epitope-specific cells with epitope peptide-MHC class I complexes. Therefore, the possibility remains that the polyspecific CD8+ T lymphocytes detected by tetramer staining may respond differently upon engagement with MHC class I expressing the WT viral epitope peptide and MHC class I expressing epitope peptides with the T2S and Y9F mutations that emerged during infection. It has been demonstrated that the avidity of the epitope peptide-MHC class I complex for the T cell receptor is correlated with the potency of CD8+ T lymphocyte activation (7). Furthermore, studies have shown that epitope peptides with mutated amino acid residues can induce quantitatively and qualitatively different CD8+ T lymphocyte responses (13, 30). Therefore, the T2S and Y9F epitope mutations may have persisted in the population of the SIVmac251-infected Mamu-A*02+ rhesus monkeys because they had lower avidities than the WT epitope peptide for the TCRs of polyspecific CD8+ T lymphocytes. The cytotoxicity assays performed in the present study support this hypothesis, as peptides representing the T2S and Y9F epitope mutations sensitized targets for p199RY-specific CD8+ T lymphocyte-mediated killing less efficiently than the WT p199RY epitope peptide. While the magnitude of the difference in killing of target cells expressing the WT and mutant epitope peptides was modest, the difference may have been enough to impart a competitive advantage for viruses with the T2S and Y9F epitope mutations to continue to replicate in the viral population. The lower frequencies of the cells that recognized the mutant forms early in infection also may affect the dynamics and extent of the immunological pressure on different forms of the virus through the course of early infection.

Studies indicate that effective CD8+ T lymphocyte responses are dependent on CD4+ T lymphocyte help and, more specifically, the secretion of IL-2 by CD4+ T lymphocytes. Many groups have demonstrated that functional CD4+ T lymphocytes are critical during the priming of naïve antigen-specific CD8+ T lymphocytes that become memory cells that can mount an effective recall response upon reexposure to antigen (28, 31). It is possible that damage to the CD4+ T lymphocytes as a result of SIV infection may have affected the priming of mutant-specific CD8+ T lymphocytes. This may account for the failure of de novo priming of T2S- and Y9F-specific CD8+ T lymphocytes as the mutant viruses emerge in the viral quasispecies. The impaired maintenance of functional memory CD8+ T cells, which is also dependent on CD4+ T cell help (29), may account for the defect in functional activity against these mutant viruses by cells that are capable of cross-recognizing the WT and mutant epitopes.

While our functional analysis of peptide-stimulated p199RY-specific CD8+ T lymphocytes focused on the cytotoxic capacity of these cells, other antiviral functions dependent upon epitope stimulation could also be affected by mutations in the epitope. Indeed, it has been shown that effector CD8+ T cells can be subdivided into subsets of lytic effector CD8+ T cells and effector CD8+ T cells that produce gamma interferon (IFN-γ), tumor necrosis factor alpha (TNF-α), and interleukin-2 (IL-2) (1). Furthermore, memory CD8+ T lymphocytes may proliferate to various extents following stimulation with WT or mutant epitope ligands. These functional responses by different subsets of effector and memory p199RY-specific CD8+ T cell populations remain to be evaluated.

These studies demonstrate an unexpected coevolution of the virus and an immunodominant epitope-specific CD8+ T lymphocyte response in SIVmac251-infected rhesus monkeys. While it was expected that the virus would evolve at the p199RY epitope to evade recognition by the population of CD8+ T lymphocytes primed by the WT virus, we found that these WT-primed p199RY-specific cells were capable of cross-recognizing mutant epitope sequences. These polyspecific CD8+ T lymphocyte populations were present in significant numbers before viral epitope evolution was clearly present, and later these polyspecific cells became enriched in the p199RY epitope-specific CD8+ T lymphocyte population, as the mutated forms became more common. This suggests that the CD8+ T lymphocytes responding to the evolving p199RY epitope were cells present at a relatively high frequency in the infected monkeys even before the mutations in the virus began to accrue.

Supplementary Material

[Supplemental material]


We thank Vitaly Ganusov, Ruy Ribeiro, and Alan Perelson for helpful discussions.

This work was supported by NIAID Center for HIV/AIDS Vaccine Immunology grant AI067854 and the LANL Laboratory Directed Research and Development.

We declare that we have no competing interests.


Supplemental material for this article may be found at http://jvi.asm.org/.

[down-pointing small open triangle]Published ahead of print on 9 February 2011.


1. Aandahl E. M., et al. 2003. CD7 is a differentiation marker that identifies multiple CD8 T cell effector subsets. J. Immunol. 170:2349–2355. [PubMed]
2. Allen T. M., et al. 2000. Tat-specific cytotoxic T lymphocytes select for SIV escape variants during resolution of primary viraemia. Nature 407:386–390. [PubMed]
3. Allen T. M., et al. 2005. De novo generation of escape variant-specific CD8+ T-cell responses following cytotoxic T-lymphocyte escape in chronic human immunodeficiency virus type 1 infection. J. Virol. 79:12952–12960. [PMC free article] [PubMed]
4. Barouch D. H., et al. 2002. Eventual AIDS vaccine failure in a rhesus monkey by viral escape from cytotoxic T lymphocytes. Nature 415:335–339. [PubMed]
5. Barouch D. H., et al. 2005. Dynamic immune responses maintain cytotoxic T lymphocyte epitope mutations in transmitted simian immunodeficiency virus variants. Nat. Immunol. 6:247–252. [PubMed]
6. Bimber B. N., et al. 2009. Ultradeep pyrosequencing detects complex patterns of CD8+ T-lymphocyte escape in simian immunodeficiency virus-infected macaques. J. Virol. 83:8247–8253. [PMC free article] [PubMed]
7. Davis M. M., et al. 1998. Ligand recognition by alpha beta T cell receptors. Annu. Rev. Immunol. 16:523–544. [PubMed]
8. Del Val M., Schlicht H. J., Ruppert T., Reddehase M. J., Koszinowski U. H. 1991. Efficient processing of an antigenic sequence for presentation by MHC class I molecules depends on its neighboring residues in the protein. Cell 66:1145–1153. [PubMed]
8a. Fischer W., et al. 2010. Transmission of single HIV-1 genomes and dynamics of early immune escape revealed by ultra-deep sequencing. PLoS One 5:e12303. [PMC free article] [PubMed]
9. Goulder P. J., et al. 2001. Evolution and transmission of stable CTL escape mutations in HIV infection. Nature 412:334–338. [PubMed]
10. Goulder P. J., et al. 1997. Late escape from an immunodominant cytotoxic T-lymphocyte response associated with progression to AIDS. Nat. Med. 3:212–217. [PubMed]
11. Haas G., et al. 1996. Dynamics of viral variants in HIV-1 Nef and specific cytotoxic T lymphocytes in vivo. J. Immunol. 157:4212–4221. [PubMed]
12. Hughes A. L., et al. 2010. Dynamics of haplotype frequency change in a CD8+TL epitope of simian immunodeficiency virus. Infect. Genet. Evol. 10:555–560. [PMC free article] [PubMed]
13. Jameson S. C., Bevan M. J. 1995. T cell receptor antagonists and partial agonists. Immunity 2:1–11. [PubMed]
14. Jin X., et al. 1999. Dramatic rise in plasma viremia after CD8(+) T cell depletion in simian immunodeficiency virus-infected macaques. J. Exp. Med. 189:991–998. [PMC free article] [PubMed]
15. Judo M. S., Wedel A. B., Wilson C. 1998. Stimulation and suppression of PCR-mediated recombination. Nucleic Acids Res. 26:1819–1825. [PMC free article] [PubMed]
16. Jukes T. H. 1987. Transitions, transversions, and the molecular evolutionary clock. J. Mol. Evol. 26:87–98. [PubMed]
17. Keele B. F., et al. 2008. Identification and characterization of transmitted and early founder virus envelopes in primary HIV-1 infection. Proc. Natl. Acad. Sci. U. S. A. 105:7552–7557. [PMC free article] [PubMed]
18. Korber B. 2000. HIV signature and sequence variation analysis, p. 55–72In Rodrigo A. G., Learn G. H. Jr., editors. (ed.), Computational analysis of HIV molecular sequences. Kluwer Academic Publishers, Dordrecht, Netherlands.
19. Lewis M. G., et al. 1994. Titration and characterization of two rhesus-derived SIVmac challenge stocks. AIDS Res. Hum. Retroviruses 10:213–220. [PubMed]
20. Loh L., Kent S. J. 2008. Quantification of simian immunodeficiency virus cytotoxic T lymphocyte escape mutant viruses. AIDS Res. Hum. Retroviruses 24:1067–1072. [PubMed]
21. Love T. M., Thurston S. W., Keefer M. C., Dewhurst S., Lee H. Y. 2010. Mathematical modeling of ultradeep sequencing data reveals that acute CD8+ T-lymphocyte responses exert strong selective pressure in simian immunodeficiency virus-infected macaques but still fail to clear founder epitope sequences. J. Virol. 84:5802–5814. [PMC free article] [PubMed]
22. Musey L., et al. 1997. Cytotoxic-T-cell responses, viral load, and disease progression in early human immunodeficiency virus type 1 infection. N. Engl. J. Med. 337:1267–1274. [PubMed]
22a. National Research Council 1996. Guide for the care and use of laboratory animals. National Academy Press, Washington, DC.
23. Nei M., Gojobori T. 1986. Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol. Biol. Evol. 3:418–426. [PubMed]
24. Newberg M. H., et al. 2006. Immunodomination in the evolution of dominant epitope-specific CD8+ T lymphocyte responses in simian immunodeficiency virus-infected rhesus monkeys. J. Immunol. 176:319–328. [PubMed]
25. Robinson S., et al. 2001. A commonly recognized simian immunodeficiency virus Nef epitope presented to cytotoxic T lymphocytes of Indian-origin rhesus monkeys by the prevalent major histocompatibility complex class I allele Mamu-A*02. J. Virol. 75:10179–10186. [PMC free article] [PubMed]
26. Rock K. L., Goldberg A. L. 1999. Degradation of cell proteins and the generation of MHC class I-presented peptides. Annu. Rev. Immunol. 17:739–779. [PubMed]
27. Schmitz J. E., et al. 1999. Control of viremia in simian immunodeficiency virus infection by CD8+ lymphocytes. Science 283:857–860. [PubMed]
28. Shedlock D. J., Shen H. 2003. Requirement for CD4 T cell help in generating functional CD8 T cell memory. Science 300:337–339. [PubMed]
29. Shin H., Wherry E. J. 2007. CD8 T cell dysfunction during chronic viral infection. Curr. Opin. Immunol. 19:408–415. [PubMed]
30. Sloan-Lancaster J., Allen P. M. 1996. Altered peptide ligand-induced partial T cell activation: molecular mechanisms and role in T cell biology. Annu. Rev. Immunol. 14:1–27. [PubMed]
31. Sun J. C., Bevan M. J. 2003. Defective CD8 T cell memory following acute infection without CD4 T cell help. Science 300:339–342. [PMC free article] [PubMed]
32. Tsibris A. M., et al. 2009. Quantitative deep sequencing reveals dynamic HIV-1 escape and large population shifts during CCR5 antagonist therapy in vivo. PLoS One 4:e5683. [PMC free article] [PubMed]
33. Vogel T. U., et al. 2002. Escape in one of two cytotoxic T-lymphocyte epitopes bound by a high-frequency major histocompatibility complex class I molecule, Mamu-A*02: a paradigm for virus evolution and persistence? J. Virol. 76:11623–11636. [PMC free article] [PubMed]

Articles from Journal of Virology are provided here courtesy of American Society for Microbiology (ASM)


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...