The preexisting HIV-1–specific T cell repertoire must influence both the immunodominance of T cells after infection and immunogenicity of vaccines. We directly compared two methods for measuring the preexisting CD4+ T cell repertoire in healthy HIV-1–negative volunteers, the HLA-peptide tetramer enrichment and T cell library technique, and show high concordance (r = 0.989). Using the library technique, we examined whether naive, central memory, and/or effector memory CD4+ T cells specific for overlapping peptides spanning the entire HIV-1 proteome were detectable in 10 HLA diverse, HIV-1–unexposed, seronegative donors. HIV-1–specific cells were detected in all donors at a mean of 55 cells/million naive cells and 38.9 and 34.1 cells/million in central and effector memory subsets. Remarkably, peptide mapping showed most epitopes recognized by naive (88%) and memory (56%) CD4+ T cells had been previously reported in natural HIV-1 infection. Furthermore, 83% of epitopes identified in preexisting memory subsets shared epitope length matches (8–12 amino acids) with human microbiome proteins, suggestive of a possible cross-reactive mechanism. These results underline the power of a proteome-wide analysis of peptide recognition by human T cells for the identification of dominant antigens and provide a baseline for optimizing HIV-1–specific helper cell responses by vaccination.
Only one candidate HIV vaccine, a canarypox vectored gp120 with a protein boost, has shown any efficacy (Rerks-Ngarm et al., 2009). The limited protection correlated with induction of nonneutralizing antibodies to the VI/V2 region of the virus Envelope protein (Env; Rerks-Ngarm et al., 2009; Haynes et al., 2012). This modest success has stimulated efforts to design vaccines that generate more efficient neutralizing antibodies, together with potent CD4+ T cell responses capable of providing help to B cells and cytotoxic T cells (Burton et al., 2012). Understanding how the magnitude and specificity of these helper T cells can be optimized will be critical to the design of an effective vaccine.
Primary immune responses are probably influenced strongly by the preexisting repertoire of B and T cells. However, characterization and quantification of these repertoires is difficult due to the extremely low number of circulating naive precursor cells (Jenkins et al., 2001; Su et al., 2013). Previous studies of naive CD4+ T cell repertoires in humans and mice have relied on magnetic beads to enrich MHC tetramer binding cells (Moon et al., 2007; Kwok et al., 2012; Su et al., 2013). However, although this approach gives precise information on responses to particular MHC-peptide epitopes, it does not measure the total repertoire and misses previously unknown epitopes. An alternative T cell library technique requires no prior knowledge of donor HLA type or epitope specificity (Geiger et al., 2009). The method presorts circulating T cells into naive and memory subsets which are seeded at limiting dilution before polyclonal expansion in the presence of PHA, allogeneic feeder cells, and IL-2. Individual cultures are then screened for proliferative responses to a protein or series of peptides representing the pathogen of interest (Geiger et al., 2009). Combined with epitope mapping and the Poisson distribution, the T cell library technique can provide quantitative data on the specificity of the entire preexisting naive and memory repertoire.
The existence of HIV-1–specific memory cells in seronegative donors was originally suggested by studies of highly exposed HIV-1 seronegative (HESN) donors. It has been shown that 25–61% of HESNs have demonstrable HIV-1–specific memory cells, probably primed by exposure to the virus. Surprisingly, HIV-1–specific CD4+ T cells were also detected in 24–44% of unexposed donors (Ritchie et al., 2011), although it was not clear whether the latter came from cross-reactive memory T cells or naive T cells primed in vitro. More recently, the existence of low frequency (1–10/million) memory CD4+ T cells, specific for a known HIV-1 Gag epitope, was demonstrated by HLA DR4 tetramers in 50% of HIV-1 unexposed HLA DR4+ adults (Su et al., 2013), but it was not clear how generalizable the HIV-1 result was beyond the single epitope–HLA DR4 combination.
The present study first validates the library technique by direct comparison with the tetramer enrichment method for measuring precursor T cell frequencies. We then use the T cell library technique to provide the first proteome-wide analysis of the frequencies and specificities of preexposure HIV-1–specific naive and memory CD4+ T cells in a HLA diverse population of HIV-1 unexposed donors.
RESULTS AND DISCUSSION
Comparison of the T cell library technique to tetramer enrichment
Before commencing a proteome-wide screen of the preexisting HIV-1–specific naive and memory CD4+ repertories, we first performed a direct comparison of the T cell library (Geiger et al., 2009) and tetramer enrichment methods (Su et al., 2013) using varicella zoster virus (VZV) as a model antigen. We established CD4+ T cell libraries of 192 wells per subset from naive (CD45RA+CCR7+), central memory (CD45RA−CCR7+), and effector memory (CD45RA−CCR7−) subsets of five anonymous HLADRB1*1501+ blood donors. Lines were polyclonally expanded for 16–20 d before screening for proliferative response to two VZV epitopes (glycoprotein E [gE] and immediate early phosphoprotein 63 [IE63]), which are known to be restricted by and commonly detected in HLADRB1*1501 donors (Jones et al., 2007; Malavige et al., 2008). In parallel 25–70 × 106 CD4+ T cells from the same time point were screened using custom-made gE and IE63 HLADRB1*1501 tetramers. We show that the two techniques are highly comparable (Pearson’s correlation, r = 0.999), with no significant difference in the precursor frequencies obtained by each technique (paired Student’s t test, P = 0.1099; Fig. 1). These data support the T cell library technique for comprehensive analysis of naive and memory CD4+ T cell repertoires.
HIV-1–specific T cells in HIV-1–unexposed, seronegative donors
Naive, central memory, and effector memory CD4+ T cell subsets from 10 HLA diverse (Table S1), healthy, HIV-1 seronegative donors were screened for HIV-1–specific responses using the T cell library method (Geiger et al., 2009). As above, 192 cell lines per subset were established at limiting dilution and polyclonally expanded. Each cell line was screened for reactivity to peptides spanning the entire consensus clade C HIV-1 proteome, pooled by proteins Gag, Env, Pol, Nef, and a mix of Vpr, Vpu, Tat, Rev, and Vif (designated Nef/Acc). Responding cells were detected by 3H thymidine incorporation.
HIV-1–specific CD4+ T cells were detected in the naive subsets of all donors (Fig. 2 A) at frequencies ranging from 28 to 129 cells/million (Fig. 3, A and D), comparable to precursor frequencies of tetanus toxoid (TT)–specific naive CD4+ cells (Fig. 3 D). Although the use of a single consensus virus sequence allowed a comprehensive assessment of preexisting specificity, it must underestimate the total response specific for this highly variable virus. However, because 70% of new HIV-1 infections are established by a single founder virus (Keele et al., 2008), the frequencies presented herein should realistically predict an individual’s CD4+ T cell response in acute infection.
In addition to the naive repertoire, HIV-1–specific T cells were also detected within the central memory (Fig. 2 B) and/or effector memory (Fig. 2 C) CD4+ T cell compartments of all donors. However, the observed frequencies were 15–20-fold lower (Fig. 3, B–D) than TT-specific memory CD4+ T cells (Fig. 3 D). The high circulating frequencies of TT-specific memory CD4+ T cells are consistent with previous reports (Geiger et al., 2009) and reflect successful prior immunization (Sallusto et al., 2010). In contrast, the low frequencies of HIV-1–specific memory CD4+ T cells in unexposed donors suggest that these T cells arose because of rare cross-reactivities with non–HIV-1 antigens.
Considerable inter-donor variation in both the specificity and frequency of naive and memory HIV-1–specific CD4+ T cells was observed (Fig. 3, A–C), probably contributing to the great variation in adaptive immune responses seen after natural infection and vaccination. Overall, the HIV-1–specific memory T cell frequencies detected herein were ∼10-fold higher than those recently reported using tetramer enrichment (Su et al., 2013). However, this is not a real discrepancy because we included peptides spanning the entire HIV-1 proteome, whereas they focused their analysis to a single known epitope; therefore, the findings are fully compatible.
After normalization of the dataset for viral protein size, HIV-1 Gag and Env proteins were the immunodominant targets of the preexposure naive and memory CD4+ T cell subsets (Fig. 3 E), similar to the known patterns of immunodominance in natural HIV-1 infection (Kaufmann et al., 2004; Ranasinghe et al., 2012). Consistent with the low frequency of Pol-specific responses observed after infection (Kaufmann et al., 2004; Ranasinghe et al., 2012), relatively few naive or memory CD4+ T cells specific for Pol were detected in the preexposure repertoire. Low expression of Pol protein in infected cells may account for the lack of Pol-specific responses after infection but cannot explain why these CD4+ T cells are rare in the naive repertoire. Because Pol is a relatively conserved protein, it is possible that T cells with this specificity are more likely to be deleted during thymic selection due to cross-reactivities with endogenous retrovirus sequences.
Peptide specificity and avidity of preexisting HIV-1–specific T cells
We next mapped the fine specificity of 68 preexisting naive, central memory, and effector memory HIV-1–specific CD4+ T cells in a total 68 T cell lines (Fig. 4 and Table S2). Epitopes were found to span the entire HIV-1 proteome (Fig. 4) with the majority of positive cell lines (76%) showing a single, unique peptide specificity, consistent with previous observations that preexisting memory CD4+ T cell responses are largely monoclonal (Geiger et al., 2009; Su et al., 2013).
The functional avidity of a subset of naive (n = 7; Fig. 5 A), central memory (n = 11; Fig. 5 B), and effector memory (n = 7; Fig. 5 C) CD4+ T cell lines was assessed by peptide titration. Overall, the avidities were comparable to those found in T cells responding to natural HIV-1 infection, with both high and low avidity responses (EC50 3.125–0.001577 µM) detected, but no significant difference between subsets.
The immunogens that primed the preexisting HIV-1–specific memory CD4+ T cells detected within seronegative donors are unknown and likely to come from many sources. Sequence identity searches (Altschul et al., 1990; Edgar, 2010) performed using the reactive HIV-1 peptide sequences identified potential epitope-length (8–12 aa) subsequence matches to a variety of human (Tables S4 B and S5 B) and human microbiome proteins (Tables S4 and S5). Epitope matches ranged from 1 to 230 microbial sequences, with 83% of all HIV-1 epitopes mapped to the CD4+ memory subsets shown to have epitope length (8–12 aa) matches to human microbiome proteins (Tables S4 A and S5 A). These data suggest that microbial proteins could have contributed to T cell priming. Our list of potential cross-reactive epitopes (Tables S4 and S5) is unlikely to be exhaustive because the degree of sequence identity required to activate cross-reactive T cell responses is unpredictable, and highly divergent epitopes can elicit cross-reactive responses in mice (Birnbaum et al., 2014). Moreover, database searches are limited to sequenced organisms and gene prediction algorithms which may miss cryptic epitopes that could be processed and presented during the course of infection (Ho and Green, 2006).
Known immunoprevalent epitopes are strongly represented in the preexposure repertoire
We next compared the mapped epitopes identified in the preexisting repertoire against HIV-1 epitope data stored in the Los Alamos National Laboratory (LANL) database (LANL-Immunology-Database, http://www.hiv.lanl.gov/content/immunology). We found that 70% of epitopes detected had previously been reported in natural HIV-1 infection (Fig. 4 A and Table S3). Many of these epitopes are presented by multiple Class II HLA types, making them immunoprevalent across the population (Table S3). Indeed, 10% of the mapped epitopes detected within this study were recognized by more than one donor (Fig. 4 and Table S2), independent of HLA type (Table S1). These data imply that T cell precursor frequency and specificity in both the naive and memory subsets before infection could play a large part in determining what is immunodominant after infection.
Analysis of the mapped HIV-1 epitopes according to CD4+ T cell subset (Fig. 4, B and C) showed that independent of donor HLA or antigen sensitivity, 88% of the epitopes recognized by preexisting naive CD4+ T cells had previously been detected in natural infection, whereas a significantly lower proportion (56%) was found for epitopes recognized by memory CD4+ T cells (P = 0.0028). HIV-1 preferentially infects HIV-1–specific memory CD4+ T cells (Douek et al., 2002), and central memory CD4+ T cells in particular are rapidly depleted during acute infection leaving very few during chronic infection (Younes et al., 2003). Thus, preexisting memory HIV-1–specific T cells could be depleted during the acute stages of infection, allowing naive T cells to expand preferentially. Alternatively, the higher proportion of novel epitopes (42%) identified in the preexisting memory CD4+ T cell repertoire may help to promote diversity of the T cell response in natural infection and could be of clinical benefit to host (Rosenberg et al., 1997).
Detection of preexisting, cross-reactive memory CD4+ T cells in unexposed donors is not restricted to HIV-1
Finally, we asked whether detection of cross-reactive memory CD4+ T cells in unexposed donors is unique to HIV-1. We screened five of the same healthy donors for proliferative responses to peptides spanning the entire envelope protein of the Zaire reference strain of Ebola virus. Because Ebola infection is associated with an extremely high mortality rate (WHO, 1978), our donors cannot have been previously exposed. In addition to naive Ebola Env-specific cells (mean 25.26 specific cells/million; Fig. S1 A), Ebola-specific memory CD4+ T cells were detected in 4/5 donors tested (mean 142 specific cells/million; Fig. S1, B and C). The frequencies of Ebola Env-specific T cells were modestly higher than those observed for HIV-1 (Fig. S1 D) but still 10-fold lower in the memory subsets than TT-specific memory cells (Fig. S1 D and Fig. 3 B). These observations demonstrate that the findings on preexposure CD4+ repertoires reported herein are not limited to HIV-1.
The present dataset provides the first comprehensive, systematic, and quantitative analysis of the preexposure HIV-1–specific CD4+ T cell repertoire in HLA diverse seronegative donors. Using peptides spanning the entire HIV-1 proteome, we show that both specificity and avidity of the preexposure HIV-1–specific CD4+ T cell repertoire has considerable overlap with those of CD4+ T cells detected after natural HIV-1 infection. Furthermore, we suggest that some preexisting memory HIV-1–specific T cells may have been primed by microbial organisms present within the human microbiome. These data help explain immunodominance and immunoprevalence in natural HIV-1 infection and the variability in human immune responses to infection and vaccines.
MATERIALS AND METHODS
Study participants and approval.
Leukapheresis samples were obtained from a total of 15 anonymous HIV-1 seronegative, healthy individuals recruited by the Basel Swiss Red Cross Blood Centre and National Blood Service (Bristol, UK). Informed, written consent was obtained from all donors and human primary cell protocols were approved by the Federal Office of Public Health (N. A000197/2 to F. Sallusto).
DNA was extracted using 5 PRIME Achieve Pure DNA kit (Prima Scientific) as per manufacturer’s recommendations. HLA typing (Weatherall Institute of Molecular Medicine, Oxford, England, UK) was performed using the sequence-specific primer method adapted from Bunce (2003), which uses allele-specific primer combination in PCR amplification to provide absolute HLA resolution to two digits and high-probability resolution to four digits. HLA types for all donors are shown in Table S1.
Synthetic peptides were synthesized by Sigma-Aldrich, and/or the Medical Research Council Human Immunology Unit, WIMM (Oxford, UK), as 18mers overlapping by 10 aa. The Zaire Ebola Reference Strain (Zaire-strain Mayinga-76; FASTA ID VGP_EBOZM, UNIPROT Q05320; UniProt Consortium, 2012) was used to design peptides (n = 82) spanning the envelope glycoprotein of Ebola. HIV-1 peptides spanned the entire HIV-1 proteome and were designed according to the 2004 consensus clade C HIV-1 sequence (LANL-Sequence-Database, http://www.hiv.lanl.gov/). In total, 386 HIV-1–specific peptides were synthesized and pooled according to proteins Gag, Env, and Pol, with a single pool comprising Nef and Accessory proteins Vif, Tat, Vpr, and Vpu (referred to as NEF/ACC). In addition, known VZV epitopes gE (531–545 aa; TSPLLRYAAWTGGLA) and IE63 (229–243 aa; QRAIERYAGAETAEY) were synthesized as 15-aa peptides and complexed to DRB1*1501 tetramers conjugated to PE (NIH Core Tetramer Facility). Quality control checks ensured that the peptide of interest was one of the three major peaks by mass spectrometry and all peptide stocks were sterility tested before use. All peptides were used at a final concentration of 2 µg/ml, and the DMSO concentration in either peptide pools or negative control wells did not exceed 0.45% DMSO. Whole protein TT (provided by G. Galli, Novartis Vaccines, Siena, Italy) was used as a positive control at a final concentration of 5 µg/ml.
Tetramer enrichment protocol.
Untouched CD4+ T cells were isolated from PBMCs using magnetic microbeads (Miltenyi Biotec). Tetramer staining and enrichment was performed as described previously (Su et al., 2013). In brief, cells were incubated for 30 min with live/dead Aqua marker (Invitrogen), washed, and then labeled with either gE or IE63 tetramers at room temperature for 45 min (14 µg/ml). Surface markers AF700-labeled anti-CD3 (UCHT3; BD), FITC-labeled anti-CD4 (SK3; BD), Pacific blue–labeled anti-CD45RA (MHCD45RA28; BD), and PE-cyanine 7 (PECy7)–labeled anti-CD56 (B159; BD), anti-CD14 (M5E2; BD), and anti-CD8 (SK1; BD) were incubated at room temperature for 15 min. Before tetramer enrichment, 1/10th staining volume was removed and added to TruCount tubes (BD) to give an absolute count of the starting number of CD4+ naive and memory T cells. The remaining staining volume was enriched for tetramer-positive cells using anti-PE microbeads (Miltenyi Biotec) and added to a separate TruCount tube. Samples were acquired using an LSR Fortessa (BD), and the frequency of tetramer-positive cells determined by dividing the absolute counts of tetramer positive cells by the starting number of CD4+ naive/memory T cells.
Cell purification and sorting for T cell library.
CD14+ monocytes and CD4+ T cells were isolated from PBMCs by positive selection with antibody-coated microbeads (Miltenyi Biotec). CD14+ monocytes were immediately cryopreserved and stored in liquid nitrogen until required for use as antigen-presenting cells in subsequent stimulation assays. CD4+ T cell subsets were cell sorted to 99% purity on a FACSAria (BD) after staining with FITC-labeled anti-CD45RA (ALB11; Beckman Coulter), allophycocyanin-labeled anti-CD4 (SK3; BD), and anti-CCR7 (150503; R&D Systems), followed by staining with biotinylated anti-IgG2a (SouthernBiotech) and streptavidin–Pacific blue (Invitrogen). PE-cyanine 5 (PC5)–labeled anti-CD56 (N901 [NHK-1]; Beckman Coulter), anti-CD25 (B1.49.9; Beckman Coulter), and anti-CD8 (B9.11; Beckman Coulter) were included as a dump channel to exclude natural killer, regulatory, and CD8+ T cells.
T cell libraries.
The medium used throughout was RPMI 1640 supplemented with 2 mM glutamine, 1% (vol/vol) nonessential amino acids, 1% (vol/vol) sodium pyruvate, 50 U/ml penicillin, 50 µg/ml streptomycin, and 5% human serum (Swiss Red Cross). Cell-sorted naive (CD45RA+CCR7+), central memory (CD45RA+CCR7+), and effector memory (CD45RA−CCR7−) CD4+ T cell populations were seeded at 2,000 (naive) and 1,000 (memory) cells. Cell lines were polyclonally expanded in the presence of 2.5 × 104 irradiated (45Gy) allogeneic feeder cells, 500 IU/ml IL-2, and 1 µg/ml PHA (Thermo Fisher Scientific). Throughout the culture period, cells were supplemented with exogenous complete media containing 500 IU/ml IL-2 and progressively transferred from 96- to 24-well tissue culture plates. The full protocol has previously been published by Geiger et al. (2009).
After a 16–20-d culture period, T cells lines were counted and 2.5 × 105 T cells/line removed. T cells were washed 4× in 180 µl PBS and rested for a minimum of 4 h at 37°C 5% CO2. Autologous CD14+ monocytes (2.5 × 104/well) were pulsed for 2 h with appropriate peptide pools, or control antigen before co-culture with T cells. Proliferation was measured on day 4 after 16 h incubation with 1 µCi/ml [3H] Thymidine (GE Healthcare). A stringent positivity criteria, defined as a stimulation index of >5 and a delta value (cpm in response to antigen-pulsed monocytes − cpm in response to unpulsed monocytes) of >3 × 103 cpm was adopted, based upon observations made across multiple negative and positive samples assessed by T cell library technique and represented the 99th percentile of delta cpm obtained from unstimulated samples (Geiger et al., 2009). Overall, <3% of lines were excluded because of nonspecific proliferation. Positive cultures identified in the first screening assay were subsequently reanalyzed using a three dimensional matrix mapping approach (Roederer and Koup, 2003) to identify epitope specificity. Precursor frequencies were calculated based on the number of negative wells according to the Poisson distribution and expressed per million cells (Lefkovits and Waldmann, 1979). EC50 values were determined from interpolated dose–response curves using Prism (version 5.00 for Windows; GraphPad Software).
All mapped epitopes were screened against the LANL HIV-1 Molecular Immunology database (LANL-Immunology-Database) to determine whether they had previously been reported in natural HIV-1. HIV-1–specific epitopes mapped in the memory subsets were screened using the National Centre for Biotechnology and Information (NCBI) basic local alignment sequence tool (BLAST) search tool (Altschul et al., 1990) for epitope-length subsequence matches to human proteome and microbiome sequences. Short-sequence optimizations were used for BLAST, as the goal was detection of sequence similarity rather than of bona fide homology. Additional human proteome sequence data were downloaded from UniProt (UniProt Consortium, 2012), whereas further human microbiome sequences (shotgun sequences derived from body location samples) were obtained from the Human Microbiome Project (HMP) clustered gene indices catalog (http://www.hmpdacc.org/HMGC/). The UniProt and HMP sequences were processed with UBLAST (Edgar, 2010) using highly non-stringent criteria (id, 0.01; evalue, 10,000; maxaccepts, 500,000) to generate matches. Both NCBI-BLAST and UBLAST matches were then filtered on the number of amino acid identities and similarities (Tables S4 and S5). For NCBI, BLAST-included epitopes showed a minimum of 8 out of 9 matched amino acids with the highest sequence identities showing 12 of 12 matched amino acids. Sequences identified with UBLAST search algorithms had a minimum of 8 of 8 matched amino acids with the highest levels of sequence identity observed with 12 of 12 amino acids matched.
Epitope maps obtained from the LANL immunology database show 60% of the HIV-1 proteome is covered by previously defined CD4+ T cell epitopes (LANL-Immunology-Database). Because of this greater proportion of sequence with previously known epitopes, we normalized the number of amino acids covered by epitopes identified in this study to the number of amino acids from the HBX2 reference strain of HIV-1 which contain known CD4+ T cell epitopes (1,875 aa) or which contained no epitopes (1,272 aa). Using a Fisher’s exact test, the percentage of HIV-1 proteome covered by previously reported epitopes from the naive and memory CD4+ T cell subsets could be compared. Significance was defined as P ≤ 0.005.
Pearson’s correlation was used to compare the precursor frequencies obtained using the T cell library technique and the tetramer enrichment protocol. Because the dataset included zero values, a log (x + 1) transformation was applied to all data points.
Online supplemental material.
Fig. S1 shows data from the naive, central and effector memory subsets of five healthy leukapheresis donors who were screened for proliferative responses to a pool of peptides spanning the envelope protein of the Zaire reference strain of Ebola virus. Table S1 shows the HLA class I and II typing for all 10 donors studied within the context of this manuscript. Table S2 shows the amino acid sequence and delta cpm values for the 68 HIV-1–specific T cell responses detected in the naive, central memory, and effector memory CD4+ T cell subsets of five healthy, HIV-1 seronegative donors. Table S3 shows the results of a Los Alamos database search to determine whether any of the mapped epitopes detected in the preexposure repertoire had previously been reported in natural infection, and lists the citation for instances where the epitope had previously been reported. Table S4 displays the sequence identity matches of HIV-1 epitopes detected in the memory CD4+ T cell repertoire to the human microbiome, human sequences, and all GenBank/EMBL sequences excepting those classified as lentiviral, synthetic, or other. Table S5 shows the results of screening mapped HIV-1–specific epitopes in the memory CD4+ T cell subsets of HIV-1 unexposed uninfected donors against predicted genes from the HMP, shotgun sequences and human sequences from Uniprot. Online supplemental material is available at http://www.jem.org/cgi/content/full/jem.20130555/DC1.
The authors wish to thank Mrs. V.E. Whale and Miss Elena Brenna for technical assistance, Dr. D. Jarrossay for cell sorting, Professor G.S. Ogg for advice and guidance on the work with varicella zoster–specific epitopes, and MHC class II tetramers, and Dr. T. Rostron for HLA typing. In addition, we acknowledge the NIH Tetramer Core Facility (contract HHSN272201300006C) for provision of GE- and IE63-specific HLA Class II tetramers.
Research reported in this publication was supported by the National Institute of Allergy and Infectious Diseases of the National Institutes of Health, and by the Center for HIV/AIDS Vaccine Immunology and Immunogen Discovery, grant number UM1-AI100645-01, the Medical Research Council, the ERC, grant number ERC-2012-ADG-2012314, and the SNSF, grant number CRSII3_147662. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
The authors declare no competing financial interests.
- Abbreviations used:
- basic local alignment sequence tool
- glycoprotein E
- human microbiome project
- immediate early phosphoprotein 63
- Los Alamos National Laboratory
- tetanus toxoid
- varicella zoster virus
- Submitted: 16 March 2013
- Accepted: 29 May 2014
This article is distributed under the terms of an Attribution–Noncommercial–Share Alike–No Mirror Sites license for the first six months after the publication date (see http://www.rupress.org/terms). After six months it is available under a Creative Commons License (Attribution–Noncommercial–Share Alike 3.0 Unported license, as described at http://creativecommons.org/licenses/by-nc-sa/3.0/).