Skip Navigation

Comprehensive Analysis of West Nile Virus–Specific T Cell Responses in Humans

  1. Marion C Lanteri1,
  2. John W Heitman1,
  3. Rachel E Owen1,2,
  4. Thomas Busch1,
  5. Nelly Gefter1,
  6. Nancy Kiely4,
  7. Hany T Kamel4,
  8. Leslie H Tobler1,
  9. Michael P Busch1,2 and
  10. Philip J Norris1,2,3
  1. 1Blood Systems Research Institute
  2. 2Departments of Laboratory Medicine
  3. 3Medicine, University of California, San Francisco, California
  4. 4Blood Systems, Scottsdale, Arizona
  1. Reprints or correspondence: Dr. Philip J. Norris, Blood Systems Research Institute, 270 Masonic Ave., San Francisco, CA 94118 (pnorris{at}bloodsystems.org)
 
Next Section

Abstract

BackgroundCellular responses have been shown to play a role in immune control and clearance of West Nile virus (WNV) in murine models. However, little is known about the immunogenic regions of the virus or the phenotype of responding T cells in human infection

MethodsFrozen peripheral blood mononuclear cells (PBMCs) from 35 WNV-infected blood donors were screened for virus-specific T cell responses by an interferon-γ (IFN-γ) enzyme-linked immunosorbent spot assay that used 452 overlapping peptides spanning all WNV proteins. More-detailed phenotypic studies were performed on subjects with high-magnitude T cell responses

ResultsIn individuals with identified responses, the total number of recognized WNV peptides ranged from 1 to 9 (median, 2 peptides), and the overall magnitude of responses ranged from 50 to 4210 spot-forming cells (SFCs) per 106 PBMCs (median, 130 SFCs/106 PBMCs). A subset of 8 frequently recognized peptides from the regions of the genome encoding membrane, envelope, and nonstructural 3 and 4b proteins was identified. Phenotypic study of the highest magnitude WNV-specific T cell responses revealed that most were mediated by CD8+ cells that expressed perforin and/or granzyme B

ConclusionsThese findings are the first to define the breadth and characteristics of the human T cell response to WNV and have implications for candidate vaccine design and evaluation

West Nile virus (WNV) is a single-stranded, positive-polarity RNA virus of the Flaviviridae family [1]. Its genome encodes for capsid (C), envelope (E), premembrane (prM), and membrane (M) proteins and for 7 nonstructural (NS) proteins that likely contribute to viral replication [2]. Since the introduction of WNV to the United States in 1999, the virus has become endemic nationwide. Humans are typically infected as incidental hosts via mosquito bites. Seasonal outbreaks in the US population have been responsible for >26,000 cases of disease and 1038 deaths. Many more underlying human infections occurred, as WNV infection is asymptomatic in 80% of cases [3]. In the 20% of infected individuals who are symptomatic, WNV fever predominates, with only 1% of infected persons presenting with neurological symptoms [4]. No specific therapy or vaccine has been approved for use in humans, leaving only supportive treatment for WNV infection [5]

Considerable progress has been made in understanding immunity to flavivirus infection, particularly in the murine model. Innate immune responses [6, 7] as well as humoral [8, 9] and cellular immune responses have been implicated in the control of WNV infection. A protective role for T and B lymphocytes against WNV has been demonstrated by transfer experiments involving mice with severe combined immunodeficiency (i.e., those with T cell and B cell deficiency) [10] and Rag1tm/Mom mice (i.e., those with B cell deficiency) [8] that succumb to WNV infection in the absence of passive immunotherapy [11, 12]. In the macaque model, WNV clearance from the blood occurs 5–11 days after inoculation, whereas anti-WNV IgM is not detectable until 9–10 days after inoculation, with neutralizing antibodies appearing an additional 1–2 days later [13]. Finally, in the natural history of WNV infection in humans, clearance of viremia occurs before detection of anti-WNV IgM [14]. Thus, WNV-specific antibodies, although known to play a critical role in initial control of viral replication and clearance of viremia, likely are not the only arm of the immune system responsible for the clearance of WNV from the body

T cells possibly play a role in viral clearance and in protection from disease. Interferon-γ (IFN-γ) secretion by γδ and αβ T cells decreases the WNV load in mouse brains, and adoptive transfer of γδ or αβ T cells reduces the susceptibility of mice to lethal WNV infection [15–17]. CD8+ T cells contribute to WNV clearance, as CD8+ T cell–deficient mice show increased mortality, with surviving mice exhibiting persistent viremia [18]. WNV can also be detected in the central nervous system of C57BL/6 mice lacking either IFN-γ [19] or perforin granules [20]. Major histocompatibility complex (MHC) class I upregulation is mediated by IFN-γ–dependent and nuclear factor-κB–dependent pathways, pointing to a role for T cell–secreted cytokines in combating WNV infection [21]

Although T cells likely play a role in controlling WNV replication, a comprehensive characterization of human T cell responses to WNV has not been previously reported. Studies of other flaviviruses suggest that NS proteins are preferentially recognized by T cells [22]. Studies of humans immunized against dengue virus also support NS proteins as the prime region of T cell immunogenicity [23]. A study that used bioinformatic approaches to T cell epitope mapping identified a number of HLA class I B7–restricted epitopes in WNV [24]. The current study was designed to determine the breadth and specificity of WNV-specific T cell responses across the entire genome expressed in WNV [25]. The phenotype and functional markers of T cells specific for immunodominant epitopes were also characterized

Previous SectionNext Section

Subjects, Materials, and Methods

Study subjectsOf >800,000 United Blood Services donors whose blood was screened for WNV in 2005, samples from 45 were confirmed to be positive for WNV RNA. Thirty-five donors with a WNV RNA–positive blood specimen were enrolled in this study after provision of informed consent, as required by the institutional review board of the University of California, San Francisco (UCSF). Questionnaires that covered 12 possible WNV-related symptoms were administered at study enrollment and 2 weeks later. Donors were considered symptomatic if they reported ⩾4 symptoms on either questionnaire (table 1). Samples were collected in EDTA and shipped overnight to the Blood Systems Research Institute (San Francisco). Ten uninfected control subjects (5 laboratory workers and 5 blood donors) were included and were not statistically different from study subjects with respect to mean age (37.6 years [range, 29–53 years] vs. 49 years [range, 26–73 years]; P>.05) and percentage of females (40% vs. 71%; P>.05)

Isolation of peripheral blood mononuclear cells (PBMCs)PBMCs were isolated on a Ficoll-Paque Plus density gradient (GE Healthcare Bio-Sciences). Aliquots of 10×106 cells were frozen in media that contained 90% FBS (HyClone) and 10% dimethyl sulfoxide (DMSO [Fisher BioReagents]) and were stored in liquid nitrogen

WNV transcription-mediated amplification (TMA), load, and antibody testingWNV TMA testing was performed at Blood Systems Laboratories (BSL; Tempe, AZ), using reagents manufactured by Gen-Probe. WNV loads were measured at the National Genetics Institute (Los Angeles, CA). WNV IgM and IgG testing was performed using ELISA kits (Focus Diagnostics) in accordance with the manufacturer’s instructions. Plaque reduction neutralization testing was performed at the Centers for Disease Control and Prevention (Fort Collins, CO) [26]

WNV peptidesA set of 452 peptides (16–18 amino acids long, overlapping by 10 amino acids) spanning the entire expressed genome of NY99-flamingo382-99 strain of WNV (GenBank accession number, AF196835) was obtained through the National Institutes of Health Biodefense and Emerging Infections Research Repository (Manassas, VA). Peptides were reconstituted in water with ⩾10% DMSO

Enzyme-linked immunosorbent spot (ELISPOT) assaysPeptides were divided into 93 pools of 8–10 peptides, with each peptide present in 2 pools [27]. The pattern of responses to the pools was indicative of candidate peptides, which were individually retested in duplicate. PBMCs were plated in 96-well polyvinylidene plates (Millipore) precoated with 2 μg/mL of anti–IFN-γ antibody (mAb1-D1K [Mabtech]) in PBS (UCSF Cell Culture Facility). Cells were added at a concentration of 100,000 cells/well in 100 μL of RPMI (UCSF Cell Culture Facility) plus 10% human AB serum (Sigma), 11 mmol/L of glucose, 2 mmol/L of l-glutamine, 23 mmol/L of NaHCO3P, and 50 U of penicillin-streptomycin (i.e., R10H). The final concentration of each peptide was 5 μg/mL. Duplicate negative control wells, which contained cells plus medium, and a positive control well, which contained 0.5 μg/mL of phytohemagglutinin (PHA; Sigma), were run on each plate. Plates were then incubated for 20 h at 37°C and 5% CO2 and washed 6 times with PBS, after which they were incubated at 25°C for 60 min with biotinylated anti–IFN-γ antibody (Mabtech). Plates were washed 6 times and incubated with streptavidin-alkaline phosphatase (Mabtech) for 60 min at 25°C. Plates were washed before addition of 5-bromo-4-chloro-3-indolylphosphate and nitroblue tetrazolium substrates (Bio-Rad). Spots were counted using an automated ELISPOT plate reader (Cellular Technology). Results were expressed as spot-forming cells (SFCs) per 106 PBMCs after subtraction of background values. Responses in 5 WNV-negative healthy control subjects ranged from 0 to 30 SFCs/106 PBMCs, with >80% of wells showing no spots. Responses were counted as positive if they were 3 times above the background value of the plate and consisted of ⩾50 SFCs/106 PBMCs

Amino acid sequence comparisonHomology to antigens potentially able to induce cross-reactive responses to WNV peptides was investigated using the protein-specific Basic Local Alignment Search Tool (BLASTP) from the National Center for Biotechnology Information (NCBI) [28]. We used the “nr” peptide sequence databases, including all nonredundant GenBank CDS translations, RefSeq Proteins, PDB, SwissProt, PIR, and PRF databases

Peptide stimulation and flow cytometryOne million PBMCs were stimulated in 250 μL of R10H with 200 ng/mL of staphylococcal enterotoxin B (SEB; Sigma) as a positive control or with 10 μg/mL of individual peptides at 37°C and 5% CO2 for 2 h before the addition of 10 μg/mL of brefeldin A (MP Biomedicals). Cells were incubated overnight at 37°C and 5% CO2 and then washed and stained. Except where noted, reagents were purchased from BD Biosciences/Pharmingen. For the phenotypic panel, cells were stained with EMA, anti–CD4-Alexa 700, anti–CD8-PE, anti–CD45RA-PECy7, and anti–CD27-FITC at 4°C in the dark for 10 min and under bright light for 10 min. Cells were fixed with medium A (Caltag-Invitrogen), permeabilized with medium B (Caltag), and stained with anti–CD3-Pacific Blue and anti–IFN-γ-APC. For the cytotoxic panel, cells were stained with anti–CD4-Alexa 700 and anti–CD8-PECy7 at 4°C for 10 min in the dark and for 10 min under bright light. A dump channel with EMA, CD14, CD16, and CD19 was used to exclude dead cells, monocytes/macrophages, NK cells, and B cells. Cells were then fixed, permeabilized, and stained with anti–CD3-Pacific Blue, anti–IFN-γ-APC, anti–perforin-FITC, and anti–granzyme B-PECy5.5 (Caltag). Data were acquired on a 3-laser BD LSRII instrument and were analyzed using FlowJo software (Tree Star)

HLA typing and restrictionDNA was extracted from PBMCs by use of QIAamp DNA Mini kits (Qiagen) in accordance with the manufacturer’s instructions. HLA class I/II typing was performed at BSL for 18 subjects with the highest-level responses, using a sequence-specific primer PCR [29]. HLA restriction analysis of responses from subjects with high-magnitude T cell responses was performed by ELISPOT, in which 5 μg/mL of peptides were preincubated for 16 h with B cell lines partially HLA matched to responders. After 2 washes, 50,000 loaded B cells were incubated with 100,000 PBMCs in duplicate wells. Positive controls included PBMCs plus peptide or PHA, and the background level was calculated using PBMCs with unloaded B cells

Previous SectionNext Section

Results

Breadth and magnitude of T cell responses to acute WNV infectionComprehensive WNV-specific T cell responses were measured by ELISPOT for each of the 35 WNV-positive participants at 3 time points: 15 days, 1 month, and 2 months after the index donation. There was a wide distribution of responses: 27 of 35 WNV-positive individuals recognized 1–9 peptides at ⩾1 of the 3 time points after the index donation, with the majority of peptides recognized at only 1 time point. The most frequently recognized regions of the WNV genome were those encoding the proteins C, M, E, NS3, NS4a, and NS4b (figure 1A and table 2). Overall, 8% of the peptides expressed in the WNV genome were recognized by the T cells of at least 1 participant. The percentage of peptides recognized within individual proteins was variable, with the greatest breadth of recognition in C (20% of peptides), M (19%), E (12%), NS4a (17%), and NS4b (18%). Peptide 30 from M was highly targeted and was recognized by >20% of individuals (table 2). Other peptides targeted by >9% of the donors were peptide 30 from M, peptides 38 and 39 from E, peptides 217, 264, and 265 from NS3, and peptides 299 and 300 from NS4b. Among the frequently targeted peptides, 3 epitopes spanned each of the following pairs of overlapping peptides: 38 and 39 from E, 264 and 265 from NS3, and 299 and 300 from NS4b (table 2). None of the frequently recognized WNV peptides elicited responses in uninfected control subjects. A BLASTP search on the NCBI Web site revealed matching amino acid sequences only between WNV and other flaviviruses (table 2)

Figure 1
View larger version:
Figure 1

Magnitude and diversity of responses to West Nile virus (WNV) peptides. A Study subjects’ response to each individual peptide. Subjects were considered to have responded if they recognized a peptide at any 1 of the following 3 time points: 15 days, 1 month, and 2 months after the index donation. B Peak positive responses, determined by interferon-γ enzyme-linked immunosorbent spot analysis after overnight incubation of peripheral blood mononuclear cells (PBMCs) with individual WNV peptides. The highest response to a given peptide from the 3 time points tested was averaged among all donors with a response. Error bars standard errors. SFC, spot-forming cell

Among the 27 donors with T cell responses, the average summed magnitude of positive responses within a given donor against individual peptides ranged from 50 to 4210 SFCs/106 PBMCs (median, 130 SFCs/106 PBMCs). The highest magnitude responses (mean magnitude, >100 SFCs/106 PBMCs) were elicited by peptides within M, E, NS3, NS4a, and NS4b (table 2). Peptides 30, 38 and 39, 264 and 265, and 299 and 300 appeared to be immunodominant epitopes, because they elicited high-magnitude responses and were among the most frequently recognized peptides (figure 1 and table 2). The breadth of WNV epitopes recognized did not differ between symptomatic and asymptomatic subjects (mean, 2.0±2.6 vs. 2.2±1.88 peptides; P=.76), nor did the magnitude of positive responses (326 vs. 425 SFCs/106 PBMCs; P=.59)

Late immune response to WNVThe durability of T cell responses to WNV was tested 1 year after the index donation in samples from 20 donors, which included the 5 donors with the highest early responses. None of the samples had ⩾50 SFCs/106 PBMCs to peptide pools at 1 year. If the ELISPOT assay cutoff was adjusted to just above the value for negative control wells, responses to the immunodominant peptides could still be detected using peptide pools in 10% of previously responsive subjects (data not shown). These data demonstrated only very low-frequency persistence of WNV-specific T cells 1 year after infection

HLA restrictionResponses to peptides 30, 38, and 299 were associated with particular HLA types. Seven of 10 donors expressing A*0101 recognized peptide 30, compared with 0 of 8 A1-negative donors. Four of 5 C*0303-positive or Cw*0304-positive donors recognized peptide 38, compared with 1 of 13 HLA C3–negative donors. The Cw*0303 and Cw*0304 alleles were treated as equivalent, because the only coding difference between them is at codon 91, which is not predicted to affect the HLA binding groove or T cell receptor contact sites [30]. Finally, 3 of 6 A*0201-positive donors responded to peptide 299, and 0 of 12 HLA A2–negative donors responded. These associations agreed with HLA restriction experiments that involved B cell lines partially HLA matched to subjects 2250 or 2350, the subjects with the highest magnitude responses (figure 2A–C). Exploration of HLA-binding motifs by use of SYFPEITHI software [31] showed that the second most likely binding motif for peptide 30 was A*0101 and that the most likely binding motif for peptide 299 was A*0201 (table 3). The binding motif analysis supported the HLA restriction results: peptide 30 was presented by A*0101, peptide 38 by Cw*0303/0304, and peptide 299 by A*0201

Figure 2
View larger version:
Figure 2

HLA restriction analysis of peripheral blood mononuclear cells (PBMCs) from donors who had the highest magnitude response. PBMCs were tested for presentation of the most frequently recognized peptides, by means of allogeneic B cell lines with partially matched HLA alleles (bold font; displayed on the left side of the charts). Peptide 30 (A) and peptide 38 (B) were tested with PBMCs from donor 2350. Peptide 299 (C) was tested with PBMCs from donor 2250. Charts show raw data for peptide-pulsed (black bars) and non–peptide-pulsed (open bars) B cell lines. Responses were counted if they were ⩾1.5 times the background level. SFC, spot-forming cell

High-magnitude WNV-specific responses are mediated primarily by CD8+ T cellsWNV-specific T cells from donors with high-magnitude responses (i.e., >500 SFCs/106 PBMCs) were examined by flow cytometry (figure 3). PBMCs from 3 subjects (symptomatic donor 2250 and asymptomatic donors 2350 and 2450) were stimulated with their respective cognate WNV peptides or with SEB as a positive control. Cells were gated by forward scatter and side scatter to determine which cells were lymphocytes. Lymphocytes were then separated into live CD4+ and CD8+ populations. IFN-γ positivity was plotted against CD3 positivity to enumerate the fraction of cells specific for WNV (figure 3A). IFN-γ–positive cells induced by peptide stimulation were predominantly CD8+ T cells, except in donor 2450, in whom low-frequency CD4+ T cell responses were also observed (figure 3B)

Phenotype of WNV-specific T cellsTo further characterize the phenotype of WNV-specific T cells, PBMCs were stained before and after stimulation (figure 4). Cells were classified as naive (CD45RA+CD27+), effector (CD45RA+/−CD27−), or memory (CD45RA−CD27+) [32] (figure 4A). Effector cell frequencies in the CD8+ T cell subset were not significantly different between WNV-infected and WNV-uninfected subjects (P=.32), whereas the frequency of effector CD4+T cells was lower among WNV-infected subjects (P=.006, by the t test) (figure 4B). For donor 2250, 47% of WNV-specific CD8+ T cells 1 month after infection were effector cells, compared with 19% of unstimulated CD8+ T cells, whereas for donor 2350, effector cells comprised 55% and 49% of WNV-specific and unstimulated CD8+ T cells, respectively, with WNV-specific CD8+ T cells almost equally divided between effector and memory populations (figure 4C). WNV-specific T cells also showed high expression of the cytolytic markers perforin and granzyme B, consistent with the effector phenotype of these cells (figure 5). Donors 2250 and 2350 exhibited higher median levels of WNV-specific CD8+ T cells that expressed cytolytic markers, compared with unstimulated (i.e., non–WNV-specific) cells (85% vs. 23%; P<.001)

Previous SectionNext Section

Discussion

In this study, we performed the first comprehensive analysis of the breadth and specificity of WNV-specific T cell responses in humans. We report WNV-specific T cell responses in the majority of WNV-infected individuals. Responses were preferentially directed against C, M, E, NS3, and NS4b, with 12%–20% of peptides from these regions recognized by at least 1 subject in our cohort. By 1 year after the index donation, the diversity of responses had waned, with only peptides associated with a strong early response still able to elicit responses directly ex vivo. Analysis of subjects with the highest magnitude responses revealed most responses to be CD8+ T cell mediated, with responding cells having an effector phenotype and expressing perforin and granzyme B

The combined magnitude of positive responses to WNV (median, 130 SFCs/106 PBMCs [range, 50–4210 SFCs/106 PBMCs]) was less than that observed in human immunodeficiency virus (HIV) infection (median, 4245 SFCs/106 PBMCs [range, 280–25,860 SFCs/106 PBMCs]) [33] but comparable to that observed in Epstein-Barr virus (EBV) infection (median, 136 SFCs/106 PBMCs [range, 27–1344 SFCs/106 PBMCs]) [34], cytomegalovirus (CMV) infection (median, 973 SFCs/106 PBMCs for IE1 protein recognition [range, 0–3900 SFCs/106 PBMCs]) [35], and HCV infection (range, 25–285 SFCs/106 PBMCs) [36]. Eight of 35 subjects did not have a WNV-specific T cell response of >50 SFCs/106PBMCs, implying that systemic infection can occur with weak induction of T cell activity. Of the 3 subjects with the strongest WNV-specific T cell responses, subject 2250 (who had symptomatic infection) had traveled to a dengue-endemic area in the past 5 years and had preexisting IgG antibody cross-reactive with WNV, and subject 2450 (who had asymptomatic infection) had been vaccinated against yellow fever. These data raise the possibility that prior flavivirus exposure increased the magnitude of WNV-specific T cell responses during primary infection but did not necessarily prevent symptoms

WNV E protein is known to be a major target for neutralizing antibodies [37, 38]. IFN-γ T cell responses to E peptides were reported after immunization of humans with a live attenuated recombinant WNV vaccine, using WNV M and E in a yellow fever virus backbone [39]. Our study confirmed the importance of WNV E protein as a major target of immune responses. Potential T cell epitopes in the E proteins of Murray Valley encephalitis virus, Japanese encephalitis virus, WNV, and dengue virus have been identified by computer prediction algorithms. One epitope in the C-terminal region of the E glycoprotein (approximate location, amino acids 420–455) was predicted to induce immune responses in all flaviviruses analyzed [40] and was confirmed in our study as peptide 95, which was 94% homologous with dengue virus and induced high-magnitude responses (table 2). Another bioinformatics approach to T cell epitope mapping identified 16 HLA B7–restricted epitopes in WNV [24]. None of these epitopes induced responses in our study, which included 2 HLA B7–positive donors who responded to other WNV peptides

Figure 3
View larger version:
Figure 3

CD4+ versus CD8+ T cell mediation of high-magnitude West Nile virus (WNV)–specific responses. A Results of flow cytometry to determine CD4+ and CD8+ populations of T lymphocytes. Gates were set on unstimulated (unstim) cells for interferon-γ (IFN-γ) secretion. CD4+ and CD8+ T cell populations were stained for IFN-γ+ expression after overnight stimulation and incubation with brefeldin A. EMA, ethidium monoazide; FSC, forward scatter; SSC, side scatter. B Rate of IFN-γ+ secretion among CD4+ or CD8+ T cell populations after no stimulation, staphylococcal enterotoxin B (SEB) stimulation, WNV peptide stimulation. Gates for IFN-γ secretion were set using an aliquot of unstimulated cells. Results are shown for responses >0.1% IFN-γ–positive cells, and a minimum of 100,000 lymphocyte-gated events were collected in each experiment

Figure 4
View larger version:
Figure 4

Phenotype of West Nile virus (WNV)–specific T cells. A Gating strategy to define naive T cells (CD45RA+CD27+), memory T cells (CD45RA−CD27+), and effector T cells (CD45RA+/-CD27−). Gates for interferon-γ (IFN-γ) secretion were set on unstimulated cells. Live lymphocytes were gated into CD4+ and CD8+ populations. CD45RA+ and CD27+ gates were set on the IFN-γ–negative (IFN-γ−) and IFN-γ–positive (IFN-γ+) CD3+ T cell populations. Fluorescence-minus-one tubes were used for CD45RA and CD27 stains. EMA, ethidium monoazide; FSC, forward scatter; SSC, side scatter. B Phenotypes of T cells in 10 WNV− control subjects and 23 WNV+ subjects 1 month after the index donation. C Phenotypes of T cells after no stimulation (unstim; i.e., IFN-γ−), staphylococcal enterotoxin B (SEB) stimulation, and antigen-specific stimulation (i.e., IFN-γ+) in 2 donors with the highest magnitude CD8+ T cell responses 1 month after the index donation. At least 100,000 lymphocyte-gated events were collected, and at least 100 events in the IFN-γ+ gate were collected

Figure 5
View larger version:
Figure 5

Cytolytic effector molecule expression by West Nile virus (WNV)–specific T cells. Gates were set on unstimulated cells for interferon-γ (IFN-γ) secretion and fluorescence-minus-one (FMO) tubes for perforin and granzyme B stains. A Unstimulated, staphylococcal enterotoxin B (SEB)–stimulated, and peptide-stimulated conditions for donor 2250 (smoothed lines) 1 month after the index donation. Peripheral blood mononuclear cells were stained for IFN-γ plus perforin and granzyme B after secretion after overnight stimulation and incubation with brefeldin A. Live lymphocytes were gated on IFN-γ–negative (IFN-γ−) and IFN-γ–positive (IFN-γ+) populations. B Rates of perforin and/or granzyme B expression among CD8+ IFN-γ+ cells after no stimulation, stimulation with SEB, or stimulation with WNV-specific peptides in donors 2250 and 2350 1 month after the index donation. Gates for perforin and granzyme B were set using FMO controls. A minimum of 100,000 lymphocyte-gated events were collected, and at least 100 events in the IFN-γ+ gate were collected

Table 1
View larger version:
Table 1

Characteristics of blood donors who tested positive for West Nile virus (WNV) RNA

Table 2
View larger version:
Table 2

Characteristics of West Nile virus (WNV) peptides frequently recognized by WNV-positive persons and findings of the Basic Local Alignment Search Tool (BLAST)

Table 3
View larger version:
Table 3

HLA class 1 binding motifs for West Nile virus peptides 30 and 299

The nonstructural proteins NS3, NS4a, and NS4b have been described as immunodominant targets for class I MHC–restricted responses in mice infected with Kunjin virus or WNV [22]. Cytotoxic T lymphocytes (CTLs) that recognize prM, E, NS1, NS2a, and NS3 proteins were also identified in volunteers immunized against dengue virus, and some responses cross-reacted with WNV [23]. Human T cell responses against dengue virus C, prM, M, E, NS1, NS2a, NS3, NS4a, NS4b, and NS5 epitopes have been reported [23, 41–45]. Several of the WNV peptides we reported are derived from proteins that are homologous to immunogenic flavivirus proteins and have similar amino acid sequences. Of the 8 most frequently recognized peptides, 3 were from NS3, and 2 from NS4b (figure 1A)

We were able to characterize the phenotype of T cells responding to WNV in the 2 subjects with the highest frequency of CD8+ T cell responses (figures 4 and 5). The results showed that WNV-specific CTLs express the cytolytic markers perforin and granzyme B and were roughly equally divided between memory and effector cell phenotypes 1 month after the index donation. Although further studies would be required to confirm these findings in more individuals, WNV would appear to induce CTLs in the middle of the spectrum of maturity, compared with chronic viral infections inducing early (e.g., EBV or HCV), intermediate (HIV), or late (CMV) memory differentiated populations in the months following the acute phase of infection [27, 32, 46]. The early high-frequency responses we observed were only very weakly detected 1 year after infection, implying that the WNV-specific memory cells persisted at very low frequencies

Potential limitations of this study include the peptides used to stimulate WNV-specific responses and the specimen handling and storage process. The WNV peptides used were 16–18 amino acids long, overlapping by 10 amino acids. Although longer than optimal CD8+ T cell epitopes, these peptide lengths have been shown to induce reasonably strong CD8+ and CD4+ T cell responses, representing a cost-effective compromise in measuring responses to the whole expressed viral genome [47, 48]. Few fixed mutations have been described in currently circulating US strains, compared with the reference WNV strain used as the template for peptide synthesis [49]. Though WNV is not highly variable, it is possible that sequence variation decreased the ability to observe responses in the current study. Finally, blood was collected at regional blood centers and shipped overnight to the laboratory, delaying cell processing by up to 24 h. This may also have reduced the magnitude of T cell responses we observed, resulting in underestimation of the true magnitude of WNV-specific responses [50]

In addition to defining frequently recognized regions of WNV, HLA restriction was performed for 3 immunodominant epitopes (figure 2). These epitopes were presented by the relatively common HLA alleles A*0101, A*0201, and Cw*0303/04, implying that these may be useful epitopes to include in vaccine candidates. Notably, despite a high diversity of HLA class I alleles among the WNV-positive individuals, 54% of the subjects and 70% of individuals with responses targeted at least 1 peptide within the subset of 13 frequently recognized peptides (table 2). Our results show that a limited number of peptides can induce WNV-specific T cell immune responses in a genetically diverse population. Our data suggest that immunodominant epitopes were present in M, E, NS3, and NS4b proteins. M and E proteins, already known to be targets for neutralizing antibodies and a focus of vaccine development [51], are also targeted by CTLs in natural WNV infection. Natural infection also induced T cell responses to C, NS3, NS4a, and NS4b proteins, and these regions of the virus should be further studied as new or complementary candidates for WNV vaccine development

Previous SectionNext Section

Acknowledgments

We thank the National Institutes of Health Biodefense and Emerging Infections Research Resources Repository for kindly providing peptide arrays spanning all West Nile virus proteins; Drs. Robert Endres and Jar-How Lee for helpful discussions on HLA-C molecular structure; and Dr. Robert Lanciotti for performing West Nile virus plaque reduction neutralization tests

Previous SectionNext Section

Footnotes

  • Potential conflicts of interest: none reported

    Presented in part: 54th Annual Meeting of the American Society of Tropical Medicine and Hygiene, Washington, DC, 11–15 December 2005

    Financial support: Centers for Disease Control and Prevention (grant R01 C1000214-01)

  • Received July 6, 2007.
  • Accepted November 9, 2007.
Previous Section
 

References

  1. ↵
    1. Hayes EB,
    2. Sejvar JJ,
    3. Zaki SR,
    4. Lanciotti RS,
    5. Bode AV,
    6. Campbell GL
    . Virology, pathology, and clinical manifestations of West Nile virus diseas. Emerg Infect Dis 2005;11:1174-9.
    MedlineWeb of Science
  2. ↵
    1. Mukhopadhyay S,
    2. Kim BS,
    3. Chipman PR,
    4. Rossmann MG,
    5. Kuhn RJ
    . Structure of West Nile viru. Science 2003;302:248.
    FREE Full Text
  3. ↵
    1. Mostashari F,
    2. Bunning ML,
    3. Kitsutani PT,
    4. et al
    . Epidemic West Nile encephalitis, New York, 1999: results of a household-based seroepidemiological surve. Lancet 2001;358:261-4.
    CrossRefMedlineWeb of Science
  4. ↵
    1. Sejvar JJ
    . The long-term outcomes of human West Nile virus infectio. Clin Infect Dis 2007;44:1617-24.
    CrossRefMedlineWeb of Science
  5. ↵
    1. Granwehr BP,
    2. Lillibridge KM,
    3. Higgs S,
    4. et al
    . West Nile virus: where are we no. Lancet Infect Dis 2004;4:547-56.
    CrossRefMedlineWeb of Science
  6. ↵
    1. Ben-Nathan D,
    2. Huitinga I,
    3. Lustig S,
    4. van Rooijen N,
    5. Kobiler D
    . West Nile virus neuroinvasion and encephalitis induced by macrophage depletion in mic. Arch Virol 1996;141:459-69.
    CrossRefMedlineWeb of Science
  7. ↵
    1. Diamond MS,
    2. Shrestha B,
    3. Mehlhop E,
    4. Sitati E,
    5. Engle M
    . Innate and adaptive immune responses determine protection against disseminated infection by West Nile encephalitis viru. Viral Immunol 2003;16:259-78.
    CrossRefMedlineWeb of Science
  8. ↵
    1. Diamond MS,
    2. Shrestha B,
    3. Marri A,
    4. Mahan D,
    5. Engle M
    . B cells and antibody play critical roles in the immediate defense of disseminated infection by West Nile encephalitis viru. J Virol 2003;77:2578-86.
    Abstract/FREE Full Text
  9. ↵
    1. Oliphant T,
    2. Diamond MS
    . The molecular basis of antibody-mediated neutralization of West Nile viru. Expert Opin Biol Ther 2007;7:885-92.
    CrossRefMedlineWeb of Science
  10. ↵
    1. Halevy M,
    2. Akov Y,
    3. Ben-Nathan D,
    4. Kobiler D,
    5. Lachmi B,
    6. Lustig S
    . Loss of active neuroinvasiveness in attenuated strains of West Nile virus: pathogenicity in immunocompetent and SCID mic. Arch Virol 1994;137:355-70.
    CrossRefMedlineWeb of Science
  11. ↵
    1. Diamond MS,
    2. Sitati EM,
    3. Friend LD,
    4. Higgs S,
    5. Shrestha B,
    6. Engle M
    . A critical role for induced IgM in the protection against West Nile virus infectio. J Exp Med 2003;198:1853-62.
    Abstract/FREE Full Text
  12. ↵
    1. Engle MJ,
    2. Diamond MS
    . Antibody prophylaxis and therapy against West Nile virus infection in wild-type and immunodeficient mic. J Virol 2003;77:12941-9.
    Abstract/FREE Full Text
  13. ↵
    1. Ratterree MS,
    2. Gutierrez RA,
    3. Travassos da Rosa AP,
    4. et al
    . Experimental infection of rhesus macaques with West Nile virus: level and duration of viremia and kinetics of the antibody response after infectio. J Infect Dis 2004;189:669-76.
    Abstract/FREE Full Text
  14. ↵
    1. Tilley PA,
    2. Fox JD,
    3. Jayaraman GC,
    4. Preiksaitis JK
    . Nucleic acid testing for West Nile virus RNA in plasma enhances rapid diagnosis of acute infection in symptomatic patient. J Infect Dis 2006;193:1361-4.
    Abstract/FREE Full Text
  15. ↵
    1. Brien JD,
    2. Uhrlaub JL,
    3. Nikolich-Zugich J
    . Protective capacity and epitope specificity of CD8+ T cells responding to lethal West Nile virus infectio. Eur J Immunol 2007;37:1855-63.
    CrossRefMedline
    1. Purtha WE,
    2. Myers N,
    3. Mitaksov V,
    4. et al
    . Antigen-specific cytotoxic T lymphocytes protect against lethal West Nile virus encephaliti. Eur J Immunol 2007;37:1845-54.
    CrossRefMedline
  17. ↵
    1. Wang Y,
    2. Lobigs M,
    3. Lee E,
    4. Mullbacher A
    . CD8+ T cells mediate recovery and immunopathology in West Nile virus encephaliti. J Virol 2003;77:13323-34.
    Abstract/FREE Full Text
  18. ↵
    1. Shrestha B,
    2. Diamond MS
    . Role of CD8+ T cells in control of West Nile virus infectio. J Virol 2004;78:8312-21.
    Abstract/FREE Full Text
  19. ↵
    1. Shrestha B,
    2. Wang T,
    3. Samuel MA,
    4. et al
    . Gamma interferon plays a crucial early antiviral role in protection against West Nile virus infectio. J Virol 2006;80:5338-48.
    Abstract/FREE Full Text
  20. ↵
    1. Shrestha B,
    2. Samuel MA,
    3. Diamond MS
    . CD8+ T cells require perforin to clear West Nile virus from infected neuron. J Virol 2006;80:119-29.
    Abstract/FREE Full Text
  21. ↵
    1. Cheng Y,
    2. King NJ,
    3. Kesson AM
    . Major histocompatibility complex class I (MHC-I) induction by West Nile virus: involvement of 2 signaling pathways in MHC-I up-regulatio. J Infect Dis 2004;189:658-68.
    Abstract/FREE Full Text
  22. ↵
    1. Lobigs M,
    2. Arthur CE,
    3. Mullbacher A,
    4. Blanden RV
    . The flavivirus nonstructural protein NS3 is a dominant source of cytotoxic T cell peptide determinant. Virology 1994;202:195-201.
    CrossRefMedlineWeb of Science
  23. ↵
    1. Mathew A,
    2. Kurane I,
    3. Rothman AL,
    4. Zeng LL,
    5. Brinton MA,
    6. Ennis FA. Dominant recognition by human CD8+ cytotoxic T lymphocytes of dengue virus nonstructural proteins NS3 and NS,
    7. Ennis FA. Dominant recognition by human CD8+ cytotoxic T lymphocytes of dengue virus nonstructural proteins NS3 and NS1
    . 2. J Clin Invest 1996;98:1684-91.
    MedlineWeb of Science
  24. ↵
    1. De Groot AS,
    2. Saint-Aubin C,
    3. Bosma A,
    4. Sbai H,
    5. Rayner J,
    6. Martin W
    . Rapid determination of HLA B*07 ligands from the West Nile virus NY99 genom. Emerg Infect Dis 2001;7:706-13.
    CrossRefMedlineWeb of Science
  25. ↵
    1. Schmittel A,
    2. Keilholz U,
    3. Bauer S,
    4. et al
    . Application of the IFN-gamma ELISPOT assay to quantify T cell responses against protein. J Immunol Methods 2001;247:17-24.
    CrossRefMedlineWeb of Science
  26. ↵
    1. Martin DA,
    2. Noga A,
    3. Kosoy O,
    4. Johnson AJ,
    5. Petersen LR,
    6. Lanciotti RS. Evaluation of a diagnostic algorithm using immunoglobulin M enzyme-linked immunosorbent assay to differentiate human West Nile virus and St
    . Louis encephalitis virus infections during the 2002 West Nile virus epidemic in the United State. Clin Diagn Lab Immunol 2004;11:1130-3.
    CrossRefMedline
  27. ↵
    1. Tomiyama H,
    2. Matsuda T,
    3. Takiguchi M
    . Differentiation of human CD8+ T cells from a memory to memory/effector phenotyp. J Immunol 2002;168:5538-50.
    Abstract/FREE Full Text
  28. ↵
    1. National Center for Biotechnology Information
    . Basic Local Alignment Search Tool [database online]. Available at: http://www.ncbi.nlm.nih.gov/blast/Blast.cgi. Accessed 26 February 2008.
  29. ↵
    1. Bunce M,
    2. O’Neill CM,
    3. Barnardo MC,
    4. et al
    . Phototyping: comprehensive DNA typing for HLA-A, B, C, DRB1, DRB3, DRB4, DRB5 & DQB1 by PCR with 144 primer mixes utilizing sequence-specific primers (PCR-SSP. Tissue Antigens 1995;46:355-67.
    MedlineWeb of Science
  30. ↵
    1. Institute für Transfusionsmedizin Web site
    . Available at: http://www.histocheck.org. Accessed 3 July 2007.
  31. ↵
    1. Rammensee H,
    2. Rammensee H-G,
    3. Bachmann J,
    4. Emmerich NN,
    5. Bachor OA,
    6. Stevanovic S
    . SYFPEITHI: database for MHC ligands and peptide motif. Immunogenetics 1999;50:213-9.
    CrossRefMedlineWeb of Science
  32. ↵
    1. Appay V,
    2. Dunbar PR,
    3. Callan M,
    4. et al
    . Memory CD8+ T cells vary in differentiation phenotype in different persistent virus infection. Nat Med 2002;8:379-85.
    CrossRefMedlineWeb of Science
  33. ↵
    1. Addo MM,
    2. Yu XG,
    3. Rathod A,
    4. et al
    . Comprehensive epitope analysis of human immunodeficiency virus type 1 (HIV-1)–specific T-cell responses directed against the entire expressed HIV-1 genome demonstrate broadly directed responses, but no correlation to viral loa. J Virol 2003;77:2081-92.
    Abstract/FREE Full Text
  34. ↵
    1. Dalod M,
    2. Dupuis M,
    3. Deschemin JC,
    4. et al
    . Broad, intense anti–human immunodeficiency virus (HIV) ex vivo CD8+ responses in HIV type 1–infected patients: comparison with anti–Epstein-Barr virus responses and changes during antiretroviral therap. J Virol 1999;73:7108-16.
    Abstract/FREE Full Text
  35. ↵
    1. Sacre K,
    2. Carcelain G,
    3. Cassoux N,
    4. et al
    . Repertoire, diversity, and differentiation of specific CD8 T cells are associated with immune protection against human cytomegalovirus diseas. J Exp Med 2005;201:1999-2010.
    Abstract/FREE Full Text
  36. ↵
    1. Lauer GM,
    2. Ouchi K,
    3. Chung RT,
    4. et al
    . Comprehensive analysis of CD8+-T-cell responses against hepatitis C virus reveals multiple unpredicted specificitie. J Virol 2002;76:6104-13.
    Abstract/FREE Full Text
  37. ↵
    1. Oliphant T,
    2. Nybakken GE,
    3. Engle M,
    4. et al
    . Antibody recognition and neutralization determinants on domains I and II of West Nile virus envelope protei. J Virol 2006;80:12149-59.
    Abstract/FREE Full Text
  38. ↵
    1. Sanchez MD,
    2. Pierson TC,
    3. McAllister D,
    4. et al
    . Characterization of neutralizing antibodies to West Nile viru. Virology 2005;336:70-82.
    CrossRefMedlineWeb of Science
  39. ↵
    1. Monath TP,
    2. Liu J,
    3. Kanesa-Thasan N,
    4. et al
    . A live, attenuated recombinant West Nile virus vaccin. Proc Natl Acad Sci U S A 2006;103:6694-9.
    Abstract/FREE Full Text
  40. ↵
    1. Kutubuddin M,
    2. Kolaskar AS,
    3. Galande S,
    4. Gore MM,
    5. Ghosh SN,
    6. Banerjee K
    . Recognition of helper T cell epitopes in envelope (E) glycoprotein of Japanese encephalitis, West Nile and Dengue viruse. Mol Immunol 1991;28:149-54.
    CrossRefMedlineWeb of Science
  41. ↵
    1. Bashyam HS,
    2. Green S,
    3. Rothman AL
    . Dengue virus-reactive CD8+ T cells display quantitative and qualitative differences in their response to variant epitopes of heterologous viral serotype. J Immunol 2006;176:2817-24.
    Abstract/FREE Full Text
    1. Green S,
    2. Kurane I,
    3. Pincus S,
    4. Paoletti E,
    5. Ennis FA
    . Recognition of dengue virus NS1-NS2a proteins by human CD4+ cytotoxic T lymphocyte clone. Virology 1997;234:383-6.
    CrossRefMedline
    1. Imrie A,
    2. Meeks J,
    3. Gurary A,
    4. et al
    . Differential functional avidity of dengue virus–specific T-cell clones for variant peptides representing heterologous and previously encountered serotype. J Virol 2007;81:10081-91.
    Abstract/FREE Full Text
    1. Simmons CP,
    2. Dong T,
    3. Chau NV,
    4. et al
    . Early T-cell responses to dengue virus epitopes in Vietnamese adults with secondary dengue virus infection. J Virol 2005;79:5665-75.
    Abstract/FREE Full Text
  45. ↵
    1. Zivna I,
    2. Green S,
    3. Vaughn DW,
    4. et al
    . T cell responses to an HLA-B*07-restricted epitope on the dengue NS3 protein correlate with disease severit. J Immunol 2002;168:5959-65.
    Abstract/FREE Full Text
  46. ↵
    1. Champagne P,
    2. Ogg GS,
    3. King AS,
    4. et al
    . Skewed maturation of memory HIV-specific CD8 T lymphocyte. Nature 2001;410:106-11.
    CrossRefMedline
  47. ↵
    1. Betts MR,
    2. Ambrozak DR,
    3. Douek DC,
    4. et al
    . Analysis of total human immunodeficiency virus (HIV)–specific CD4+ and CD8+ T-cell responses: relationship to viral load in untreated HIV infectio. J Virol 2001;75:11983-91.
    Abstract/FREE Full Text
  48. ↵
    1. Draenert R,
    2. Altfeld M,
    3. Brander C,
    4. et al
    . Comparison of overlapping peptide sets for detection of antiviral CD8 and CD4 T cell response. J Immunol Methods 2003;275:19-29.
    CrossRefMedlineWeb of Science
  49. ↵
    1. Herring BL,
    2. Bernardin F,
    3. Caglioti S,
    4. et al
    . Phylogenetic analysis of WNV in North American blood donors during the 2003–2004 epidemic season. Virology 2007;363:220-8.
    CrossRefMedlineWeb of Science
  50. ↵
    1. Bull M,
    2. Lee D,
    3. Stucky J,
    4. et al
    . Defining blood processing parameters for optimal detection of cryopreserved antigen-specific responses for HIV vaccine trial. J Immunol Methods 2007;322:57-69.
    CrossRefMedlineWeb of Science
  51. ↵
    1. Arroyo J,
    2. Miller C,
    3. Catalan J,
    4. et al
    . ChimeriVax-West Nile virus live-attenuated vaccine: preclinical evaluation of safety, immunogenicity, and efficac. J Virol 2004;78:12497-507.
    Abstract/FREE Full Text
| Table of Contents

This Article

  1. J Infect Dis. (2008) 197 (9): 1296-1306. doi: 10.1086/586898
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

    • Major Articles
      • Viruses
        • Major Articles

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 205 (5)
  1. Journal of Infectious Diseases
  1. Alert me to new issues

Most