mediastinal lymph nodes
influenza virus A/Memphis/1/71 x A/Bellamy/42
influenza virus A/Puerto Rico/8/34
The longevity of the influenza virus-specific CD8+ T cell response following intranasal delivery of a synthetic lipopeptide was investigated and the characteristics and location of the cells associated with viral clearance examined. The lipopeptide, incorporating an epitope for CD8+ T cells and another for CD4+ T cells with the lipid moiety S-[2,3-bis(palmitoyloxy)propyl]cysteine (Pam2Cys) attached, induced potent and long-lived pulmonary protection. Both the lipopeptide and its largely unprotective non-lipidated counterpart elicited comparable numbers of CD8+ T cells in the spleen, which was the main location of the memory pool. However, the lipopeptide, unlike the non-lipidated peptide, also induced a substantial memory population that remained in the lungs and was rapidly activated upon viral challenge months later. These lipopeptide-induced lung-resident CD8+ T cells were also very similar in number and IFN-γ-secreting potential to those induced by prior exposure to the virus itself and are likely mediators of initial viral clearance prior to recruitment from the expanding lymph node T cell pool. Significant clearing responses were demonstrated as late as 9 months post-lipopeptide vaccination. This study shows that CD8+ T cells primed by the lipopeptide are not only long-lived but can take up residence in the lung where they are important early mediators of pulmonary protection.
Induction of CD8+ T cells by vaccination is considered to be highly desirable, particularly in the context of viral infections, such as influenza, where the surface glycoproteins of the virus rapidly accumulate amino acid substitutions that allow the virus to escape from pre-existing antibody responses. Such CD8+ T cell-inducing vaccines, which could target epitopes located on the highly conserved internal components of the virus, would be predicted to provide some degree of pulmonary protection against newly emerging viral strains, including those with pandemic potential.
In the mouse model of pulmonary influenza infection, it has been estimated that immunity that protects the lung against reinfection with a heterologous subtype of virus wanes rapidly (within 3–4 months) after initial infection 1 despite the presence of significant numbers of virus-specific CD8+ T cells being present in the spleen and lymph nodes of the mice for much longer periods 2. More recently, it has been suggested that a better correlate of protective cellular immunity may be the population of virus-specific T cells that become established in the lungs of mice after virus infection 2. These cells have been found to persist in both the airways and lung parenchyma after infection with influenza, Sendai and respiratory syncytial viruses 2–4; they do not cycle, have a half life of about 40 days, but upon re-exposure to antigen can proliferate and express a highly activated phenotype and cytolytic function. It has been proposed that these cells secrete cytokines and other mediators in the lung upon re-exposure to virus, which recruit additional T cells from the conventional memory pool in the lymph nodes and spleen back into the lungs to control the infection. If an association between these lung-resident T cells and protection can be established, then this would be an important population to elicit by vaccination.
The lipopeptide technology has shown its strength in recent years by providing much needed immunogens for the safe and efficient induction of CD8+ T cells (reviewed in 5, 6). We have recently shown that synthetic epitope-based vaccines incorporating the lipid moiety S-[2,3-bis(palmitoyloxy)propyl]cysteine, (Pam2Cys), which corresponds to the lipid component of macrophage-activating lipopeptide 2 (MALP-2) from Mycoplasma fermentans7, are efficacious in the clearance of viral, bacterial and tumour agents in mouse models of infection and disease 8. Their mode of action appears to rely on the ability of the lipid to interact with Toll-like receptor 2 (TLR2) on DC, which triggers DC maturation resulting in an increased efficiency of presentation of the peptide epitopes 8. A lipopeptide based on T cell epitopes from influenza virus, when delivered by the intranasal (i.n.) route in the absence of additional adjuvant, induced lung viral clearance after challenge of mice on day 28 post-priming, while the equivalent non-lipidated peptide was largely non-protective. In that study, pulmonary protection was accompanied by high levels of specific CD8+ T cells in the lungs of mice 7 days after lipopeptide priming but the fate of those cells beyond the primary effector phase of the response was not addressed. In the present study we examine the longevity of the CD8+ T cell response following i.n. delivery of the Pam2Cys-containing vaccine and determine the characteristics and location of the cells that are associated with long-term viral clearance.
Protective efficacy of i.n. peptide-based vaccines
Peptide vaccines were synthesised that comprised a Th epitope peptide and a CTL epitope peptide from influenza virus in a structure shown diagrammatically in Fig. 1A. The lipopeptide differed from the non-lipidated peptide only by the addition of Pam2Cys to an intervening lysine through two serine residues. The data in Fig. 1B show that the primary effector response, induced 7 days after i.n. vaccination of mice with the lipopeptide, could mediate an average of at least 99% reduction in the level of pulmonary viral clearance upon challenge with influenza virus A/Memphis/1/71 x A/Bellamy/42 (Mem71) (p=0.008). In contrast, mice vaccinated with the non-lipidated vaccine showed significant (p=0.008) but much reduced levels of clearance when challenged 7 days later. Furthermore, in the lipopeptide-vaccinated mice, the viral clearing immunity persisted at a reduced (to 88%) but highly significant level (p=0.008) for at least 3 months, whereas in mice primed with the non-lipidated peptide the clearing response decayed at a very rapid rate, such that mice challenged at 3 months showed no benefit (p=0.15) from the vaccination (Fig. 1B). Therefore, both vaccines were able to induce viral clearing responses but these varied dramatically in magnitude or quality or both. As such, a comparison of the effector and memory populations elicited by the two vaccines may be able to reveal parameters critical for pulmonary protection, particularly in the memory phase of the response, where the differences between the vaccines are most pronounced.
The peptide epitopes that form the core of these vaccines do not elicit any anti-viral antibodies and we have previously determined by depletion studies that the CD4+ T cells induced in response to lipopeptide vaccines do not participate in viral clearance 9. The pulmonary protection we observe is therefore largely, if not completely, dependent upon the activity of CD8+ T cells directed towards the highly conserved epitope within the nucleoprotein.
Distribution of vaccine-induced CTL epitope-specific T cells
To determine whether the difference in the efficacy of the two vaccines is due merely to a difference in the numbers of CD8+ T cells induced, we enumerated the CTL epitope-specific cells in different locations by staining with an H-2KdNP147–155 tetramer (Table 1). An example of the data used to construct this table is displayed in Fig. 2, which shows the results for the recall response.
|Organ||Primingb)||Acute responsec)||Memory responsed)||Recalled responsee)|
|Lung||Mem71 virus||3.2 × 104 (26.0)||1.6 × 104 (36.8)||1.6 × 104 (23.6)|
|lipopeptide||1.7 × 104 (53.1)||3.5 × 104 (56.0)||4.8 × 104 (15.8)|
|peptide||6.7 × 103 (61.5)||6.7 × 103 (54.0)||3.8 × 103 (48.4)|
|PBS||naf)||na||4.2 × 103 (81.0)|
|MLN||Mem71 virus||3.9 × 104 (65.6)||-g)||3.9 × 103 (28.0)|
|lipopeptide||2.5 × 103 (69.5)||-||5.6 × 104 (43.8)|
|peptide||6.8 × 102 (75.9)||-||5.2 × 103 (83.6)|
|PBS||na||na||3.1 × 103 (56.6)|
|SPLEEN||Mem71 virus||1.2 × 105 (83.1)||1.2 × 105 (74.7)||8.7 × 104 (59.7)|
|lipopeptide||8.3 × 104 (80.1)||1.2 × 105 (71.9)||1.5 × 105 (67.8)|
|peptide||6.5 × 104 (74.7)||9.8 × 104 (62.0)||5.4 × 104 (70.4)|
|PBS||na||na||5.6 × 104 (63.2)|
On day 7 post vaccination (Table 1), only slightly fewer CD8+Tet+ cells were found in the lungs, draining mediastinal lymph nodes (MLN) and spleen of non-lipidated peptide vaccinated mice compared to lipopeptide-vaccinated mice; the greatest difference of 3.6-fold being observed in the MLN. By 28 days post vaccination, when the MLN had contracted in size and no longer yielded sufficient cells for analysis, CD8+Tet+ cells were still present in the lungs and spleens of both vaccine groups. At this point, the lipopeptide-induced CD8+Tet+ cells dominated in the lungs, being over 5-fold more numerous than those induced by the non-lipidated peptide, and, somewhat surprisingly, similar in number to the CD8+Tet+ cells induced by virus infection. At both days 7 and 28 in all sites, the CD8+Tet+ cells were heterogeneous in their expression of the lymph node homing marker CD62L, which is down-regulated on activated effector cells (CD62Llo phenotype) 10, but as expected, the CD62Llo population was more frequent in the lungs than in the lymphoid organs (Table 1).
At 3 months post immunisation, similarly vaccinated or infected mice were challenged with Mem71 virus and the CD8+ T cells that were recalled from the memory phase 5 days later were investigated (Table 1, Fig. 2). A control group of mice initially inoculated with PBS and then challenged in parallel with the vaccinated or infected mice was also included to reveal the effects of the viral challenge itself. This challenge control group had relatively low levels of specific CD8+ T cells, most of which expressed the CD62Lhi phenotype. The group of mice previously infected with Mem71 virus had completely cleared the homologous challenge virus from their lungs within 5 days (data not shown), presumably through the action of strain-specific neutralising antibodies to the viral hemagglutinin 11. There was little, if any, expansion in the numbers of specific CD8+ T cells in the MLN and spleen of these mice compared to the challenge control group, presumably due to the low viral load, and the numbers in the lung were the same as those established on day 28. In contrast to the challenge control mice, however, at least 70% of the CD8+ T cells in the lungs and MLN of mice previously exposed to virus had switched to the CD62Llo effector phenotype on secondary exposure to virus.
In lipopeptide-primed mice, relatively high levels of specific CD8+ T cells were observed in the MLN 5 days after challenge and over half of these were CD62Llo. In the lungs, the numbers of CD8+ T cells were only slightly increased above the day 28 levels but the vast majority of these had down regulated CD62L expression. In contrast to lipopeptide-primed mice, mice primed with the ineffective non-lipidated vaccine prior to challenge showed very similar numbers of specific CD8+ T cells to those from the challenge control group for all three organs. This led to an overall 13-fold deficit in the numbers of specific CD8+ T cells in the lungs of non-lipidated peptide-primed mice compared to lipopeptide-primed mice and a lesser percentage of these had the activated phenotype.
Significant disparity in the numbers of functional CD8+ T cells induced by the vaccines
Cells from the same groups of mice examined above for tetramer staining were also examined for their capacity to acquire lytic function after in vitro stimulation (Fig. 3A) and to produce IFN-γ in response to overnight incubation with the CTL peptide NP147–155 (Fig. 3B). The results revealed a hierarchy of functionality, with the CD8+ T cells induced by virus infection being the most active followed by those induced by lipopeptide vaccination. Overall, the numbers of IFN-γ-producing cells on days 7 and 28 in the different organs reflected the numbers of specific CD8+ T cells detected by tetramer staining. However, there were some exceptions, most notable being the large pool of non-lipidated peptide-induced CD8+ T cells that remained in the spleen on day 28 post vaccination (Table 1) but had little functional capacity (Fig. 3A, B). In the early memory phase examined on day 28, the lipopeptide-induced CD8+ T cells with IFN-γ-secreting capacity that remained after the contraction phase were as numerous as those induced by virus infection (p=0.07), despite the fact that they were under-represented on day 7. This indicated that the priming of CD8+ T cells by lipopeptide vaccination favoured the induction of a stable memory population.
In mice challenged 3 months after initial vaccination (Fig. 3B), IFN-γ-secreting CD8+ T cells were undetectable 5 days later in the challenge control group (equivalent to a primary response to the virus). However, in mice previously vaccinated with lipopeptide, a massive expansion or recruitment of functional cells was observed in both the MLN (note the difference in scale compared to day 7) (p=0.017) and to a lesser extent in the spleen (p<0.0001), while those induced by non-lipopeptide-priming were barely detectable except for a small population in the spleen. Of interest is the fact that at 5 days post challenge, when several logs of virus have already been cleared from the lungs of lipopeptide-primed animals (Fig. 1), the IFN-γ-secreting CD8+ T cell population at the site of infection is not very different in size to the day 7 (p=0.67) and day 28 (p=0.088) levels. This indicates that the expansion of the effector population occurring elsewhere in the animal may have little effect on initial viral clearance, and that control of infection is greatly reliant on the specific memory population in the lung.
Long-term pulmonary protection induced by i.n. lipopeptide vaccination
To examine the longer-term memory responses, CD8+ T cells were also measured 5 days after viral challenge at 9 months post vaccination with lipopeptide or non-lipidated peptide (Fig. 4). In this experiment, the magnitude of the anti-influenza response induced by the vaccines was compared to that induced by priming with an influenza virus (PR8), which has a haemagglutinin of a different subtype to the challenge virus (Mem71) and, therefore, any clearance observed will be independent of neutralising antibody. Despite the fact that the virus grew to an approximately tenfold higher titre in lungs of these older mice, a single dose of lipopeptide 9 months previously was sufficient to induce a strong viral clearing response (p=0.008), again far superior to the non-lipidated peptide (p=0.008) which was non-protective (p=0.84) (Fig. 3A). The protective efficacy was mirrored by the capacity of the cell populations to gain lytic function after in vitro incubation with virus, as shown here for spleen-derived cells (Fig. 4B). Cells induced after Mem71 challenge of mice having prior exposure to PR8 virus or to the lipopeptide had a high cytolytic capacity, while those from non-lipidated peptide-primed mice were similar to cells induced by challenge of naïve control mice. Of interest is the fact that the clearance is associated with a very small number of specific IFN-γ-producing CD8+ T cells in the lungs of the lipopeptide-vaccinated mice at this time point. At 5 days after challenge, the cells were barely detectable by ELISPOT assay (mean of 110 ± 28 cells/lungs of lipopeptide-primed mice) but could be enumerated by the more sensitive intracellular cytokine assay (Fig. 4C), which revealed similar numbers of cells in the lipopeptide-primed and heterologous virus-primed animals.
This study has revealed that a single dose of lipopeptide, which greatly decreased the levels of virus in the lungs 5 days after infection, induces a different CD8+ T cell response than does vaccination with the unprotective non-lipidated vaccine. In particular, the lipopeptide induces a lung-resident memory CD8+ T cell population, which can be implicated as an important mediator of viral clearance after challenge, at least in the initial stages, because (i) the pulmonary memory CD8+ T cells were fivefold more numerous than those induced by the unprotective vaccine, whereas both vaccines induced similar numbers in the spleen, (ii) they rapidly gained the activated CD62Llo phenotype after exposure to virus, unlike those induced by the unprotective vaccine, and (iii) despite massive expansion in the number of lipopeptide-induced specific CD8+ T cells with effector function in the MLN and spleen 5 days post infection, the numbers of effector CD8+ T cells in the lung remained low, suggesting that significant recruitment of cells from the lymphoid tissues to the site of infection had not yet taken place, even though significant viral clearance had already occurred.
These lung-resident CD8+ memory T cells induced after lipopeptide vaccination were very similar in number and activation status to those induced by virus infection itself; they had a similar potential to gain lytic function after in vitro stimulation and contained similar numbers of the IFN-γ-producing subset. Lung tissue CD8+ T cells induced by influenza infection came under scrutiny after it was noted that this population had different characteristics to the CD8+ T cells in other sites affected by the infection 12. These virus-induced specific T cells remained in the peribronchiolar and perivascular interstitial spaces of the lung long after the virus infection had resolved 2. Here they existed as an effector memory population, which continued to express a higher frequency of CD69, normally a marker of recent activation, and a lower frequency of CD62L than did the specific memory CD8+ T cells in the spleen 2, 13. The point has been made that these lung-resident cells can theoretically confer a particular advantage on the host as they are already present at the appropriate site of viral invasion and can, therefore, counter an infection in its initial stages before the viral load becomes too great and before additional effectors are recruited from other sites 2, 14.
It has been shown that the lung-resident memory CD8+ T cell population established 1 month after primary influenza infection are increased approximately tenfold 1 month following a second heterologous infection 2. The ability to boost this population is an important consideration for vaccination. It has been shown that if the overall pool of cross-reactive memory CD8+ T cells is sufficiently large, substantial control of the replication of a highly virulent influenza virus can be achieved within three days post challenge, leading to significant pulmonary protection against damage 15. In that study, high levels of influenza-specific memory CD8+ T cells were achieved by sequential infection with heterologous influenza viruses. In the absence of a human-compatible live attenuated influenza vaccine that replicates within the lung to prime or boost CD8+ T cell immunity 16, it will be necessary to devise alternate strategies to achieve this goal. The results obtained with the i.n. lipopeptide vaccine used in the present study suggest that vaccines for humans, which build on this same principle, may provide a safe and effective means of regularly boosting CD8+ T cell immunity including the important lung-resident population.
Whereas the protective efficacy of the non-lipidated peptide decayed very rapidly, benefit from lipopeptide vaccination was still observed as much as 9 months post vaccination, when lung CD8+ T cells after challenge were quite low. Even at this late time point, the numbers were comparable to those induced by prior exposure to live virus and, although the pulmonary protection was not as complete as that observed in virus-exposed mice, it should be remembered that virus infection also induces heterosubtypic immunity in the respiratory tract that is CD8+ T cell independent 1, which may further control the infection.
In regard to the longevity of the CD8+ T cell response, the role of CD4+ T cells for optimal CD8+ T cell memory in influenza has been well documented 17, 18. The lipopeptide incorporates an epitope for the co-induction of CD4+ T cells but clearly the lipid component is also critical for the establishment of the lung-resident memory population examined here because the corresponding vaccine without the lipid did not induce an equivalent population. CD4+ T cells fulfil their role in providing help for the establishment of CD8+ T cell memory during initial antigen priming 19–21, and it is likely that the lipid also has a role at this early phase of immune induction through its ability to mature the priming DC in a TLR2-dependent manner 8.
We report here that the i.n.-delivered Pam2Cys-containing vaccine provides all the necessary signals to DC for the priming of a CD8+ memory T cell population with the capacity to reside in the lung and mediate the initial clearance of pulmonary virus upon influenza infection. This study provides extremely valuable information for strategies for the control of respiratory pathogens and has particular relevance to pandemic influenza where a vaccine to boost appropriate cross-protective CD8+ T cell immunity, in the period prior to the manufacture of specific antibody-inducing vaccines, may lessen the severity of infection with a highly virulent emerging strain.
Materials and methods
Synthesis and assembly of lipidated and non-lipidated vaccines
Synthetic peptide-based immunogens were synthesised and purified as described previously 22. The vaccines consisted of a Th epitope peptide synthesised contiguously with and N-terminally to a CTL epitope peptide. The I-Ed-restricted Th epitope, sequence ALNNRFQIKGVELKS, was derived from the light chain of the influenza haemagglutinin and elicits CD4+ T cells that are cross-reactive with all H3 influenza viruses 23. The H-2Kd-restricted CTL epitope NP147–155, sequence TYQRTRALV, was derived from the nucleoprotein of the virus and is common to all type A influenza strains 24, 25. The Th peptide and the CTL peptide were separated in sequence by a single lysine residue. The lipid moiety Pam2Cys was attached to the intervening lysine through two serine residues 22 to yield the lipopeptide as described in 8 and shown in Fig. 1A.
The type A influenza viruses used in this study were an H3N1 subtype virus referred to as Mem71, derived by genetic reassortment of A/Memphis/1/71 (H3N2) x A/Bellamy/42 (H1N1), and the H1N1 virus A/Puerto Rico/8/34 (PR8). Virus stocks were grown for 2 days in the allantoic cavity of 10-day embryonated hen's eggs. Allantoic fluid containing virus was stored at –70°C. Infectious virus titres were determined by a plaque assay using monolayers of Madin Darby canine kidney (MDCK) cells 26.
Immunization and viral infection
Female BALB/c mice, bred at the University of Melbourne, were used at 6 weeks of age. After penthrane anaesthesia, either 10 or 45 nmol peptide immunogen resuspended in 50 μl PBS, or 104.5 PFU of Mem71 virus in 50 μl PBS, was delivered to the mice by the i.n. route. Control mice received PBS alone. For the challenge experiments, mice given a single dose of immunogen were infected i.n. with 104.5 PFU of Mem71 virus at 7 days, or at 3 or 9 months after immunization. Heterologous challenge experiments involved i.n. infection with 50 PFU of the PR8 virus 9 months prior to a challenge with 104.5 PFU Mem71 virus. The research complies with the University of Melbourne's Animal Experimentation Ethics guidelines and policies.
Determination of viral titres
Lungs taken from mice 5 days after challenge with Mem71 virus were homogenised, and the virus-containing supernatant, remaining above the cell debris following centrifugation, was harvested and stored at –70°C 9. Titres of infectious virus in the lung supernatants were determined by plaque assay on monolayers of MDCK cells.
Cell culture medium
T cell culture medium consisted of RPMI 1640 without glutamine (CSL Ltd., Parkville, Australia) supplemented with 10% heat-inactivated foetal calf serum (JRHBioscience), 2 mM L-glutamine, 2 mM sodium pyruvate, 30 µg/mL gentamicin, 100 µg/mL streptomycin, 100 IU/mL penicillin and 10–4 M 2-mercaptoethanol.
51Cr-release cytotoxicity assay
Secondary effector cells were generated either from spleens, lungs or mediastinal lymph nodes of mice that had been immunized i.n with either peptide immunogens resuspended in PBS or infected with 104.5 PFU Mem71. Briefly, 4 × 107 cells, depleted of erythrocytes by treatment with Tris-buffered ammonium chloride (ATC; 0.15 M NH4Cl in 17 mM Tris-HCl at pH 7.2) were cultured with 1 × 107 virus-infected then irradiated (2200 rad) syngeneic spleen cells in 25-cm2 tissue culture flasks (Falcon) containing 15 mL T cell culture medium. The virus-infected spleen cells had been preincubated at 37°C for 30 min with 3000 HAU of Mem71 virus in 1 mL serum-free RPMI 1640, and washed once prior to addition to the flask. After 5 days of culture at 37°C in a humidified atmosphere containing 5% CO2, cells were washed three times and used in 51Cr-release assays. The 51Cr-release assays were performed as described 27 using P815 mastocytoma cells (H-2d, DBA/2) as targets. Virus-infected and uninfected targets were prepared incubating 2 × 106 P815 cells in 250 μL of infectious solution (5000 HAU/mL) or serum-free RPMI, respectively. After 1-h incubation at 37°C, the cells were washed once and resuspended in 200 μL of T cell medium containing 200 μCi 51Cr (Amersham). After 2-h incubation at 37°C, the cells were washed three times and their concentration adjusted to 105 cells/mL. Aliquots of target cells (100 μL) were then dispensed into 96-well U-bottom tissue culture plates, and 100 μL aliquots of effector cells, at various effector to target (E:T) cell ratios were added.
After the target and effector cells had been incubated together at 37°C for 4 h in 5% CO2, 100 µL supernatant were removed from each well and the amount of radioactivity determined. The specific 51Cr release at each E:T ratio is derived from the radioactivity (in cpm) of the test sample following subtraction of the cpm released spontaneously in wells containing target cells incubated with medium only. These values are then expressed as a percentage of the maximal releasable counts derived from the cpm in samples from wells in which the target cells are incubated in Triton X-100, minus the spontaneously released cpm. Spontaneous release usually ranged from 1% to 10% of the maximum releasable counts. Data are presented as the mean and SD of the values obtained from triplicate cultures.
IFN-γ ELISPOT assay
IFN-γ-producing CD8+ T cells specific for the peptide TYQRTRALV were determined by an ELISPOT assay as previously described 9. Briefly, flat-bottom 96-well plates (Dynatech, Australia) were coated overnight with 5 µg/mL rat anti-mouse IFN-γ antibody (clone R4–6A2) in 50 µL PBS. Unoccupied sites on the wells were then blocked by incubation for 1 h with 10 mg/mL BSA in PBS. Twofold dilutions of spleen, lung or lymph node cells in T cell medium were then added to the wells together with 5 × 105 irradiated (2200 rad, 60Co source) syngeneic spleen cells from unimmunised mice and 10 U/well recombinant human IL-2 (Pharmingen, San Diego, CA). Cells were incubated at 37°C in 5% CO2 for 18 h in the presence or absence of the CTL peptide at a concentration of 1 µg peptide/mL, then lysed and removed with distilled water. The plates were incubated with biotinylated anti-mouse IFN-γ antibody (clone XMG 1.2, Pharmingen), washed, incubated with streptavidin-alkaline phosphatase (Pharmingen), and subsequently with ELISPOT substrate as previously described 9. When blue-green spots had developed, the plates were washed with water, dried and the spots counted using an inverted microscope.
Intracellular cytokine production assay for IFN-γ
Cells obtained from lungs, MLN and spleens of either immunised or immunised/challenged mice were cultured with or without 10–6 M of the NP147–155 (TYQRTRALV) peptide for 5 h in the presence of brefeldin A (5 μg/mL) and human recombinant IL-2 (10 U/mL) in 96-well round-bottom plates (Costar, Corning, NY). Lymphocytes were then washed and stained with phycoerythrin (PE)-conjugated rat anti-mouse CD8α antibody (BD Biosciences Pharmingen, San Diego, CA) for 30 min on ice, followed by two washes. Cells were fixed with 1% formaldehyde in PBS for 20 min and permeabilised in PBS/0.5% saponin for 10 min on ice. After two further washes lymphocytes were stained with anti-IFN-γ antibody conjugated to fluorescein isothiocyanate (FITC) (BD Biosciences Pharmingen) for 30 min at 4°C. Cells were resuspended in FACS buffer (PBS containing 1% BSA and 0.02% sodium azide). Data was acquired on a Becton Dickinson FACSCalibur flow cytometer and analysed using FloJo software. The total numbers of cells producing IFN-γ were calculated from the percentage CD8+IFN-γ+ cells in the samples determined by flow cytometry and the numbers of viable cells in the sample determined microscopically.
Tetramer staining of CTL epitope-specific CD8+ T cells
NP147–155-specific CD8+ T cells were identified by staining with tetrameric complexes of H-2Kd molecules presenting the CTL epitope peptide TYQRTRALV (provided by Drs W. Xie and J. Lin, St Jude Children's Research Hospital, Memphis, TN), conjugated to streptavidin-PE (Molecular probes, Eugene, OR). Staining was for 60 min at room temperature, followed by two washes in FACS buffer. Cells were then stained with both anti-CD8 mAb conjugated to FITC and anti-CD62L mAb conjugated to allophycocyanin (Pharmingen) for 30 min on ice, followed by two washes. Cells were resuspended in FACS buffer and analysed by flow cytometry 28.
The viral clearance data were analysed using the non-parametric Mann-Whitney test, calculated using Prism software. Viral titres below the limits of detection of the assay were assigned a log value of 2.5. The cytokine production data were analysed using the two-tailed Student's t-test, also calculated using Prism software. The resultant p value for particular comparisons is given.
This work was supported by grants from the National Health and Medical Research Council of Australia and the Australian Government's Cooperative Research Centres Program. K.K. is a Peter Doherty National Health and Medical Research Council Postdoctoral Fellow. We wish to thank Prof Peter Doherty for his input to the project.