Vaccine—screening method

Abstract

The present invention provides screening methods which may be regarded as in vitro or ex vivo methods of interrogating the immune system to understand what viral antigens are “seen” and responded to by T cells of the immune systems during viral infection. The screening methods further link in vitro or ex vivo responses to progression of infection in subjects.

Claims

1. A method for screening a plurality of test peptides having a level of identity with a sequence of a protein of a virus, to identify any peptide or peptides which are capable of ameliorating infection by the virus in a human, the method comprising: a) obtaining T0 samples comprising T cells from at least five different human test subjects who are seronegative for antibodies to the virus, where each of the T0 samples is obtained from a single human test subject, b) contacting the T cells in each of the T0 samples with the plurality of test peptides, quantifying the level of response of the T cells in each of the T0 samples to each of the test peptides, and identifying any peptide or peptides which induce a T cell response in one or more of the T0 samples, c) inoculating and infecting each one of the at least five human test subjects with the virus after obtaining the T0 sample from the subject in step a), d) obtaining a measure of disease severity in each one of the at least five human test subjects following infection with the virus in step c); e) for each peptide identified in step b), correlating the level of response to the peptide of the T cells in each T0 sample with the measure of disease severity in the test subject from whom the T0 sample was obtained and f) for any one peptide where an inverse correlation is found in step e), identifying the peptide as a peptide that is capable of ameliorating infection by the virus in a human subject.

2. The method of claim 1, wherein the subjects have not been vaccinated against the virus with a vaccine designed to elicit a T cell response.

3. The method of claim 1, wherein the measure of disease severity is i) a score of severity of symptoms of the viral infection experienced by the subjects, ii) a measure of viral shedding or viral load in the test subjects, iii) duration of illness experienced by the test subjects, iv) a measure of a biomarker typically increased during infection with the virus, or v) a measure of a biomarker typically decreased during infection with the virus.

4. The method of claim 1, wherein the peptide is about 7to about 25 amino acids long.

5. The method of claim 4, wherein the peptide is 9-25 amino acids or 10-20 amino acids and preferably 15-18 amino acids long.

6. The method of claim 1, wherein the peptide has at least 70% identity with a sequence of a viral protein.

7. The method of claim 6, wherein the peptide has at least 80% identity with a sequence of a viral protein, optionally wherein the peptide is identical a sequence of a viral protein.

8. The method of claim 1, wherein a library of peptides is screened.

9. The method of claim 8, wherein the library of peptides spans a protein of a viral proteome.

10. The method of claim 1, wherein the peptide is synthetic.

11. The method of claim 1, wherein the virus is capable of causing acute, self-limiting infection in humans.

12. The method of claim 1, wherein the virus is a respiratory virus, an enteric virus or a mucosal virus, or a virus that infects by the mucosal route.

13. The method of claim 1, wherein the virus is an enteric virus, optionally the virus is norovirus.

14. The method of claim 1, wherein the virus a) is a virus from the Orthomyxoviridae, Paramyxoviridae, Coronaviridae, Picornaviridae, Caliciviridae, Flaviviridae or Reoviridae families, or b) is a respiratory virus, optionally the virus is selected from an influenza virus, rhinovirus or respiratory syncytial virus, human metapneumovirus, adenovirus, coronavirus, boca virus or other acute respiratory virus, preferably the virus is selected from an influenza virus, rhinovirus or respiratory syncytial virus.

15. The method of claim 14, wherein the virus is an influenza virus, optionally an influenza A virus.

16. The method of claim 9, wherein the library of peptides spans the protein nucleoprotein (NP) and/or matrix (M1 or M2) protein of influenza virus.

17. The method of claim 1, wherein CD4+T cell responses are quantified.

18. The method of claim 1, wherein the test samples comprising T cells are obtained from more than 10, 15 or 20 different subjects.

19. The method of claim 1, wherein the peptide is a polypeptide, an oligopeptide or a protein.

20. The method of claim 1, further comprising the step, prior to collecting the T0 test samples, of placing the at least five human test subjects into controlled conditions in isolation.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows an Elispot layout of experimental influenza A infection in humans. (a) H3N2 challenge study (A/Wisconsin/67/05) T cell Elispot layout. (b) H1N1 challenge study (A/Brisbane/59/07) T cell Elispot layout. Freshly isolated 300,000 PBMC were put into each well and stimulated with peptide pool at 2 μg/ml for 18-24 hours.

(2) FIG. 2 shows viral shedding in nasal wash, seroconversion and symptom development in seronegative healthy volunteers experimentally infected with influenza A. (a) Volunteers infected with cell grown H3N2 (WS/67/05) virus or egg-grown H1N1 (BR/59/07) virus. The virus in the nasal sample was titrated by TCID.sub.50 assay. Presence of flu-specific antibody was measured by haemagglutination inhibition assay. (b) Correlation of total symptom scores against the peak nasal virus shedding in H3N2 infected subject by Spearman rank correlation test. (c) Mean symptom scores and oral temperatures of volunteers infected with H3N2 virus. (d) Mean symptom scores and oral temperatures of volunteers infected with H1N1 virus. Symptom assessments were performed by the volunteers twice daily on a four-point scale (absent to severe). The score for each symptom group was obtained by adding the total individual symptom scores for that particular group on that particular day. Oral temperatures were determined four times a day for the duration of the study and the highest temperature was represented.

(3) FIG. 3 shows symptom scores in each infected volunteer infected with influenza A. (a) Total symptom scores (y axis) in volunteers infected with H3N2 (WS/67/05) (x axis showing days after H3N2 challenge infection). (b) Upper respiratory symptom scores (y axis) in volunteers infected with H3N2 (WS/67/05) (x axis showing days after H3N2 challenge infection). (c) Lower respiratory symptom scores (y axis) in volunteers infected with H3N2 (WS/67/05) (x axis showing days after H3N2 challenge infection). (d) Systemic symptom scores (y axis) in volunteers infected with H3N2 (WS/67/05) (x axis showing days after H3N2 challenge infection). (e) Total symptom scores (y axis) in volunteers infected with H1N1 (BR/59/07) virus (x axis showing days after HiN1 challenge infection). (f) Upper respiratory symptom scores (y axis) in volunteers infected with H1N1 (BR/59/07) virus (x axis showing days after H1N1 challenge infection). (g) Lower respiratory symptom scores (y axis) in volunteers infected with H1N1 (BR/59/07) virus (x axis showing days after H1N1 challenge infection). (h) Systemic symptom scores (y axis) in volunteers infected with H1N1 (BR/59/07) virus (x axis showing days after H1N1 challenge infection).

(4) FIG. 4 shows T cell responses in seronegative healthy volunteers experimentally infected with influenza A virus. Flu-specific T lymphocyte responses were measured from freshly isolated PBMC ex vivo from each volunteer by IFN-γ release after stimulation with corresponding peptide pools spanning the entire challenge influenza proteome. Each bar represented the total T cell responses to entire influenza proteome and each colour box represented the response to each protein. X axes denote subject number.

(5) FIG. 5 shows antibody and T cell responses in seronegative healthy volunteers experimentally infected with influenza A virus. (a) Presence of flu-specific antibody was measured by haemagglutination inhibition assay. (b) Plot of proportion of infected subjects demonstrating positive T cell responses to NP and M flu proteins at baseline. The Y-axis represents the proportion (%) of subjects from both challenge studies with positive response to NP and M proteins and their CD4 and CD8 dependency. (c) Activated and proliferating cells (CD38+Ki67+) could be detected ex vivo in H1N1-infected subjects by flow cytometry (d) Changes in the proportion of activated and proliferating (CD38+Ki67+) T cells in both CD4 and CD8 population of freshly isolated PBMC from volunteers infected with H1N1 virus by flow cytometry. (e) Correlation between proportion of CD38.sup.+Ki67.sup.+ T cells on day 7 of H1N1 infected subjects with their magnitude of Elispot response by Spearman rank correlation test.

(6) FIG. 6 shows correlations between flu-specific total T and CD4 T cell responses to internal proteins and measure of influenza severity (viral shedding, symptom severity or illness duration) in volunteers infected with (a) H3N2 (WS/67/05) or (b) H1N1 (BR/59/07). Correlations between total symptom scores or length of illness duration against flu-specific total T cell responses or CD4 flu-specific T cells specific to internal proteins including nucleoprotein and matrix of challenge virus. All tests were run by spearman rank correlation test.

(7) FIG. 7 shows phenotypic and functional studies of CD4 and CD8 cells at baseline and day 7. (a) Expression of IFNγ and CD107a in memory T cells of a representative H3N2 infected subject after stimulation with peptide pools to influenza proteins. PBMC from baseline and day 7 samples were stimulated with different peptide pools (Flu, NP, M) for 6 hours and the ex vivo response was measured by FACS staining. Both memory CD4 and CD8 responses in the same sample were measured. Staphylococcus enterotoxin B (SEB) was used as positive control. (b) Killing function of CD4+ T cell lines from the same baseline sample upon recognition of autologous target cells pulsed with peptides was measured by chromium Release Assay. Perforin-dependent cytotoxicity was measured by sensitivity to concanamycin.

(8) FIG. 8 a) Human parenchymal (i) and (ii) and bronchial tissue stained (iii and iv) for MHC II (HLA-DR) 2 mm sequentially cut sections and immunostained using i subtype control monoclonal antibodies (i) and (iii) or antibodies specific for HLA-DR (ii) and (iv) at the same concentration. Signal was amplified using the ABC system, and colour developed using DAB stain. Specific staining is shown in brown, haematoxylin counterstain is shown in blue. Size bar represents 50 um. b) (i) Representative histograms showing specific staining of HLA-DR expression on primary bronchial epithelial cells (PBECs) by flow cytometry using cells incubated in the presence or absence of HLA-DR APCCy7 or IgG2a APCCy7 (isotype). (ii) Graph of mean fluorescence intensity of HLA-DR expression on PBECs using flow cytometry. NT —non treated control, X31 influenza infected cells, UVX31—UV inactivated viral controL HLA-DR is constitutively expressed on primary respiratory epithelial cells, there is a small rise in expression following in vitro infection of these cells with influenza virus which was significant in comparison to stimulation with UV-treated (inactivated) virus. This confirms that respiratory epithelial cells are potential target cells for cytotoxic CD4+ T cells.

(9) IDENTIFICATION OF PEPTIDES FOR FLU VACCINE AS VALIDATION OF SCREENING METHODS

(10) This study was designed to allow several effects to be demonstrated, including the following points.

(11) Firstly, the study confirmed that pre-existing cell mediated immunity to a virus (in other words the existence of memory T cells within a subject, which respond to peptide antigens of a virus) results in a reduction in disease symptoms, duration of disease and viral shedding, when the subject is infected with that virus.

(12) Secondly, the cognate peptide antigens for the pre-existing memory T cells were identified. These peptides can be used to induce T cell immunity to the virus.

(13) Thirdly, the study demonstrates the effectiveness of the screening method of the invention for identifying peptides which correspond to antigens inducing T cell responses during immune response to viral infection. These peptides can be used to induce T cell immunity in a virus.

(14) Materials and Methods

(15) Study Design

(16) Between October, 2008 and October, 2009, two separate prospective, randomised, and double blinded, parallel group clinical studies of experimental human influenza A infections were undertaken in a single site in Cambridge, UK. The two studies were carried out 9 months apart. An H3N2 challenge study was carried out between 24 Oct. and 24 Nov., 2008 whereas H1N challenge study was carried out between 18 Aug. and 18 Sep., 2009. Healthy, non-pregnant adults between the ages 18 and 45 were eligible for the enrolment. Exclusion criteria included health care workers, history of acute respiratory illness, chronic illness or medications. In H3N2 challenge study, a total of 17 healthy adult volunteers, which are haemagglutination-inhibition (HI) titres less than 1:8 to influenza A/Wisconsin/67/05, were enrolled in the study. Whereas, in H1N1 challenge study, a total of 24 healthy adult volunteers with HI titres less than 1:8 to influenza A/Brisbane/59/07 were enrolled in the study.

(17) Both studies were conducted in compliance with Good Clinical Practice guidelines (CPMP/ICH/135/95) and declaration of Helsinki. The protocols were approved by East London and City and the Southampton and Southwest Hampshire ethics review committees. Written informed consent was obtained from each participant with an ethics committee approved form. No medications, except acetaminophen for treatment of severe symptoms, were permitted. Subjects were compensated for their participation of the study.

(18) Study Outline

(19) Screening assessments began within 45 days of the scheduled viral inoculation. Volunteers were confined to individual rooms in an isolation unit 2 days before the day of inoculation, and remained in isolation for 7 days thereafter. Therefore, contact with any pathogens such as viruses or bacteria is completely controlled. Isolation and monitoring of subjects allows study of infection and symptoms of the infection. Inoculation occurs under clinical conditions so that the exact time of inoculation is known. Therefore samples obtained from the subject can be taken at known time points after inoculation.

(20) The subjects were randomised into 4 groups and each group of the participants were inoculated intra-nasally with different doses of influenza A virus on day 0. The dose of the virus was designated as 1:10 (high), 1:100 (medium-high), 1:1000 (medium-low) and 1:10,000 (low) from the original virus stock. Group 1 received high dose, Group 2 received medium-high dose, Group 3 received medium-low and Group 4 received low dose of virus. Nasopharyngeal swab were collected daily from baseline day 0 during the quarantine period for virus isolation. Serum samples were taken daily for serum cytokine and biomarker study. Fresh whole blood for cellular assays was taken on day −2 or 0, 7 and day 28. An additional time point day 3 was taken for H1N1 study.

(21) Oral temperatures were measured four times daily. Fever was defined as an oral temperature >37.7° C. Symptom assessments were performed by the volunteers twice daily on a four-point scale (0-3 corresponding to absent to severe) (Hayden et al., J. Clin. Invest. 101(3), 643-649, 1998). The symptoms assessed were nasal stuffiness, runny nose, sore throat, cough, sneezing, earache/pressure, breathing difficulty, muscle aches, fatigue, headache, feverish feeling, hoarseness, chest discomfort, and overall discomfort. The total symptom score for each day was obtained by adding the individual symptoms scores for that particular day including morning and evening sessions. The individual symptoms contributing to the total symptoms scores were divided into three subgroups: systemic symptoms (muscle aches, fatigue, headache, and fever), upper respiratory symptoms (nasal stiffness, ear ache/pressure. runny nose, sore throat, and sneezing) and lower respiratory symptoms (cough, breathing difficulty, hoarseness and chest discomfort).

(22) Viruses

(23) In both challenge studies, GMP grade viruses were manufactured and processed by GlaxoSmithKline, UK. The stock virus were diluted to four different inoculum titres and prepared in individual aliquots intended for single use and then administered. The titre of the stock virus was 10.sup.7 TCID.sub.50 infectious dose. They were ten-fold diluted and the titre were ranged from high titre (1:10), medium-high (1:100), medium-low titre (1:1,000) and low titre (1:10,000). Subjects were observed for potential allergic reactions for 30 min following inoculation. In H3N2 challenge study, tissue culture grown A/Wisconsin/67/05 virus was used. In H1N1 challenge study, egg grown A/Brisbane/59/2007 virus was used.

(24) Virus Titration by TCID.sub.50 (Tissue Culture Infectious Dose 50%) Assay

(25) Viral load in the nasopharyngeal samples were determined by TCID.sub.50 assay as described by the WHO manual of Animal Influenza: Serial ten-fold dilutions of virus-containing samples were inoculated into 96-well microtitre plates seeded with Madin-Darby canine kidney (MDCK) cells, and incubated for 5-6 days at 37.degree. C. Cytopathic effects in individual wells were determined via light microscopy. Titre greater than 1:5 was considered positive.

(26) Hemagglutination Inhibition (HI) Assay

(27) Hemagglutinin-specific antibody titers against H1N1 (A/Brisbane/59/2007) or H3N2 (A/Wisconsin/67/05) in the serum samples were determined by HI assay using chicken erythrocytes as described in WHO manual

(28) Synthetic Peptides

(29) 18-mer peptides overlapping by 10 amino acid residues and spanning the full proteome of the H1N1 and H3N2 influenza A viruses were designed using the Los Alamos National Library web-based software PeptGen and synthesized (purity >70%; PEPscreen; Sigma-Aldrich) using the sequences of the following strains: A/Brisbane/59/2004 (H1N1), A/New York 388/2005 (H3N2) (surface proteins), and A/New York 232/2004 (H3N2) (internal proteins). In H3N2 peptides, the amino acid sequence homology between challenge Wisconsin strain and New York strain was greater than 99%. The total numbers of peptides used in detecting antigen-specific responses for H1N1 and H3N2 were 554 and 601 respectively.

(30) Identifying Peptides “Seen” by T Cells of Immune System

(31) Ex vivo IFNγ ELISPOT assays were used to identity T cells which respond to stimulation with a specific peptide and therefore secrete IFNγ. In the each influenza Elispot assay, all overlapping peptides in each individual were simultaneously tested using 2-dimensional matrices with a total of 50 pools (1.sup.st D=25 pools; 2.sup.nd D=25 pools; up to 25 peptides/pool) so that each peptide was present in two different pools (see FIG. 1 for Elispot layout). Peptides were used at a final concentration of 2 μg/ml each. The putative peptide from each positive response well could be deconvoluted from a 2-dimensional matrix system where each peptide only appeared once in each dimension. The putative peptides were then confirmed individually in the second Elispot assay with the same input cell number per well.

(32) Ex Vivo IFN-γ ELISPOT Assay

(33) Peripheral mononuclear cells (PBMC) were separated from 50 ml heparinised blood by density gradient centrifugation using Lymphoprep (Axis-Shield, Norway) and Leucosep tube (Greiner, UK) within 3-6 hour upon each bleed (Li et al., J. Immunol. 181, 5490-5500 2008). To detect influenza-specific effector memory cells (CD45RO+), PBMC were added into 96-well Elispot Multiscreen plates (MAIPS4510, Millipore) at 300,000 cells/well and cultured with peptide pools for 18-24 h incubation at 37° C. and 5% CO.sub.2. The end concentration of each peptide in each well was 2 μg/ml, for both peptide pools and individual peptides. All ELISPOT assays were performed using the human IFN-γ ELISPOT kit (Mabtech) according to the manufacturer's instructions. The internal negative control was no peptide in quadriplicates, and positive controls were EC (a mixture of EBV and CMV T cell epitope peptides) or PHA (10 μg/ml). The spots on each well were counted using an automated ELISPOT reader and AID ELISPOT 3.1.1 HR software (Autoimmune Diagnostika). In pool responses, wells containing spot numbers greater than the mean+4 SD of three negative control wells (no peptide) were regarded as positives in each individual, provided that the total was greater than 50 spot forming cells (SFC)/million PBMC, to rule out false positives where background was very low. In all assays, values of no peptide control wells were 1.8±4.6 SFC/million PBMC for 150 healthy subjects and 2±5.7 SFC/million PBMC for 150 influenza-exposed subjects. Values of T cell responses were all background subtracted and presented in SFC/million PBMC. To determine whether T cells were CD4 or CD8, in the second ELISPOT assay, cell depletion was also conducted by Dynal CD8 beads, as described in the manufacturer's instructions (Invitrogen, UK), before the ELISPOT assay. Undepleted PBMC served as positive controls. For single peptide confirmation Elispot assay, response greater than 10 SFC/million PBMC was considered positive after background substraction and when T cell lines could be generated from respective peptides and tested positive again with ICS.

(34) Generation of Short-Term T Cell Lines

(35) Short-term T cell lines were generated to confirm influenza peptides and the CD4 or CD8.sup.+ property of each peptide by ICS and flow cytometry, as described previously (Li et al., J. Immunol. 181, 5490-5500, 2008). In brief, frozen samples of PBMC were thawed and rested for 2 h before stimulating with 10 μg/ml of each peptide at final concentration for 1 h. Cells were cultured in RPMI 1640 supplemented with 10% human serum (National Blood Services, UK) and 25 ng/ml IL-7 (PeproTech) for 3 days, and then 100 U of IL-2/ml (Proleukin, Novartis UK) was added every 3 to 4 days thereafter. On day 14, cells were washed three times with sterile PBS and then rested in fresh RAB-10 for 25 to 35 h at 37° C., 5% CO.sub.2.

(36) FACS Staining Assay

(37) Activated (CD38+) and proliferating (Ki67+) cells in freshly isolated PBMC were stained by were stained with mAbs against human Ki67-FITC (Clone B56, BD Biosciences), DR-PE (clone TU36, BD) CD38-APC (clone HB7, BD), CD4-pacific blue (Clone MT130, DakoCytomation), and CD8-PE-Cy5 (Clone SK1, BD). Cytotoxicity as measured by expression CD107a (clone H4A3, BD) and IFN-γ (clone XMG1. 2, BD) in both CD4 and CD8 memory cells were also studied ex vivo using frozen PBMC as described previously (Li et al., J. Immunol. 181, 5490-5500 2008). PBMC (1 million per stimulation) were stimulated with peptide pools for 6 hours in the presence of brefeldin A and monensin. For each stimulation condition, at least 500,000 total events were acquired using LSRII (BD immunocytometry Systems, San Jose, Calif.). Data analysis was performed using FlowJo (version 8.8.4; TreeStar, Ashland, Oreg.). Response greater than 3 times background was considered positive.

(38) Chromium Release Assay

(39) A standard .sup.51Cr release assay was used as described previously (McMichael A J et al., N Engl J Med 309, 13-17, 1983). T cell lines generated from PBMC samples were used as effector cells and their autologous EBV-transformed B cell lines were used as target cells. Inhibition of perforin-mediated cytotoxicity was obtained by incubating the CD4+ T cells for 2 h with 100 nM concanamycin (Sigma). Specific .sup.51Cr release was calculated from the following equation: ([experimental release-spontaneous release]/[maximum release-spontaneous release])×100%.

(40) Statistics

(41) All graphs were presented by GraphPad Prism (version 5) and statistical analysis was done by GraphPad Prism and SPSS. Magnitude of T cells response was presented by SFC/million PBMCV and breadth of T cell response was defined by the number of proteins recognized by each subject. To study the role of T cell in the virus shedding (viral control) and symptom development (immunopathology), correlation was run between pre-existing T cells and measures of infection and illness (virus titre, symptom assessments, temperature) by Spearman rank correlation analysis. Correlation analysis was based on data collected from all infected (culture positive and/or four fold or greater rise in HI antibody titre) individuals.

(42) Epithelial Cell MHC Class II Expression

(43) Immunohistochemistry

(44) Lung explants were harvested from lung tissue recovered from patients undergoing routine thoracic surgery under additional consent. Human parenchymal and bronchial tissue was fixed in acetone prior to embedding in GMA resin. Two millimetre sections were cut sequentially and immunostained using isotype control monoclonal antibodies or antibodies specific for MHC II (HLA-DR) at the same concentration. Signal was amplified using the ABC system, and colour developed using DAB stain. Specific staining is shown in brown, haematoxylin counterstain is shown in blue.

(45) Flow Cytometry

(46) Primary bronchial epithelial cells (PBECs) were obtained from subjects undergoing research bronchoscopies in the Wellcome Trust Clinical Research Facility at Southampton General Hospital. Bronchial brushings were cultured in Bronchial Epithelium Growth Media (BEGM), (Lonza, Wokingham, UK) in collagen coated flasks (PureCol™, Inamed Biomaterials, California, USA) and incubated in a humidified atmosphere at 37° C., 5% CO.sub.2. The collection and use of these samples was approved by the Southampton and South West Hampshire Research Ethics Committee (REC No: 06/Q1701/98 & 08/H0504/138).

(47) Influenza A virus strain X31 was supplied at a concentration of 4×10.sup.7 pfu/ml (a kind gift of 3VBiosciences). Inactivated virus (UVX31) was prepared by exposure to an ultra-violet (UV) light source for 2 h.

(48) PBECs were seeded at 1×10.sup.5 cells per well onto a collagen-coated 24 well plate and left at 37° C., 5% CO.sub.2 for 24 h. Cells were then growth media starved for 24 h in 0.5 ml Bronchial Epithelium Basal Media (BEBM) supplemented with 1 mg/ml BSA, insulin, transferrin and selenium (BEBM+ITS). Cells were incubated for 2 h with no virus, or 2×10.sup.3 pfu of X31 or UVX31. Cells were then washed three times with BEBM+ITS and incubated for a further 20 h at 37° C., 5% CO.sub.2 in 0.5 ml of BEBM-ITS. Cells were dispersed by trypsinisation and prepared for flow cytometric analysis as previously described.

(49) Samples were incubated on ice in the dark for 30 min with Allophycocyanin-Cyanine 7 (APC-Cy7)-conjugated anti-HLA-DR (BD Biosciences, Oxford, UK) or appropriate isotype control (IgG2a BD Biosciences Oxford, UK). After washing, intracellular staining for viral nucleoprotein (NP)-1, was performed using BD Cytofix/Cytoperm kit according to manufacturer's instructions, and AlexFluor 488 (AF488)-conjugated anti-NP-1 antibody (HB-65, a kind gift of 3VBiosciences). Flow cytometric analysis was performed on a FACSAria using FACSDiva software v5.0.3 (all BD).

(50) Results

(51) Human Influenza Infection Model

(52) In order to study the impact of existing cell mediated immunity (CMI) on influenza infection, an experimental infection model was established using live influenza A virus in human volunteers (Oxford J S et al Expert Rev Anti Infect Ther. 3, 1-2 (2005)). A total of 41 healthy volunteers aged between 19 and 41 were inoculated intra-nasally with serial 10 fold dilution of influenza A viruses; a cell grown H3N2 WS/67/05 and an egg grown H1N1 BR/59/07. Subjects were studied prospectively from inoculation in a clinical isolation facility with measures of viral shedding, symptom development, cellular and humoral immune responses for the first 7 days and again at day 28. These measures provide information on the severity of the infection, duration of infection and on the immune responses of the subject to the infection. In the H3N2 challenge study, 8 out of 17 (47%) volunteers were female and the median age was 26.5 yr (range 22-41) (Table 2). In H1N1 challenge study, 7 out of 24 (29%) were female and the median age was 24 yr (range 19-35).

(53) TABLE-US-00003 TABLE 2 Demography, virus shedding and antibody titre of the study groups H3N2 challenge group H1N1 challenge group Group 1 Group 2 Group 3 Group 4 Group 1 Group 2 Group 3 Group 4 N 4 4 4 5 6 6 6 6 Age-yr Mean ± SD   26 ± 3     28 ± 5     25 ± 2   29 ± 8     25 ± 5   25 ± 5     27 ± 4    23 ± 3   Median 26 27 25 28 24 24 26 22 Range 23-29 25-35 22-27 22-41 20-32 22-35 23-31 19-27 Sex-no. (%) Female 2(50) 2(50)  2(50)  2(40) 2(33) 1(17) 2(33) 2(33) Male 2(50) 2(50)  2(50)  3(60) 4(67) 5(83) 4(67) 4(67) Virus shedding no. (%) 1(25) 4(100) 2(50)  2(40) 1(17) 3(50) 1(17) 1(17) HAI titre on Day 28 Positive 3(75) 2(50)* 1(25)** 1(25) 2(33) 2(33) 2(33) 1(17) GMT 96 33 0.6 0 1 5 0 0 Mean symptom scores Mean ± SD 10.5 ± 19.7 60.8 ± 10.7 13.8 ± 14 4.6 ± 5.5 11.2 ± 17.7 39 ± 23.2 14.5 ± 14.1 8.3 ± 15.2 Median 1 57.5 7.7 1 11.5 44.5 14.5  0.5 Range  0-40 52-76  5-22 0-9  0-22  3-65  0-31  0-38 *One subject was unavailable for D28 visit. **Two subjects were unavailable for D28 visit.

(54) All volunteers selected were seronegative to the challenge strain and virus PCR negative in nasal lavage at the time of challenge. This confirmed that subjects were not currently infected with the challenge virus, or been infected recently with the challenge virus. The overall infection rate was defined by evidence of virus shedding and/or seroconversion by day 28. This was higher in subjects (14/17, 82%) challenged with H3N2 virus than subjects (9/24, 38%) challenged with H1N1 virus.

(55) In the H3N2 challenge group, virus shedding persisted in individuals for as long as 7 days but most subjects (8/14, 57%) cleared the virus completely by day 4. (FIG. 2a). The H1N1 challenge group, did not exhibit reliable viral shedding —a recognised phenomenon with this egg grown virus (Steel J, et al. J Virol. 2009 February; 83(4):1742-53).

(56) In the H3N2 challenge infection, total symptoms closely tracked peak viral load (FIG. 2b, r=0.6977, p=0.0055, Spearman coefficient). Similar symptom profiles were observed between the two challenge cohorts and were comparable to wild type infections in this population (Newton D W at al. Am J Manag Care. 6, 265-75 (2000)). In the H3N2 challenge group, 11 out of 14 (79%) infected subjects developed one or more symptoms, and as a group, exhibited symptom scores that peaked on day 3 and returned to normal by day 7 after viral inoculation (FIG. 2c). 3 out of 14 subjects (21%) developed fever (oral temperature>37.7° C.) and the highest temperatures were detected on day 2. In the H1N1 group, 8 out of 9 (89%) infected subjects developed one or more symptoms and showed mean symptom scores that peaked on day 4 and returned to normal by day 7 after viral inoculation (FIG. 2d). Also, 1 out of 9 infected subjects (11%) developed fever and the highest temperatures were detected on day 2.

(57) In both challenge groups, the total symptoms were dominated by upper respiratory illness as defined by the presence of symptoms such as runny nose and sore throat, occurred in 10/14 (71%) subjects in H3N2 group and 8/9 (89%) subjects in H1N1 group. Lower respiratory symptoms such as cough and hoarseness were much milder in severity and occurred in 3/14 (21%) in H3N2 group and 2/9 (22%) in H1N1 group. Scores for systemic symptoms such as muscle aches and fatigue were also present in 6/14 (43%) in H3N2 group and 5/9 (56%) in H1N1 group. For more details on the distribution of symptoms of each infected subject, see FIG. 3.

(58) Antibody and T Cell Responses of Infected Volunteers

(59) All volunteers enrolled were screened to ensure they were sero-negative for antibodies to the challenge virus. However, the antibody responses (HAI titre) were detectable after 7 days post challenge (FIG. 5a), at which time the viruses were completely cleared as indicated in FIG. 2a.

(60) Prior to viral challenge the nature of pre-existing T cell memory from previous infection exposure was determined. T cell responses to proteins expressed by the challenge virus were present in most volunteers in both studies prior to challenge despite the absence of detectable antibodies to the same strains. The size of total T cell responses was below 1000 SFC/million PBMC in all subjects studied at baseline (FIG. 4). At baseline, in the H3N2 group, 11 out of 14 (79%) infected subjects showed memory T cell responses recognizing one or more H3N2 proteins, with an average of two proteins recognized (range 1-5). The most immunodominant proteins were nucleoprotein (8/14, 57%) and matrix proteins (7/14, 50%), which are highly conserved across strains, based on the number of subjects and the magnitude of IFN-γ response. In the H1N1 challenge group, 8 out of 9 (89%) infected subjects showed memory T cell responses that recognized one or more proteins at the baseline, with an average number of one protein recognized (range 1-3). The most immunodominant protein was matrix protein (6/9, 67%). These results show that conserved viral peptide sequences from nucleoprotein and the matrix proteins are important in cell mediated immunity and T cell responses, and the present methods allow identification of these peptides.

(61) On day 7 after challenge infection, both the breadth and magnitude of memory T cell responses increased dramatically in the peripheral blood by an average of 10 fold in both study groups (FIG. 4). In the H3N2 group, 14 out of 14 (100%) of infected subjects demonstrated positive T cell responses with an average of five proteins recognized (range 1-8). Pre-existing T cell responses against each protein were expanded in addition to new responses that had not been detected at baseline. In the H1N1 challenge group, 9 out of 9 (100%) infected subjects were T cell positive responding to an average of five proteins (range 2-7). No significant changes in the T cell responses against known CD8 epitopes of CMV and EBV were found in control wells, suggesting bystander activation was minimal (data not shown). Therefore the body dramatically responds to viral peptides during infection raising T cell responses and the present methods allow identification of those peptides.

(62) On day 28, the total memory T cell response had returned to baseline levels (<1000 SFC/million PBMC) in both challenge groups. Immunodominant protein responses such as NP and M persisted at a baseline level whereas most newly generated responses against other proteins had vanished after the acute phase of infection. In the H3N2 challenge group, 7 out of 14 infected subjects (50%) were T cell positive, with the average number of proteins recognized reduced to 1 (range 1-2). In the H1N1 challenge group, 8 out of 9 infected subjects (89%) were T cell positive, with the average number of proteins recognized reduced to 2 (range 2-4). However, 4 out of 9 (44%) newly generated HA responses persisted at lower levels (average 60 SFC/million PBMC). In addition, epitope mapping and whether they were mediated by CD4 or CD8 T cells was determined for responses to the immunodominant proteins (NP and M) on all baseline samples from both challenge groups. T cell response against immunodominant proteins were predominantly CD4 T cell mediated in both groups (CD4 vs CD8 56% vs 44% for H3N2, and 71% vs 28% for H1N1) (FIG. 5b), consistent with a previous report (Lee L Y et al J Clin Invest. 118, 3478-3490 (2008)).

(63) To understand better the kinetics of T cell responses the functional status of both CD4 and CD8 cells during the course of infection with H1N1 virus was studied. Activated (CD38.sup.+) and proliferating cells (Ki67.sup.+) of both CD4 and CD8 cells from freshly isolated PBMC were undetectable before the challenge (FIG. 5c). Both markers were present on the greatly expanded T cell population on day 7 before returning to baseline level on day 28. (FIG. 5d). The number of Ki67.sup.+CD38.sup.+ T cells correlated with the frequency of SFC by Elispot on day 7 (FIG. 5e, r=0.9, p=0.002, Spearman coefficient).

(64) Impact of Pre-Existing T Cell Responses on Viral Shedding and Symptom Scores in Experimental Influenza Infection

(65) The role of T cells in controlling virus shedding (viral control) (Li I W et al. Chest. 137, 759-68 (2010)) and symptom development (immunopathology) (La Gruta N L et al. Immunol Cell Biol. 85, 85-92 (2007)) was studied. The relationship between pre-existing T cells responding to total and immunodominant influenza proteins (NP+M), virus shedding, total symptom scores and illness duration was analysed. A correlation test (Spearman rank correlation test, Prism 5) was run to see if the magnitude of flu-specific CD4 or CD8 cells were correlative in virus shedding and disease severity as indicated by total symptom scores and length of illness duration in both H3N2 and H1N1 challenge studies. The results clearly showed that the magnitude of CD4 response against immunodominant nucleoprotein (NP) and matrix (M) proteins was inversely correlative to peak virus shedding, symptom scores and illness durations (Table 2). This demonstrates that pre-existing T cell immunity to a virus can ameliorate subsequent infection with that virus. This also shows that the present methods allow determination of the peptides which induce a response.

(66) As shown in FIG. 6 and Tables 3a and 3b, the magnitude of total pre-existing T cells was strongly correlated with illness duration in H3N2 challenge study (Table 3a, FIG. 6a, r=−0.5740, p=0.0318, Spearman coefficient) and total symptom scores in H1N1 challenge (Table 3b, FIG. 6b, r=−0.7113, p=0.0369, Spearman coefficient,).

(67) TABLE-US-00004 TABLE 3a Correlation of pre-existing CD4 or CD8 cell responses to immunodominant proteins with control of virus shedding and symptom development in H3N2 challenge infection Peak viral load (TCID.sub.50/ml) Symptom scores Illness duration (days) Correlation Correlation Correlation Protein coefficient P-value coefficient P-value coefficient P-value Total −0.3330 0.2447 −0.2257 0.4379 −0.5740 0.0318 NP and M −0.4972 0.0704 −0.3402 0.2340 −0.6918 0.0061 NP and M −0.6087 0.0209 −0.5390 0.0467 −0.7886 0.0008 (CD4) NP and M 0.0127 0.9657 0.09640 0.7430 −0.1617 0.5808 (CD8)

(68) TABLE-US-00005 TABLE 3b Correlation of pre-existing CD4 or CD8 cell responses to influenza proteins with control of virus shedding and symptom development in H1N1 challenge infection Symptom scores Illness duration Correlation Correlation Protein coefficient P-value coefficient P-value Total −0.7113 0.0369 −0.6102 0.0857 NP and M −0.852 0.0108 −0.7404 0.0255 NP and M (CD4) −0.6908 0.0433 −0.6110 0.0857 NP and M (CD8) −0.2079 0.5809 −0.1053 0.7756

(69) When the T cell responses to the immunodominant proteins (NP and M) were examined in detail, it was observed that these protective T cell responses were mediated by pre-existing CD4 (FIG. 6a, right panel; FIG. 6b top right), but not CD8 T cell responses. The observed correlation was independent of the size of flu-specific CD8 response in that the magnitude of pre-existing CD4, but not CD8, cells against the internal proteins NP and M were inversely associated with total symptom scores in both challenge groups. More importantly, virus shedding of H3N2 was predominantly controlled by the level of pre-existing CD4 responses to internal proteins NP and M (r=−0.6087, p=0.0209, Spearman coefficient) but this was not the case for CD8 cells (r=−0.0127, p=0.9657, Spearman coefficient).

(70) To determine the relationship between the acutely expanding T cell population and illness metrics, the relationship between peak T cells on day 7, viral load and symptom severity was determined. The size of the developing acute T cell response correlated positively with viral shedding and illness severity for both models. These findings suggests that pre-existing memory CD4 T cells are the key in the CMI response in limiting illness that once illness is established acutely expanding cell populations tracked peak viral load and thus symptoms.

(71) Phenotypes of Pre-Existing T Cells Against NP and M Flu Proteins

(72) Pre-existing T cell responses against internal protein NP and M as measured by IFN-γ responses were largely CD4 T cell mediated in both H1N1 and H3N2 study groups. In the H3N2 challenge group, 9 subjects had NP and M responses at baseline and 8/9 (89%) had their peptides identified at a single peptide level (Table 4a). For the M protein, 5 out of 8 (63%) peptide responses were CD4 T cell mediated (partial results shown) and for the NP protein, 9 out of 12 (75%) peptide responses were CD4 T cell mediated (results for 9 peptides shown). In the H1N1 challenge group, 7 subjects which were positive with NP and M at baseline and 7/7 (100%) had their peptides identified at a single peptide level (Table 4b). For the M protein, 5 out of 5 (100%) peptides were seen by CD4 T cells (partial results shown) and for the NP protein, 3 out of 6 (50%) peptides were seen by CD4 T cells (partial results shown).

(73) Phenotypes of Induced T Cells Against NP and M Flu Proteins

(74) In the day 7 antigen-specific T cell response to NP and M proteins most of the response was by CD4 T cells (Table 5). Upregulation of CD107a expression on memory CD4 T cells was observed following ex vivo stimulation of peptide pools to NP or M proteins (FIG. 7a). Their cytotoxic function was further examined by a Cr-release assay using short-term T cell lines generated from baseline PBMC and on the killing of peptide pulsed autologous B cell lines. As shown in FIG. 7b, these CD4 T cells killed autologous target cells in a peptide specific manner. The killing was sensitive to concanamycin, suggesting cytotoxicity was dependent on the perforin/granzyme pathway. Therefore, these memory CD4 T cells possess a cytotoxic activity as described previously (Cazazza J P et al. J. Exp Med. 203, 2865-77 (2006)).

(75) MHC Class H Expression on Respiratory Epithelium and Changes During Infection

(76) A role for cytotoxic CD4.sup.+ cells in limiting viral infection would implicate the need for expression of MHC class II on the respiratory epithelium—the target of influenza infection. To investigate this we analysed the constitutive expression of the MHC class II molecule, HLA-DR, in explanted lung tissue and on primary bronchial epithelial cells in culture (PBECs) and the effect of in vitro influenza infection on expression in PBECs. We found significant constitutive expression of this molecule in both lung tissue and cultured PBECs with a rise in HLA-DR expression after infection of PBECs compared to cells treated with UV-inactivated virus (data in FIG. 8a,b).

(77) TABLE-US-00006 TABLE 4a T cell peptide responses in H3N2 challenge study subjects Amino SEQ ID NO SFC/million Number Peptide acid SEQ ID in sequence CD4 or CD8 PBMC positive Protein ID  position Amino acid sequence NO listing dependency (range) (%) M M15 103-119 LKREITFHKAKEIALSY  6  1 4   67 (35-96) 3 (38) M M23 159-175 HRSHRQMVATTNPLIKH  7  2 4 158 1 (13) M M25 173-189 IKHENRMVLASTTAKAM  8  3 4 118 1 (13) NP NP05 24-41 EIRASVGKMIDGIGRFYI  9  4 4 141.5 (35-248) 2 (26) NP NP08 48-65 KLSDHEGRLIQNSLTIEK 10  5 4  26 1 (13) NP NP14  95-111 PIYRRVDGKWMRELVLY 11  6 4   113 (45-181) 2 (26) NP NP15 102-119 GKWMRELVLYDKEEIRRI 12  7 4    93 (45-141) 2 (26) NP NP20 141-156 SNLNDATYQRTRALVR 14  8 8 391 1 (13) NP NP21 147-163 TYQRTRAVLRTGMDPRM 15  9 8 331 1 (13) NP NP32 229-246 KFQTAAQRAMVDQVRESR 18 10 8 104 1 (13) NP NP57 404-420 GQTSVQPTFSVQRNLPF 19 11 4  10 1 (13) NP NP58 411-428 TFSVQRSLPFEKSTIMAA 20 12 4  10 1 (13)

(78) TABLE-US-00007 TABLE 4b T cell peptides responses in  H1N1 challenge study subjects Amino SEQ ID NO SFC/million Peptide acid SEQ ID in sequence CD4 or CD8 PBMC Positive Protein ID position Amino acid sequence NO listing dependency (range) no(%) M M13  97-114 VKLYRKLKREITFHGAKE 23 13 4 41 (30-52) 2 (28) NP NP09 65-82 RMVLSAFDERRNKYLEEH 26 14 4 38 1 (14) NP NP27 209-226 GENGRKTRIAYERMCNIL 29 15 8 30 1 (14) NP NP28 217-234 IAYERMCNILKGKFQTAA 30 16 8 40 1 (14) NP NP52 409-426 QPTFSVQRNLPFDKTTIM 31 17 4 26 1 (14)

(79) TABLE-US-00008 TABLE 5 T cell responses to H1N1 challenge study SFC/million PBMC Amino  SEQ ID NO Days after Peptide acid SEQ ID in sequence CD4 or CD8 challenge Subjects ID position Amino acid sequence  NO listing dependency −2 7 28 B017(2C)  NP52 409-426 QPTFSVQRNLPFDKTTIM 31 17 4 26  76  10 B017(2C)  M13  97-114 VKLYRKLKREITFHGAKE 23 13 4 52 316  33 B005 (15C) NP27 209-226 GENGRKTRIAYERMCNIL 29 15 8 30 157  10 B005 (15C) NP28 217-234 IAYERMCNILKGKFQTAA 30 16 8 40  67  10 B005 (15C) M13  97-114 VKLYRKLFREITFHHAKE 23 13 4 30  70  10 B005 (20C) NP09 65-82 RMVLSAFDERRNKYLEEH 26 14 4 38 187 113