General enquiries on this form should be made to:randd.defra.gov.uk/Document.aspx?Document=SE3222_8531... · Web viewAntigens that were prioritized by in silico analysis were then

General enquiries on this form should be made to:Defra, Science Directorate, Management Support and Finance Team,Telephone No. 020 7238 1612E-mail: [email protected]

SID 5 Research Project Final Report

SID 5 (Rev. 3/06) Page 1 of 23

NoteIn line with the Freedom of Information Act 2000, Defra aims to place the results of its completed research projects in the public domain wherever possible. The SID 5 (Research Project Final Report) is designed to capture the information on the results and outputs of Defra-funded research in a format that is easily publishable through the Defra website. A SID 5 must be completed for all projects.

This form is in Word format and the boxes may be expanded or reduced, as appropriate.

ACCESS TO INFORMATIONThe information collected on this form will be stored electronically and may be sent to any part of Defra, or to individual researchers or organisations outside Defra for the purposes of reviewing the project. Defra may also disclose the information to any outside organisation acting as an agent authorised by Defra to process final research reports on its behalf. Defra intends to publish this form on its website, unless there are strong reasons not to, which fully comply with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.Defra may be required to release information, including personal data and commercial information, on request under the Environmental Information Regulations or the Freedom of Information Act 2000. However, Defra will not permit any unwarranted breach of confidentiality or act in contravention of its obligations under the Data Protection Act 1998. Defra or its appointed agents may use the name, address or other details on your form to contact you in connection with occasional customer research aimed at improving the processes through which Defra works with its contractors.

Project identification

1. Defra Project code SE3222

2. Project title

Development of improved diagnostic tests for the detection of bovine tuberculosis

3. Contractororganisation(s)

Veterinary Laboratories AgencyWoodham LaneNew HawAddlestoneSurrey KT15 3NB

54. Total Defra project costs £ 1,770,821(agreed fixed price)

5. Project: start date................ 01 July 2005

end date................. 31 December 2008


6. It is Defra’s intention to publish this form. Please confirm your agreement to do so...................................................................................YES NO (a) When preparing SID 5s contractors should bear in mind that Defra intends that they be made public. They

should be written in a clear and concise manner and represent a full account of the research project which someone not closely associated with the project can follow.Defra recognises that in a small minority of cases there may be information, such as intellectual property or commercially confidential data, used in or generated by the research project, which should not be disclosed. In these cases, such information should be detailed in a separate annex (not to be published) so that the SID 5 can be placed in the public domain. Where it is impossible to complete the Final Report without including references to any sensitive or confidential data, the information should be included and section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No" answer.In all cases, reasons for withholding information must be fully in line with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.

(b) If you have answered NO, please explain why the Final report should not be released into public domain

Executive Summary7. The executive summary must not exceed 2 sides in total of A4 and should be understandable to the

intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together with any other significant events and options for new work.

The incidence of bovine tuberculosis in GB has been increasing since 1988 despite the use of a control strategy based on tuberculin skin testing and slaughter of animals that react positively to the test. To develop vaccination strategies for cattle is an important part of Defra's research into future control strategies. Currently, the most promising cattle TB vaccine strategies are based on the boosting of BCG induced immunity with viral or protein subunit vaccines (so-called heterologous prime-boost strategies). However, application of BCG results in cattle, at least for some time post-vaccination, becoming positive in diagnostic tests using tuberculin. Therefore, in order to allow cattle vaccination using such heterologous prime-boost approaches to become a viable control policy option, diagnostic tests are required that can differentiate between infected and vaccinated cattle (differential diagnosis). This requires the development of so-called DIVA (Differentiation of Infected and Vaccinated Animals) reagents which could be applied alongside current test and slaughter control strategies. This project was proposed in response to a Defra research requirement document (September 2004) and directly addressed requirement R2 of this document. It was aimed at the continued development and optimisation of such reagents. It was based on successful earlier ROAME projects undertaken by our group that defined prototype DIVA reagents based on the use in the Bovigam IFN- test of synthetic peptides derived from antigens such as ESAT-6 and CFP-10. These previous studies demonstrated that the sensitivity of ESAT-6/CFP-10 was still lower compared to tuberculin sensitivity. The main objective of this proposal was therefore to identify additional DIVA antigens complementing ESAT-6/CFP-10 to increase test sensitivity without loss of specificity. As this report demonstrates, these objectives were fully met.The approaches taken in this project are based on comparative genome analysis and comparative transcriptome screening to list proteins that are potentially species-specific and can be applied to increase the specificity. The elucidation of the genome sequences of M. bovis, M. tuberculosis, M. bovis BCG Pasteur, M. avium subsp. paratuberculosis and M. avium subsp. avium allowed us to extend the search for candidates for differential diagnostics based on comparative genome analysis (objective 01) as well as by comparing their transcriptomes after in vitro macrophage infection or culture under defined conditions (objective 02). Potential antigens that were prioritized by these approaches were then prepared as overlapping sets of synthetic peptides and experimentally assessed both for immunogenicity and specificity using


blood samples obtained from experimentally and naturally M. bovis infected cattle, BCG vaccinated cattle as well as naïve cattle.

Objective 1: Using genome comparison, a total 40 genes were selected for immunological analysis based on the following selection criteria: genes without homology to M. a. avium and M. a. paratuberculosis genes; genes with frame-shifting point mutations in BCG Pasteur that lead to truncated gene products; or genes deleted entirely from the genome of BCG Pasteur. In general, the proteins were poorly recognised in M. bovis infected cows, compared to recognition of ESAT-6 and CFP-10. Therefore none of these 40 antigens showed diagnostic potential when used to stimulate cell-mediated immune responses. This also highlights the limitations of genome comparison to prioritise antigens, as bioinformatical approaches based on primary sequence analysis to-date are not yet robust enough to predict antigenicity or specificity.

Objectives 02 and 04:In this work package we assessed the transcriptome of BCG Pasteur and M. bovis 2122/97 during growth of these bacilli in bovine macrophages or under different defined in vitro conditions in a chemostat (objective 02). Microarray expression analysis was then performed using a DNA microarray containing all genes from M. bovis and M. tuberculosis. 41 gene products thus identified were tested for immunogenicity and specificity in infected and vaccinated cattle. This approach proved to be highly successful as it lead to the identification and characterisation of Rv3615c as DIVA reagent. Significantly, Rv3615c could be shown to preferentially recognise M. bovis infected cattle that were not detected using ESAT-6/CFP-10 and therefore increased combined DIVA test sensitivity to levels comparative to tuberculin, whilst not being recognised by BCG vaccinated or naïve cattle. We also characterised its epitopic peptides and demonstrated that it was recognised by CD4+ and CD8+ T cells. In objective 04, we were able to substantiate these data in a larger number of experimentally infected or vaccinated as well as naturally infected cattle by further demonstrating that Rv3615c constitutes a potent DIVA reagent that increased overall test sensitivity (when used in combination with ESAT-6 and CFP-10) to levels comparable with tuberculin without loss of specificity in BCG vaccinated and naïve cows (relative sensitivity of tuberculin, ESAT-6/CFP-10, ESAT-6/CFP-10 plus Rv3615c = 96.5, 85.8, 94.1 %, respectively). This antigen combination is therefore a prime candidate for further field evaluation.

Objective 03:Careful response kinetics analysis using experimentally infected cattle established that the defined DIVA antigens tested, including ESAT-6, CFP-10 and Rv3615c, are recognised early post-infection with similar kinetics as tuberculin. Thus, using defined antigens for diagnosis is unlikely to result in reduced sensitivity by detecting cattle later after infection than tuberculin.

Objective 05:The collaborative network has been maintained and extended to include an industrial partner. This network has lead to exchange of information, reagents, and staff. By testing recombinant Rv3615c produced by one of the collaborators (VSD, Stormont, Northern Ireland), we were able to establish that cocktails of synthetic peptides are at least, if not more, potent and sensitive when used as diagnostic reagents in the IFN test. In a collaborative study with AgResearch, Palmerston North, New Zealand) we determined that in vitro neutralisation of IL-10 increased test sensitivity. However, the effects of this intervention on specificity need to be evaluated in further studies.

The results obtained in this project have been published in a number of scientific papers, and a patent have been filed to protect the IP generated. General conclusions on antigen mining approaches are being discussed in the report, and future research directions are being proposed.


Project Report to Defra8. As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with

details of the outputs of the research project for internal purposes; to meet the terms of the contract; and to allow Defra to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information obligations. This short report to Defra does not preclude contractors from also seeking to publish a full, formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms. The report to Defra should include: the scientific objectives as set out in the contract; the extent to which the objectives set out in the contract have been met; details of methods used and the results obtained, including statistical analysis (if appropriate); a discussion of the results and their reliability; the main implications of the findings; possible future work; and any action resulting from the research (e.g. IP, Knowledge Transfer).

IntroductionIn an attempt to improve the detection of cattle infected with M. bovis, previous research has resulted in the development of a diagnostic test for bovine tuberculosis based on the measurement of interferon-gamma (IFN-) in vitro. In its original form, this assay measures the production of IFN- after in vitro stimulation with bovine and avian tuberculin PPD. This test is now permitted under EU law as an adjunct to the tuberculin skin test in cattle. It is a rapid and practical test and has potential to detect animals at an earlier stage of infection, but has slightly lower specificity than the skin test used in the UK. Moreover, BCG vaccination sensitises animals to antigens present in bovine tuberculin and therefore compromises tuberculin-based diagnosis strategies in BCG vaccinated animals. The most promising cattle vaccine strategies to-date encompass the boosting of BCG with subunit vaccines based on recombinant attenuated viruses or protein subunits. In the absence of a vaccine that results in 100% protection, the development of more specific diagnostic reagents capable of discriminating between infected and uninfected vaccinated animals (DIVA test) is a pre-requisite for the use of a vaccine against tuberculosis in cattle so that test and slaughter control strategies can be carried out alongside vaccination regimens. This project was proposed in response to a Defra research requirement document (September 2004) and directly addressed requirement R2 of this document.. It was based on the success of previous Defra-funded projects that identified several prototype DIVA reagents based on antigens such as ESAT-6 (Rv3875) or CFP-10 (Rv3874). The main objective of this proposal was therefore to identify additional DIVA antigens complementing ESAT-6/CFP-10 to increase test sensitivity without loss of specificity. The objectives have been fully met, in particularly by the identification and characterisation of the antigen Rv3615c as DIVA reagents that was recognised preferentially by infected cattle escaping ESAT-6 and CFP-10 detection.

The approaches taken in this project are based on comparative genome analysis and comparative transcriptome screening to list proteins that are potentially species-specific and can be applied to increase the specificity of immuno-diagnostic reagents detecting cattle infected with Mycobacterium bovis. Specifically, these antigens are to be used in conjunction with the Bovigam IFN- test to allow the differentiation between infected and vaccinated cattle (DIVA). In addition, to assign immunological activity to any thus prioritised candidate, we applied a high-throughput immunological screening system based on sets of synthetic peptides representing the proteins that are then tested in whole blood cell stimulations. As the main objective of this proposal was to identify diagnostic antigens that would complement ESAT-6 and CFP-10 by identifying infected animals that fail to respond to these prototype DIVA reagents, we endeavoured to source infected field reactor cattle that, whilst tuberculin skin and IFN- test positive, did not recognize these two antigens. We were fortunate to acquire a number of these animals, and their usefulness was demonstrated particularly in objectives 02 and 04.

Objective 01. Extended antigen mining using comparative genomics.


The elucidation of the genome sequences of M. bovis [1], M. tuberculosis [2], M. bovis BCG Pasteur [3], M. avium subsp. paratuberculosis [4] and M. avium subsp.avium have been completed. This allowed us to extend the search for candidates for differential diagnostics based on comparative genome analysis. Antigens that were prioritized by in silico analysis were then assessed experimentally both for immunogenicity and specificity in cattle.

Annotation of the BCG Pasteur sequence was used to search for potential proteins that were deleted in BCG compared to M. bovis, or were found to contain truncated gene products or altered amino acid sequences (see below). Briefly, a file of all proteins encoded by the 3592 genes of M. bovis was generated and used to search the available mycobacterial genome sequence data using the BLAST suite of programmes. Stand-alone BLASTP and TBLASTN was used to search the M. avium subsp. avium and M. avium subsp. paratuberculosis sequences to identify M. bovis proteins with low score and E values, i.e. potentially M. bovis unique proteins. In addition, the file of 3592 M. bovis gene was also searched for open-reading frames (ORF), i.e. potential proteins, whose genes were absent from the BCG Pasteur genome, or BCG genes with frame-shift mutations resulting in truncated gene products (by frame-shift mutation introducing a stop codon), or altered amino acid sequences (frame-shift mutations changing codons). The potential specificity of these selected M. bovis genes was then confirmed by demonstrating the absence of sequence homology with other M. bovis genes.

Using this approach, a total 40 genes were selected for immunological analysis: 10 genes without homology to M. a. avium and M. a. paratuberculosis genes, thus potentially applicable to improving tuberculin specificity; 15 genes with frame-shifting point mutations in BCG Pasteur that lead to truncated gene products; potentially applicable to differential diagnosis; and 15 genes deleted entirely from the genome of BCG Pasteur (see table 1 for list of these genes).

Table 1. List of ORFs selected for immunological analysis based on genome comparison. Gene ID based on M. bovis genome

Function (based on annotation on Bovilist website (http://genolist.pasteur.fr/BoviList/

Selection criteria (aa = amino acid; bp = base pair)

Mb0037c CHP Frameshift: 3’ end, different C-terminal (55 aa) in BCG compared to M. bovis

Mb0138 Putative acetyltransferase Frameshift: 3’ end, different C-terminal (41 aa) in BCG compared to M. bovis

Mb0386 Possible protein transport protein SecE2 No homology with M. avium and M. paratuberculosis

Mb0619 Possible exported protein No homology with M. avium and M. paratuberculosis

Mb0771c HP No homology with M. avium and M. paratuberculosis

Mb0924 IMP No homology with M. avium and M. paratuberculosis

Mb1000c Probable acyl-CoA dehydrogenase Frameshift: 3’ end, different C-terminal (20 aa) in BCG compared to M. bovis

Mb1057 IMP (Kpdf) No homology with M. avium and M. paratuberculosis

Mb1222 CHP Entire gene deleted in BCGMb1223 CHP Frameshift: 5’ end deleted in BCG

(180 amino acids)Mb1522c CHP Frameshift: 3’ end, different C-

terminal (18 aa) in BCG compared to M. bovis

Mb1588c Maltooligosyltrehalose trehalohydrolase Frameshift: at 3’ end BCG causing different 3’ end compared to M. bovis


Mb1599 Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1600 Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1601c Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1602c Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1603c Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1604c Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1605c Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1606c Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1607c Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1608c Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1609c Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1610c Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1611c Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1612c Probable prophage protein (phiRV1) Entire gene deleted in BCGMb1974 Probable oxidoreductase Frameshif: different 3’ end in BCG

with different C-terminal aa 77 residues compared to M. bovis

Mb2107c TMP Frameshift: 3’ end due to 8 bp insertion, different C-terminal (55 aa) in BCG compared to M. bovis

Mb2152 TMP No homology with M. avium and M. paratuberculosis

Mb2196 IMP Frameshift: 3’ end, different C-terminal (18 aa) in BCG compared to M. bovis

Mb2295 IMP No homology with M. avium and M. paratuberculosis

Mb2437c HP Frameshift: 3’ end, different C-terminal (40 aa) in BCG compared to M. bovis



Mb3512c IMP Frameshift: 3’ end, different C-terminal (39 aa) in BCG compared to M. bovis

Mb3610c Putative tRNA/rRNA methyltransferase Frameshift: 3’ end with 86 terminal aa residues different in BCG compared to M. bovis

Mb3678 HP No homology with M. avium and M. paratuberculosis

Mb3865 IMP Frameshift: 3’ end with 70 C-terminal aa different compared to M. bovis

Mb3890 CHP Frameshift: 3’ end, different C-terminal (40 aa) in BCG compared to M. bovis

Mb3931c IMP Frameshift: 3’ end (2 bp deletion), different C-terminal (18 aa) in BCG compared to M. bovis

CHP= conserved hypothetical protein; IMP = integral membrane protein; HP = integral membrane protein, TMP = transmembrane protein.

Next, a set of about 1000 overlapping synthetic peptides (20mer peptides, with a 12 residues off-set) spanning the complete amino acid sequences of these target gene products were synthesised, and formulated into pools of around 10 peptides/pool. These pools were screened, using in vitro IFN- production as read-out system, with blood samples from cattle experimentally and naturally M. bovis infected cattle, BCG vaccinated cattle and uninfected/unvaccinated controls. The results of this analysis are presented in Figure 1. In general, the proteins were poorly recognised with responder frequencies below 25 % of M. bovis infected cows, compared to recognition of ESAT-6 and CFP-10 (90%, Figure 1). Further, responses of individual responder animals were low when IFN- production


was assessed by comparing OD450 (optical density measured at 450 nm) values, which were in most cases just above the background value set for positivity (OD450 value with background subtracted > 0.1, data not shown). Apart from two proteins (Mb1000c and Mb1602c), responses were also detected in BCG vaccinated and control animals. However, the low responses to stimulation with these two antigens meant that they were not suitable to improve diagnostic sensitivity, nor did they complement ESAT-6/CFP-10.

Consequently, none of these 40 antigens showed diagnostic potential when used to stimulate cell-mediated immune responses. This also highlights the limitations of genome comparison to prioritise antigens, as bioinformatical approaches based on primary sequence analysis to-date are not robust enough to predict either antigenicity or specificity (or X-reactivity).

ESAT-6/CFP10

Mb2437

c

Mb3865

Mb1000

c

Mb1522

c

Mb1588

c

Mb1602

c

Mb1612

c

Mb2784

c

Mb3512

c

Mb3678

Mb3890

Mb0037

c

Mb0771

c

Mb1057

Mb2152

Mb2196

Mb2479

c

Mb0386

Mb0619

Mb1599

Mb1603

c

Mb1611

c

Mb2107

c

Mb3931

c

Mb0924

Mb1222

Mb1223

Mb1604

c

Mb1608

c

Mb1974

Mb2295

Mb0138

Mb1600

Mb1601

c

Mb1605

c

Mb1606

c

Mb1607

c

Mb1609

c

Mb1610

c

Mb3610

c0

20

40

60

80

100

Res

pond

er fr

eque

ncy

[%]

Figure 1. Recognition of antigens listed by comparative genome analysis (Table 1). Data are presented as % of M. bovis infected animals (n=20) recognising listed antigens represented by peptide pools (responder frequencies). Comparison is with recognition on ESAT-6 and CFP-10 peptide pool.

Objective 02. Antigen mining based on differential gene expression between M. bovis and BCG inside bovine macrophages (comparative transcriptomics) to identify antigens for differential diagnosis.

The transcriptome is the complete gene expression profile of an organism under defined conditions. In this objective we assessed the transcriptome of BCG Pasteur and M. bovis 2122/97 during growth of these bacilli in bovine macrophages. A method of infecting alveolar macrophages with these two tubercle bacilli was developed by adapting the methodology of Schnappinger and colleagues [5] which they used to define the transcriptome of macrophage-grown M. tuberculosis. Following infection, mycobacterial RNA was isolated, and a single-step amplification protocol developed, which resulted in sufficient cDNA to perform microarray analysis. Microarray expression analysis was then performed using a DNA microarray containing all genes from M. bovis and M. tuberculosis developed at VLA in collaboration with the BUGS Group at St George’s Hospital, London (bugs.sghms.ac.uk/index.php). This allowed the comparison of the transcriptomes of BCG and M. bovis when grown intracellularly in alveolar macrophages. In addition, the transcriptomes of in vitro grown bacilli with those of intracellularly grown BCG and M. bovis has also been performed. Further, the transcriptomes of bacilli grown in vitro under different conditions were also assessed (this latter approach was done in part in collaboration with Professor N. Stoker, Royal Veterinary College, London) and the combined data used to prepare a list of potential antigens (n= 41 ORFs, see table 2).

Next, peptides (about 1100 overlapping synthetic 20mer peptides, with a 12 residues off-set) spanning the complete amino acid sequences of these target gene products were synthesised, and formulated into pools of around 10 peptides/pool. These pools were then screened, using in vitro IFN- production as read-out system, with blood samples from cattle infected naturally infected with M. bovis (n=20), BCG vaccinated cattle (n=30) as well as uninfected/unvaccinated control cattle (n=10). Generally, low responder frequencies in M. bovis infected field reactor animals were observed: 44 % (18/41) ORFs


http://bugs.sghms.ac.uk/index.php

tested were not recognised, 54 % (22/41) were recognised by < 15 % of animals tested. However, one antigen, Rv3615c/Mb3645c, was recognised at a high frequency of 37 % (11/30 animals tested). Results from 14 representative antigens including Rv3615c, tested using blood from M. bovis infected and naïve control animals are shown in figure 2. Interestingly, as shown in Fig. 2, Rv3615c did not elicit IFN- responses in naïve control animals (or BCG vaccinated animals, see below) whilst antigens such as Rv2876 did, thereby ruling them out as diagnostically relevant antigens. Rv3615c was therefore recognised in a statistically significantly larger proportion of infected animals compared to naïve control animals (Fisher’s exact test, p < 0.05, power > 80%).

Table 2. List of ORFs selected for immunological analysis based on transcriptome analysisRv number Mb number Size (aa) Function etc1222 1254 154 CHP967 992 119 CHP997 1024 143 CHP1088 1117 144 PE91089 1118 101 PE101211 1243 257 SigE1375 1410 439 CHP2232 2257 291 CHP2408 2431 239 PE242874 2899 695 IMP2876 2901 104 TMP2941 2966 580 HP3407* 3441 99 CHP3478 3505 393 PPE603633 3657 291 CHP3866 3896 283 CHP0445c 0453c 183 RNA polym*0446c 0454c 256 TMP*0447c 0455c 427 Lipid synthesis*0448c 0456c 221 CHP*0449c 0457c 439 CHP*1169c 1202c 100 PE111388c 1433c 85 CHP1813c 1843c 143 CHP1963c 1998c 406 mce31993c 2016c 90 CHP2081c 2107c 147 TMP2356c 2377c 614 PPE402626c 2659c 143 DosR regulon protein2877c 2902c 287 IMP3222c 3249c 183 CHP3231c 3260c 169 CHP3271c 3299 222 IMP3477^^ 3504 98 PE313529c 3559c 384 CHP3613c 3643c 53 HP3614c 3644c 184 CHP3615c 3645c 103 CHP3702c 3728c 233 CHP3739c/3738c** 3765c 118 PPE663750c 3776c 130 ExcisionaseAbbreviations: see table 1. In addition: PE= PE protein family member; PPE = PPE protein family member (see also [2] for further info on PE and PPE protein families).

Next, it was of particular interest to determine the overlap in responses in individual infected animals between Rv3615c and ESAT-6/CFP-10. We observed that there were responses to the Rv3615c peptide pool in M. bovis-infected cattle that did not respond to ESAT-6 or CFP-10 (Fig. 3A, 4 of 7,


57%). To confirm Rv3615c’s potential as a DIVA antigen for the differential diagnosis of BCG-vaccinated and M. bovis-infected animals, we screened the peptide pool in 20 BCG-vaccinated cattle. In contrast to M. bovis-infected animals, none of the BCG vaccinated cattle generated a significant IFN- response to Rv3615c (Fig. 3A, p < 0.0001). In contrast, 80% (16 of 20) BCG vaccinated animals demonstrated a significant IFN- response to PPD-B when compared to the naïve animals (Figure 3B). This is despite the fact that M. bovis BCG possesses an identical copy of this gene, indicating that antigenic diversity can occur independent of sequence diversity. To confirm the presence and location of the T cell epitopes within Rv3615c, we determined the response to constituent peptides from the Rv3615c pool using an IFN-γ EliSpot assay with PBMC (peripheral blood mononuclear cells) isolated from M. bovis-infected cattle. Peptides 6-11 were recognised promiscuously by the majority of cattle tested thus defining the C-terminus of the protein as the most antigenic part of the protein (data not shown).

Figure 2. Proportion of infected or control animals recognising antigens selected by transcriptome analysis. Results for 14 representative/41 mycobacterial antigens tested are shown which were screened for an IFN-γ response in 30 M. bovis infected and 10 naïve cattle. The responder frequencies, as percentages, for each antigen from M. bovis-infected (blue bars) and naive cattle (yellow bars) are shown here.

SID 5 (Rev. 3/06) Page 10 of 23

Figure 3: Rv3615c is specific DIVA reagent and can complement ESAT-6/CFP-10 to increase sensitivity. IFN- production upon stimulation with Rv3615c (2A) or PPD-B (2B) for each animal within each group are shown here. The data is expressed as background subtracted optical density (ΔOD450), and the dashed line represents the cut-off for positivity. Significance testing was performed using the non-parametric Wilcoxon rank sum test adjusted for multiple comparisons.

To further characterise the specific lymphocyte response to Rv3615c, we measured intracellular IFN-γ production in the CD4+ and CD8+ T cell sub-populations after stimulation with the Rv3615c peptides using FACS (Fluorescence Activated Cell Sorting) analysis. Mirroring the EliSpot data, we found peptides 1 - 6 stimulated little IFN-γ. Markedly greater levels of IFN-γ were observed from the CD4+ T cells stimulated with peptides 7-11 confirming the C-terminal half of the protein as its immunodominant

SID 5 (Rev. 3/06) Page 11 of 23

region (Figure 4). In addition to being recognised more frequently by CD4+ T cells, the results in figure 4 demonstrate that peptides 7, 8, 10 and 11 were also recognised by CD8+ T cells suggesting that the response to these peptides are both MHC class I and II-restricted.

Figure 4: FACS analysis on PBMC from 3 M. bovis infected cattle. Cells were stimulated with either RPMI medium (unstimulated), each individual Rv3615c peptide (5μg ml-1) or the Rv3615c peptide pool (5μg ml-1). Analysis: intracellular IFN-γ production by FACS. Data: lymphocyte population staining as CD4+/IFNγ+ (bars above x-axis) or as CD8+/IFNγ+ (bars below x-axis) from each animal.

As this work demonstrates, it is possible to identify promising diagnostic antigens by screening highly expressed genes. It was beneficial, therefore, to gain an understanding of the contribution mRNA level makes to a protein’s antigenicity. To assess this, we explored the correlation between mRNA levels and antigenicity further by including the responder frequencies of an additional 43 mycobacterial proteins, alongside the 41 screened here. Correlation of the responder frequencies for these antigens with their mRNA levels in both M. tuberculosis and M. bovis growing in chemostats revealed a low but statistically significant relationship in both organisms: Spearman’s Rank = 0.35 (p < 0.005) and 0.3 (p < 0.005) respectively. However, despite this level of statistical significance, as this analysis combined with our empirical screening results with the 41 proteins showed, transcriptome data or mRNA abundance levels alone is not sufficient to define a protein’s antigenicity. Therefore immunological screening procedures using high-throughput peptide-based experimental approaches like ours are still needed to assign immunogenicity and specificity to a given protein. The results described for this objective have been published (Sidders et al., [6], and a patent was filed.

In conclusion, the work presented in the context of this objective lead to the identification and characterisation of Rv3615c as DIVA reagent that, by preferentially recognising infected cattle escaping ESAT-6/CFP-10 detection, has the potential to increase DIVA test sensitivity by complementing ESAT-6/CFP-10.

SID 5 (Rev. 3/06) Page 12 of 23

Objective 03. Kinetics of antigen recognition post-infection

To determine at what time post-experimental infection diagnostically relevant antigens will be recognised in the IFN-g assay, a group of 10 animals were intratracheally infected with around 2000 CFU of M. bovis strain AF2122/97 as described [7]. Blood samples were taken regularly post-infection from these animals and stimulated with bovine tuberculin PPD, ESAT-6 and CFP-10 peptides, Rv3615c peptides, and a peptide cocktail composed of the complete set of ESAT-6 and CFP-10 peptides, plus selected immunogenic peptides from Rv3019c, Rv0288, Rv3873, Rv3879c (PC2 cocktail). PC-2 was developed from DIVA peptide cocktail PC1, described by us previously [8,9]. Whilst PC1 contained only 12 ESAT-6 and CFP-10 derived peptides, PC2 contains the full complement of 25 ESAT-6/CFP-10 peptides. The results are shown in Figure 5 and demonstrate that IFN- responses to ESAT-6/CFP10 and the antigens represented in cocktails PC2 specific developed alongside PPD-B responses (beginning 2 weeks post-infection) whilst responses to Rv3615c developed only marginally later post-infection (starting at 4 weeks post-infection). Responses to all antigens tested increased and were maintained at high levels until the termination of the experiment 14 weeks post-infection. In conclusion therefore, these defined DIVA antigens are recognised early post-infection with similar kinetic to tuberculin responses. Thus, using defined antigens for diagnosis is unlikely to result in reduced sensitivity by detecting cattle later after infection than tuberculin.

Figure 5. Kinetic of IFN- responses following M. bovis infection. Results are expressed as mean OD450 values with background (medium control) responses subtracted (n = 10). Similar response kinetics were observed in a second challenge experiment (n = 10, data not shown).

Objective 04: Optimisation of antigen cocktails tested in objectives 1-3.

Based on published results and data generated through objectives 1-3, the following peptide cocktails were studied in larger sets of cattle to determine their performance as DIVA reagents: 1) ESAT-6/CFP-10 cocktail of overlapping peptides covering complete sequences [7,8,10]; 2) PC2 [8](see definition under objective 03), and 3) the peptide cocktail composed of overlapping peptides covering the complete sequence of Rv3615c [6]. Rv3615c was studied on its own to determine further its potential to complement ESAT-6 and CFP-10 to detect animals not detected by these two antigens. Based on unpublished data by our collaborators at Prionics (Schiller and Marg, unpublished) we also had concerns that we would add too many peptides into a single cocktail, potentially resulted in inhibitory effects.

The first set of data describes results from experimental BCG vaccination and challenge experiments. Confirming earlier observations from the literature [6,7,8,9,10] and in this report (objectives 02 and 03), BCG vaccination, whilst inducing strong PPD-B responses, did not induce IFN- responses after

SID 5 (Rev. 3/06) Page 13 of 23

stimulation with either ESAT-6/CFP-10, Rv3615c, or PC2 (Fig. 6). No responses against these defined reagents were seen in the naïve control animals (Fig. 6), although one animal tested positive using tuberculin as antigen (Fig. 6).

Figure 6. Responses of DIVA reagents in naïve (n=10) and BCG (n=20) vaccinated animals prior to M. bovis infection. Blood was cultured in the presence of antigens indicated, and IFN- determined by Bovigam assay. Results are mean responses + SEM. Horizontal line = cut-off for positivity.

These animals were then re-sampled 14 weeks after challenge and re-tested with the same set of reagents. The disease and protection status of these animals was determined at a post-mortem and the results are reported for unvaccinated controls (all visibly lesioned), BCG vaccinates that had no visible lesions (NVL), i.e. where vaccination was successful, and BCG vaccinates calves with lesions (VL). The results are shown in Figure 7. Infected naïve animals gave rise to strong responses to all reagents including PPD-B (and PPD-A, not shown), thus confirming together with data presented in Figure 6 that all three peptide cocktail constitute DIVA reagents. Weaker, but still highly potent responses were observed to all reagents in vaccinated calves presenting with visible lesions at post-mortem. Interestingly, lower responses, in particular to Rv3615c were evident in vaccinated animal without lesions (NVL). However, due to the nature of the model used to infect these cattle, at the post-mortem time point M. bovis cases could still be recovered from some NVL animals, and it was therefore not expected that all these animals would be all test-negative when stimulated with the DIVA reagents at this time point: all NVL BCG vaccinates were PPD-B-positive, whilst much lower proportions were were respectively positive to ESAT-6/CFP-10, PC2, and Rv3615c (table 3, 25-50 %). In contrast, VL BCG vaccinates displayed responder frequencies comparable to unvaccinated animals towards the defined peptide pools (table 3), 42 – 92 % compared ot 78-89 % in naïve animals). Further, the data confirmed that responses to ESAT-6, CFP-10, and Rv3615c correlated with disease severity considering that naïve/unvaccinated animal had pathology scores compared to VL vaccinates, which have in turn presented with higher scores than NVL vaccinates (data not shown). The data are suggestive that BCG vaccination does not mask detection of truly infected animals with these DIVA reagents. It is also likely (as we have shown in an earlier study [11]) that animals that will clear fully the challenge inoculi, will become test-negative when these DIVA cocktails will be used. However, as M. bovis culture results are not yet available for all NVL vaccinates, these results will need to be re-evaluated once this data has become available.

SID 5 (Rev. 3/06) Page 14 of 23

Table 3. Responder frequencies (Bovigam IFN- test) in naïve and vaccinated M. bovis Infected calves.

Animal categoryAntigensPPD-B ESAT6/CFP10 PC2 Rv3615c

Unvaccinated (all VL) 100 % (9/9) 89 % (8/9) 100 % (9/9) 78 % (7/9)BCG vaccinate-VL 100 % (12/12) 92 % (11/12) 83% (10/12) 42 % (5/12)BCG vaccinate - NVL 100 % (8/8) 50 % (4/8) 50 % (4/8) 25 % (2/8)VL= visible lesions found at post-mortem, NVL = no visible lesions found.

Figure 7. Responses of DIVA reagents in naïve and BCG vaccinated animals following M. bovis infection. Blood was cultured in the presence of antigens indicated, and IFN- determined by Bovigam assay. Animal groups represented are: unvaccinated cattle (blue, n=9), BCG vaccinates with (red, n=12) and without (green, n=8) visible lesions (Results are mean responses + SEM. Horizontal line = cut-off for positivity.

To further define the sensitivity and specificity of Rv3615c and PC2 compared to ESAT-6/CFP10 and tuberculin (PPD-B minus PPD-A), blood samples were obtained from field reactor animals from herds with multiple culture-positive reactor animals. The disease status of these animals was confirmed by post-mortem. In addition, we also tested samples from TB-free herds without recent TB history and from non-endemic areas including samples from Switzerland that were tested by our collaborator Prionics. The results of these experiments are tabulated in Table 3. Confirming previous results including data from the different trials determining tuberculin and ESAT-6/CFP-10 specificity and sensitivity, 96.5 and 85.8 % of infected field reactor animals were detected by ESAT-6/CFP-10 and tuberculin, respectively. Whilst Rv3615c detected less infected animals overall (54.1 %), it detected the majority of cattle that escaped ESAT-6/CFP-10 detection (7/12, Table 3). Taking responses to ESAT-6/CFP-10 and Rv3615c together increased overall sensitivity to 94.1 %, which is therefore approaching that of tuberculin (Table 3). The data in table 3 also confirmed the higher specificity of ESAT-6/CFP-10 and Rv3615c, both when used alone and in combination, compared to tuberculin).

Despite displaying good specificity (Table 3), cocktail PC2 was only marginally more sensitive than ESAT-6/CFP-10 in this study (88.5 compared to 85.8 %, Table 3).Further, in a study conducted in collaboration with several groups in Mexico, it also detected equal percentages of animals compared to ESAT-6/CFP-10 (data not shown). Therefore, the performance of PC2 was disappointing compared to Rv3615c.

In summary, we were able to substantiate our findings described in objective 2 and in Sidders et al., [6]that prioritised Rv3615c as a potent DIVA reagent capable of increasing overall test sensitivity when

SID 5 (Rev. 3/06) Page 15 of 23

used in combination with ESAT-6 and CFP-10. This antigen combination is therefore a prime candidate for further field evaluation.

Table 3. Relative sensitivity and specificity of DIVA reagents. SENSITIVITYa SPECIFICITYb

Antigen % (95% CI) Numberspositive

% (95% CI) Numberspositive

PPD-B-A 96.5 (90.0, 99.3) 82/85 94.4 (86.4, 98.5) 4/72ESAT-6/CFP-10* 85.8 (76.6, 92.5) 73/85 98.6 (92.5, 99.9) 1/72Rv3615c* 54.1 (42.9, 65.0) 46/85 100 (95.0, 100) 0/72Rv3615c (ESAT-6/CFP-10-negative subset)

58.3 (27.7, 84.8) 7/12 NA NA

Rv3615c plus ESAT-6/CFP-10

94.1 (86.8, 98.1) 80/85 99 (92.5, 99.9) 1/72

PC2 88.5 (73.3, 96.8) 31/35 97 (90.3, 99.7) 2/72a Based on blood samples from reactor animals from farms culture-positive M. bovis, with their infection status confirmed by PM and/or culture. b Animals from TB-free herds from non-endemic areas. *Pools of peptides.

Objective 05. Continue collaboration with VSD (Stormont, Northern Ireland) and AgResearch (Palmerston-North, New Zealand).

We have met with these collaborators frequently in person and also regularly had discussions by telephone and email. The collaboration with these two groups during this project was also complemented by additional collaborations with Prionics (Schlieren, Switzerland), the producer of the Bovigam test kit, and the National Animal Disease Centre (Ames, Iowa, USA). Prionics brought an industry perspective into this project as well as opened up their wider international collaborative network, including participation in a study in Mexico sponsored by the USDA. However, this study was not conducted within the remit of this project and confidentiality agreements between the partners prevents me from reporting specific data at present in a report like this aimed for public release. But I have reported on this study’s data to Defra in a confidential steering committee meeting (VPAG). In addition, I have presented in this report some of the data generated by Prionics using TB-free animals from Switzerland (see table 3). Peptide cocktails were shared with Prionics, with whom we also exchanged research staff for short-term placements, and with AgResearch who are using them in their on-going studies. Results will be reported to Defra as and when they will become available.

VSD was subcontracted to produce recombinant Rv3615c to enable us to compare protein responses with responses stimulated by a peptide pool. 2 mg of the protein were delivered in 2008 and assayed at VLA using blood from 25 naturally and experimentally infected cattle. This experiment was conducted to validate our peptide-based reagents, i.e. to investigate whether peptide-based reagents are as sensitive as recombinant proteins. The data, summarised in table 4, demonstrated no significant differences in detection rates between peptide and recombinant protein based Rv3615c (p = 0.321), although the Rv3615c peptide cocktail detected 4 more animals than tested positive after protein stimulation (Table 4). The amounts of IFN- produced, were also comparable (Linear correlation between peptide and protein induced OD450 values: r2 = 0.96, p < 0.005). Therefore, the Rv3615c peptide cocktail representing the complete sequence of this protein is as at least as sensitive in detecting infected animals than the corresponding recombinant protein.

Table 4. Comparison between responses to Rv3615c peptide cocktail and recombinant protein

N* = Peptide ProteinPositive 21 17Negative 4 8

Total 25 25Fisher’s exact test: p = 0.321, not significantly different.* Based on blood samples from reactor animals from farms culture-positive M. bovis, with their infection status confirmed by PM and/or culture.

SID 5 (Rev. 3/06) Page 16 of 23

VSD also attempted to clone and express a second protein, Rv3616c, from the same operon as Rv3615c that I had described in a previous study as being highly immunogenic ([12]. However, they did not succeed in producing this antigen so far, despite several time-consuming attempts. I have consulted several other groups who attempted to express this antigen which similarly failed or produced only minute quantities as Rv3616c is toxic when expressed in E. coli.

Furthermore, we also conducted a joint study with Drs Buddle and Denis (AgReserach) assessing the role of IL-10 on the diagnostic sensitivity of the Bovigam IFN- assay. The hypothesis tested being that, particularly in so-called anergic animals, IL-10 produced for example by regulatory T cells during the blood stimulation phase, or present in the plasma before culture commencement, would down-regulate IFN- responses. A similar effect had been previously demonstrated in M. paratuberculosis infected cattle [13]. Consequently, we determined if the sensitivity of the in vitro Bovigam test could be enhanced by adding antibodies neutralizing bovine IL-10. Blood was obtained from uninfected control cattle, experimentally infected cattle, cattle responding positively to the skin test in tuberculosis-free areas (false positives), and cattle naturally infected with Mycobacterium bovis from New Zealand and Great Britain. IFN- responses to PPD-B and PPD-A, avian purified protein derivative, and a fusion protein of ESAT-6 and CFP-10 were measured. We could demonstrate that whole blood cells from field reactors produced substantial amounts of IL-10 upon stimulation with PPD-B or ESAT-6/CFP-10 (proteins as well as peptides), thus demonstrating that IL-10 is directly induced during the blood culture step with antigen and not present in the plasma before culture initiation (data not shown). Further, adding the neutralizing antibody against IL-10 enhanced antigen-specific responses. In particular, addition of anti-IL-10 to ESAT-6/CFP-10-stimulated blood cultures enhanced test sensitivity (Figure 8).

Although initial testing “false-positive” cattle from tuberculosis-free areas of New Zealand revealed that addition of anti-IL-10 did not compromise the test specificity further preliminary screening of larger groups of tuberculin skin test-negative cattle has revealed that some cattle that produced IFN- in response to PPD-A also produced IFN- responses to ESAT-6/CFP-10, which were further amplified by the addition of anti-IL-10 (Denis and Buddle, unpublished). Therefore further studies are needed to determine the full effects of IL-10 neutralization on specificity.

Figure 8. Modulation of IFN- production by anti-IL-10 in naturally infected cattle from Great Britain. Blood samples from 31 field reactor cattle from herds with culture-confirmed bovine tuberculosis were stimulated in the absence or presence of MAb neutralizing bovine IL-10. Antigen-specific responses, determined by BOVIGAM ELISA, are shown, with medium control values subtracted.

SID 5 (Rev. 3/06) Page 17 of 23

Objective 06: Facilitate technology transfer

An advisory panel was created, chaired by the project leader, consisting of 2 Defra policy staff and 2 VLA staff members (one statistician, one from the Laboratory Testing Division responsible for the routine Bovigam testing). They recommended that cocktail PC2, derived from work under SE3028 [8,9], to be assessed in animals from whole herd slaughter operations to assess its sensitivity in detecting animals escaping skin detection and to test its relative specificity compared to culture and meat inspection. This was done (see below). However, the project leader was asked early in this project to report progress, and to make recommendations on project modifications, to VPAG, the Defra scientific steering committee overseeing TB vaccine and DIVA test development. The two defra staff that served on the advisory panel, are also VPAG members. Thus progress on SE3222 was reported six-monthly to VPAG. Since 2006, SE3222 updates were also given to another Defra committee on the IFN- testing. Thus the role of the advisory panel, as envisaged in objective 06, was now within the remits of these steering committees. To avoid further and unnecessary duplication, no separate advisory panel discussions were held as such discussions became part of the regular half-yearly VPAG meetings (and the IFN- committee meetings). At present Rv3615c field evaluation is viewed by these committees as a priority, as well as further validation of ESAT-6/CFP-10 with a view to obtain OIE validation of its use (a project proposal addressing these recommendations is in preparation).

The PC2 cocktail was tested in the IFN- test with 467 blood samples obtained from 3 herds that were candidates for whole or partial herd removal. Responses were compared to ESAT-6/cFP-10 and PPD-B and PPD-A. As can be expected from such high prevalence herds, high positivity rates were observed [PPD-B minus PPD-A: 18.4 % (86/467); ESAT-6/CFP-10: 15.2 % (71/467), PC2: 19.4 % (91/467)]. Thus, PC2 detected larger numbers of animals than either tuberculin and ESAT-6/CFP-10 (as we had initially reported [8]). However, whole herd removal operations were not authorized, and we were therefore not able to interpret the results in respect to detection of animals escaping skin testing, or to relative test specificity. However, other studies, reported under objective 4, are suggesting that the advantages of PC2 compared to ESAT-6/CFP-10 may be less than expected from pilot studies, and that the confirmed properties of Rv3615c in complementing ESAT-6 and CFP-10 (see obj. 04) made this protein the reagent of choice for further studies rather than PC2.

General lessons learned and future ways ahead:

Assessing the various antigen mining approaches taken, we can now draw the following conclusions: Bio-informatical approaches based on comparative transcriptome or genome analysis are very

useful to prioritize target genes for analysis. Such in silico approaches combined with the high-throughput peptide immunological screening system that we have developed, and allows rapid evaluation of a large number of proteins.

However, bio-informatical methods are at present unable to predict with high probability immunogenicity or antigenicity. We also do not know enough about the mechanisms of T cell X-reactivity to predict specificity without knowing the individual epitopes recognized by bovine T cells.

Therefore, immunological screening remains a vital part of antigen mining, and the notion that in silico approaches can in the foreseeable future replace in vitro or in vivo ‘wet’ experimentation is unrealistic.

When comparing the properties of the most immunogenic antigens defined in this and previous studies, it was apparent that secreted antigens and PE/PPE antigens were significantly over-represented in the proteins that displayed the highest immunogenicity values. This information will focus future studies.

Suggestions for future work are based on a triple-track approach to improved diagnostics which includes (i) a product-development arm pursuing incremental improvement to current tests using targeted approaches testing all secretome proteins (i.e. secreted proteins) and more unbiased library approaches (such as employing genome-wide gateway libraries, (ii) translational research to validate promising reagents in the field, and (iii) a basic research

SID 5 (Rev. 3/06) Page 18 of 23

arm to uncover new categories of infection in cattle by defining stage-specific antigens that may be recognized, for example during bacterial latency and to immune profile CMI responses in respect to their cytokine production profiles using such methods as cytokine multiplex analysis or FACS.

Annexe 1: Methods used throughout this project (if not referred to in the context of individual objectives).

CattleNaïve uninfected control animals were obtained from herds within 4 yearly testing parishes with no history of a BTB breakdown in the past 4 years and tested for the absence of an in vitro IFN response to avian tuberculin PPD (PPD-A) and bovine tuberculin PPD (PPD-B). Such animals were, when about 6 months old, experimentally infected with M. bovis (M. bovis field strain from Great Britain AF 2122/97 by intratracheal instillation of circa 2 x 103 CFU as described previously [7]) and for BCG vaccination. BCG vaccinated animals were vaccinated with 106 CFU of BCG Danish strain 1331 (Statens Serum Institute, Copenhagen, Denmark) according to the manufacturers instructions (reconstituted in Sautons medium and 1ml injected subcutaneously). In experiments where BCG vaccinated animals were challenged by intratracheal M.bovis, infection was carried out about 14 weeks after BCG vaccination, and results compared to identically infected matched naïve control animals. Blood samples were obtained also from naturally infected, tuberculin skin test-positive reactors within herds known to have BTB. All animals were additionally screened for an in vitro IFN-γ response to PPD-B and the presence or absence of a response to ESAT-6 and CFP-10 was recorded. These animals were housed at the VLA at the time of blood sampling. Infection was confirmed by necropsy and/or M. bovis culture.All procedures involving animals were carried out under a project licence granted by the GB Home Office under the Animals (Scientific Procedures) Act 1984. This project was approved by the local VLA Animal Ethics Committee prior to submission to the Home Office.

Production & Preparation of Antigens & PeptidesBovine tuberculin (PPD-B) and avian tuberculin (PPD-A) were supplied by the Tuberculin Production Unit at the Veterinary Laboratories Agency, Weybridge, Surrey, UK and were used to stimulate whole blood at 10g ml-1. Staphylococcal enterotoxin B was included as a positive control at 1g ml-1. Peptides representing our candidates were pin-synthesised as 20-mers spanning the length of all 14 proteins with each peptide overlapping its neighbour by 12 amino acid residues (Pepscan, Lelystad, Netherlands). These were dissolved in Hanks Balanced Salt Solution (Gibco) and 20% DMSO to 5mg ml-1 and grouped into pools of 8 to 12 peptides, with some genes represented by more than one pool. Pools were used to stimulate whole-blood at a final concentration of 10µg ml-1 total peptide. Peptides from the ESAT-6, CFP-10 and Rv3615c proteins were synthesised by conventional solid-phase synthesis technology, quality assessed and formulated into peptide cocktails as previously described.

IFN enzyme-linked immunosorbent assay (ELISA)Incubation of 250µl whole blood aliquots with antigen (as described above) was performed in 96-well plates. Plasma supernatants were harvested after 24 hours of culture at 37°C and 5% CO2

in a humidified incubator. The IFNγ concentration was determined using the BOVIGAM ELISA kit (Prionics AG, Switzerland). Results were deemed positive when the optical densities at 450 nm (OD450) with antigens, minus the OD450 without antigens, were ≥ 0.1. For comparative analysis of PPD-B versus PPD-A responses, a positive result was defined as a PPD-B OD450 minus PPD-A OD450 of ≥0.1, and a PPD-B OD450 minus unstimulated OD450 of ≥ 0.1.

Ex vivo IFNγ EliSpot assay

SID 5 (Rev. 3/06) Page 19 of 23

Peripheral blood mononuclear cells (PBMC) were isolated from heparinised blood taken from three cattle which had previously demonstrated an in vitro response to the Rv3615c peptide pool. Separation was performed using Histopaque 1077 (Sigma) gradient centrifugation and resuspended in RPMI 1640 tissue culture medium containing 25mM HEPES (Gibco), 10% Foetal Calf Serum, 1% non-essential amino acids, 5 x 10-5 M ß-mercapto-ethanol, 100U/ml penicillin and 100μg ml-1 streptomycin. Cells were enumerated and prepared to 2x106 cells ml-1.IFNγ production by PBMC was analysed using the Mabtech bovine IFNγ EliSpot kit (Mabtech, Stockholm, Sweden). The EliSpot plates (Multiscreen HTS-IP, Millipore) were coated at 4°C overnight with the bovine IFNγ specific monoclonal antibody after which the wells were blocked for two hours using 10% foetal calf serum in RPMI 1640. The primary antibody and blocking buffer were removed from the plates and PBMC suspended in tissue culture medium were then added (2x105 cells well-1) and cultured overnight at 37oC + 5% CO2 in the presence of the individual antigens. Stimulations were performed using the peptides at 5μg ml-1 or a pool of all 12 peptides at 5μg ml-1 of each peptide. Wells were washed using phosphate-buffered saline (PBS) plus 0.05% Tween 80. A secondary biotinylated antibody was used at 0.025 μg ml-1 followed by incubation with streptavidin-linked horse radish peroxidase. After a further wash, the spot-forming cells were visualised using the AEC Chromogen Kit (Sigma). Spots were counted using an AID EliSpot Reader and EliSpot 4.0 Software (Autoimmun Diagnostika, Germany).

Fluorescence-activated cell sorting (FACS) analysisPBMC were isolated from fresh heparinised blood as for the EliSpot, enumerated and prepared to 2x106 cells ml-1 and cultured overnight in a 24 well plate (Nunc) at 37oC + 5% CO2 with either RPMI medium (unstimulated control), PPD-B, pokeweed mitogen (positive control), individual peptides at 5μg ml-1 or a pool of all 12 peptides at 5μg ml-1. After incubation, Brefeldin A was added at 10μg ml-1 and allowed to incubate for a further 4 hours. The plate was centrifuged at 300g for 5 min, and the cells resuspended in a final volume of 250μl for transfer to a 96 well plate. Surface antibody staining was performed using AlexaFluor 647 conjugated anti-CD4 (Serotec, code MCA1653A627) and fluorescein-isothiocyanate conjugated anti-CD8 (Serotec, code MCA837F) antibodies. A differential “live/dead” stain was performed using Vivid (Invitrogen). After incubation for 15 min at 4oC, cells were washed and centrifuged before being permeabilised using Cytofix/Cytoperm (BD) at 4oC for 20 minutes, and stored overnight at 4oC. Intracellular staining for IFNγ was performed using phycoerythrin-conjugated anti-IFNγ (Serotec, code MCA1783PE) for 30 min at 4oC. Cells were finally suspended in 600μl of buffer and analysed using a Dako Cyan ADP and Summit 4.3 software (Dako, Switzerland).

Statistical analysisThe statistical software package GraphPad InStat v3. was used for the statistical analysis of IFNγ responses and responder frequencies. Comparison of IFNγ responses in naïve, infected and vaccinated cattle was performed using the non-parametric Wilcoxon rank sum test, with all p values corrected for multiple testing using the Bonferroni method. Comparison of the responder frequencies in infected cattle versus naïve and vaccinated cattle was performed using Fisher’s Exact test. Power analyses were performed using GraphPad StatMate 2.00 (San Diego CA).

Annexe 2: References cited in report

1. Garnier T, Eiglmeier K, Camus JC, Medina N, Mansoor H, et al. (2003) The complete genome sequence of Mycobacterium bovis. Proc Natl Acad Sci U S A 100: 7877-7882.

2. Cole ST, Brosch R, Parkhill J, Garnier T, Churcher C, et al. (1998) Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393: 537-544.

3. Brosch R, Gordon SV, Garnier T, Eiglmeier K, Frigui W, et al. (2007) Genome plasticity of BCG and impact on vaccine efficacy. Proc Natl Acad Sci U S A 104: 5596-5601.

SID 5 (Rev. 3/06) Page 20 of 23

4. Li L, Bannantine JP, Zhang Q, Amonsin A, May BJ, et al. (2005) The complete genome sequence of Mycobacterium avium subspecies paratuberculosis. Proc Natl Acad Sci U S A 102: 12344-12349.

5. Schnappinger D, Ehrt S, Voskuil MI, Liu Y, Mangan JA, et al. (2003) Transcriptional Adaptation of Mycobacterium tuberculosis within Macrophages: Insights into the Phagosomal Environment. J Exp Med 198: 693-704.

6. Sidders B, Pirson C, Hogarth PJ, Hewinson RG, Stoker NG, et al. (2008) Screening of highly expressed mycobacterial genes identifies Rv3615c as a useful differential diagnostic antigen for the Mycobacterium tuberculosis complex. Infect Immun 76: 3932-3939.

7. Vordermeier HM, Chambers MA, Cockle PJ, Whelan AO, Simmons J, et al. (2002) Correlation of ESAT-6-specific gamma interferon production with pathology in cattle following Mycobacterium bovis BCG vaccination against experimental bovine tuberculosis. Infect Immun 70: 3026-3032.

8. Cockle PJ, Gordon SV, Hewinson RG, Vordermeier HM (2006) Field evaluation of a novel differential diagnostic reagent for detection of Mycobacterium bovis in cattle. Clin Vaccine Immunol 13: 1119-1124.

9. Cockle PJ, Gordon SV, Lalvani A, Buddle BM, Hewinson RG, et al. (2002) Identification of novel Mycobacterium tuberculosis antigens with potential as diagnostic reagents or subunit vaccine candidates by comparative genomics. Infect Immun 70: 6996-7003.

10. Vordermeier HM, Whelan A, Cockle PJ, Farrant L, Palmer N, et al. (2001) Use of synthetic peptides derived from the antigens ESAT-6 and CFP-10 for differential diagnosis of bovine tuberculosis in cattle. Clin Diagn Lab Immunol 8: 571-578.

11. Dean GS, Rhodes SG, Coad M, Whelan AO, Wheeler P, et al. (2008) Isoniazid treatment of Mycobacterium bovis in cattle as a model for human tuberculosis. Tuberculosis (Edinb) 88: 586-594.

12. Mustafa AS, Skeiky YA, Al-Attiyah R, Alderson MR, Hewinson RG, et al. (2006) Immunogenicity of Mycobacterium tuberculosis antigens in Mycobacterium bovis BCG-vaccinated and M. bovis-infected cattle. Infect Immun 74: 4566-4572.

13. Buza JJ, Hikono H, Mori Y, Nagata R, Hirayama S, et al. (2004) Neutralization of interleukin-10 significantly enhances gamma interferon expression in peripheral blood by stimulation with Johnin purified protein derivative and by infection with Mycobacterium avium subsp. paratuberculosis in experimentally infected cattle with paratuberculosis. Infect Immun 72: 2425-2428.

For a list of publication generated during this project, as well as patent applications filed/submitted, see section 9.

SID 5 (Rev. 3/06) Page 21 of 23

References to published material9. This section should be used to record links (hypertext links where possible) or references to other

published material generated by, or relating to this project.

SID 5 (Rev. 3/06) Page 22 of 23

Denis M, Wedlock DN, McCarthy AR, Parlane NA, Cockle PJ, et al. (2007) Enhancement of the sensitivity of the whole-blood gamma interferon assay for diagnosis of Mycobacterium bovis infections in cattle. Clin Vaccine Immunol 14: 1483-1489.

Sidders B, Pirson C, Hogarth PJ, Hewinson RG, Stoker NG, et al. (2008) Screening of highly expressed mycobacterial genes identifies Rv3615c as a useful differential diagnostic antigen for the Mycobacterium tuberculosis complex. Infect Immun 76: 3932-3939.

Garcia Pelayo MC, Garcia JN, Golby P, Pirson C, Ewer K, et al. (2009) Gene expression profiling and antigen mining of the tuberculin production strain Mycobacterium bovis AN5. Vet Microbiol 133: 272-277.

Vordermeier HM, Gordon SV, Hewinson RG (2008). Antigen mining to define M. bovis antigen for the differential diagnosis of vaccinated and infected animal. Transboundary and Emerging Infectious Diseases, in press.

Vordermeier HM (2009). Development of cattle TB vaccines. In: Tuberculosis vaccines. Editors: A Acosta, ME Sarmiento, NM Nor (Havanna, Cuba), in press.

Three more publications are submitted, or in the final stages of preparation.

UK Patent application No. 0722105.4 (Tuberculosis Diagnosis)

SID 5 (Rev. 3/06) Page 23 of 23

Documents

General enquiries on this form should be made to:randd.defra.gov.uk/Document.aspx?Document=SE3222_8531... · Web viewAntigens that were prioritized by in silico analysis were then