General enquiries on this form should be made...

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 (2/05) of 31

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.

A SID 5A form must be completed where a project is paid on a monthly basis or against quarterly invoices. No SID 5A is required where payments are made at milestone points. When a SID 5A is required, no SID 5 form will be accepted without the accompanying SID 5A.

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 SE3209

2. Project title

Testing of Vaccine Candidates for Bovine Tuberculosis using a Low Dose Aerosol Challenge Guinea Pig Model

3. Contractororganisation(s)

Prof Glyn Hewinson,TB Research Group,VLA Weybridge,New HawSurreyKT15 3NB

54. Total Defra project costs £ 1,068,045

5. Project: start date................ 01 July 1999

end date................. 30 June 2004

SID 5 (2/05) of 31

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.Bovine tuberculosis remains an economically important problem in Great Britain with potential zoonotic consequences. As such, Defra continues to have a statutory obligation to control tuberculosis in farm animals in Great Britain under the Animal Health Act of 1981, the Tuberculosis Orders, and various EC directives. Despite implementation of a test and slaughter strategy using the tuberculin skin test to detect infected animals, the number of cases of bovine tuberculosis in cattle has been increasing by 18% per annum since 1988. By 2002 about 4% of the national herd was under restriction and in 2003 bovine tuberculosis resulted in a direct annual cost to the government of ca. £80 million. This is thought to be due, at least in part, to the relatively high occurrence of M. bovis infection in badgers, which act as reservoirs of infection in this area. In 1996, an independent scientific commission chaired by Professor John Krebs to review the situation of bovine TB in GB concluded that the development of a cattle vaccine and associated diagnostic test had the best prospect of controlling the disease in the National Herd and that the option of badger vaccination should be retained. It was envisaged that an effective cattle vaccine would have a significant impact in reducing the economic burden of the TB control programme and the psychological and economic hardship inflicted on the farming industry. This conclusion has been re-affirmed in the House of Commons Environment, Food and Rural Affairs Committee’s report on Bovine TB (2004) and by the findings of the Independent Scientific Group Vaccine Scoping Sub-committee, which highlighted that work on development and testing of vaccines should be maintained in order to produce a vaccine that is more effective than BCG in cattle. In the Krebs report it was recommended that in the first five years of a vaccine development programme candidates should be generated and tested in laboratory animal models.

The aim of this 5 year project was to test vaccine candidates in the low dose aerosol challenge guinea pig model so that suitable candidates could be taken forward for testing in cattle. The model was developed in collaboration between VLA and Health Protection Agency, Porton Down, Salisbury (formerly CAMR, Porton Down). All the aerosol challenge work was performed at Health Protection Agency, Porton Down.

The development of effective vaccine candidates is still by and large an empirical process especially against ‘stealth pathogens’ such as TB, HIV and malaria and there is no guarantee of success. However, there are a number of approaches to vaccine development which have proved successful against other organisms and the aim of this project was to test vaccines developed as part of SE3028 in the guinea pig model. Vaccines candidates were developed in SE3208 by exploiting a number of recent advances in (i) vaccinology, (ii) manipulation of M. bovis DNA, (iii) the understanding of the immune response against M. bovis infection and (iv) the recently completed genome sequences of M. tuberculosis, M. bovis (SE3206) and BCG (SE3206).

SID 5 (2/05) of 31

The first approach that we took, based on the BCG paradigm, was the attenuation of M. bovis to produce a live TB vaccine. In this approach key genes essential for the growth and survival of M. bovis within the host are inactivated, thereby disabling the micro-organism. For a successful vaccine, the disabled micro-organism will be sufficiently robust to induce a protective immune response but be contained or eliminated such that the disease process is self-limited. A number of methods were used to generate live, attenuated vaccines including transposon mutagenesis, illegitimate recombination and targeted mutagenesis to knock out key genes identified by either genomics, proteomics or detailed study of important metabolic pathways. Early studies with auxotrophic mutants i.e. those attenuated due to disruption of key metabolic genes suggested that protective immunity to tuberculosis may, at least in part, be achieved without sensitisation to the tuberculin skin test. We therefore investigated this phenomenon further using either mutants that persisted for different times within the host or by using different doses of the attenuated vaccine. Our results suggest that either persistence of the organism, replication of the organism or high doses of an organism that is subsequently cleared are required to stimulate a strong enough T cell response to sensitise animals to the tuberculin skin test and provide good protective immunity. Armed with this knowledge we then targeted genes involved in virulence of M. bovis that might give improved protection over auxotrophic strains of BCG. These approaches generated a number of promising vaccine candidates that gave at least as good protection as BCG against M. bovis challenge in guinea pigs. These included a knockout mutant of the M. bovis glycosylated lipoprotein, MPB83 and an M. bovis mutant lacking the original attenuating lesion of BCG, RD1. Future work will be required on these new vaccines: 1.) to confirm the vaccine efficacy results 2.) to better characterise their stability and genetic functionality, 3.) to investigate their vaccine potential in hosts other than guinea pigs, 4.) to characterise the host response after vaccination and challenge in appropriate hosts and 5.) to remove antibiotic resistance genes and incorporate another attenuating mutation. However we recommend that both vaccines should be tested in cattle. Of particular priority would be the M. bovis RD1 knockout mutant because vaccination of cattle with this recombinant strain would also allow differential diagnosis using RD1 antigens that include ESAT6 and CFP10 which have shown promise as diagnostic antigens for use in the Bovigam gamma interferon assay. In this way vaccinated cattle could be differentiated from animals infected with M. bovis.

One of the recommendations of the WHO/FAO/OIE consultation on animal tuberculosis vaccines in 1994 (WHO/CDS/VPH/94.138) was that the efficacy of all candidate vaccines for bovine tuberculosis should be compared against that of BCG Pasteur. Therefore, the efficacy of BCG vaccination against disease induced by M. bovis was always assessed in the guinea pig alongside any vaccine candidate. We also compared the efficacy of BCG Pasteur with other strains of BCG and demonstrated no statistical difference between the protection conferred by these strains.

The second approach to generating candidate vaccines for M. bovis was based on induction of immune responses to mycobacterial components delivered in the form of a subunit vaccine. This approach is preferable in relation to quality control and safety considerations and has produced a number of effective vaccines against diseases other than TB. In this project we tested killed vaccines in adjuvant and DNA and protein subunit vaccines based on antigens that had been identified as immunodominant in cattle as part of projects SE3212, SE3028 and SE3222. Before testing protein vaccine candidates in guinea pigs we first identified adjuvants suitable for use in guinea pigs. Both DNA vaccines and protein subunit vaccines proved disappointing and not one candidate gave protection equivalent to BCG or better.

Better success was achieved using killed organisms. In these experiments formalin-inactivated BCG was mixed with non-phospholipid liposome adjuvants (Novasomes) and administered to guinea pigs as a single subcutaneous inoculation. All formulations were well tolerated and one conferred a significant survival advantage against lethal aerogenic challenge with M. bovis. We have therefore provided preliminary evidence that it is possible to achieve protection against TB using a killed whole mycobacterial cell vaccine administered once in an appropriate adjuvant. These vaccine formulations are cheap and safe to produce (being already based on BCG) and had no reactogenicity in the naive guinea pig. Future studies should address titration of the vaccine dose, identification of the optimum Novasome™ composition and the effect of boosting. Given that the profile of expressed mycobacterial proteins alters during their culture in liquid media, there is scope to optimize the conditions in which the BCG are grown prior to harvest and formalin inactivation and in this way maximize the number and concentration of protective antigens in the final preparation. We also recommend that the lead candidate from this study should be taken forward for assessment and optimisation in cattle.

In our final approach we combined both approaches described above in a vaccine strategy based on priming the immune response with one type of vaccine and then boosting with the other (a heterologous prime-boost vaccination strategy). First we tested a combination of DNA vaccination followed by boosting with a recombinant vaccinia virus, MVA, expressing the same antigens as the DNA vaccine. This demonstrated that the use of heterologous prime-boost immunisation regimes did improve upon the protection conferred by single subunit vaccines and highlighted a clear strategy for future TB vaccine development. We then sought to improve the efficacy of BCG by boosting BCG-induced immunity with

SID 5 (2/05) of 31

subunit vaccines. Two advances in our vaccine development programme at VLA had allowed us to pursue this strategy. First, we had developed specific diagnostic tests that differentiated between cattle that had been vaccinated with BCG and animals that were infected with M. bovis using antigens that are present in M. bovis but absent in BCG (see SE3212, SE3028 & SE3222). Second, we had demonstrated that a cocktail of DNA vaccines could enhance the protection afforded by BCG in cattle (SE3212). In this current project, the vaccines that we used to boost BCG immunity in guinea pigs were the same cocktail of DNA vaccines that had proved efficacious in cattle i.e. DNA vaccines encoding HSP65, HSP70 and Apa. The results showed that the cocktail of DNA vaccines used alone had no protective effect against M. bovis challenge. Furthermore the guinea pig model was unable to predict the results obtained in cattle (and mice) i.e. that priming with DNA vaccine could improve the protective efficacy of BCG in an heterologous prime-boost vaccination strategy. For this reason we decided that in future the mouse model and not the guinea pig model should be used to screen vaccines for use in prime-boost strategies to improve the protective efficacy of BCG. This is now our strategy for testing vaccine candidates and has been taken forward in SE3224.

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).

The aim of this 5 year project was to test vaccine candidates in the low dose aerosol challenge guinea pig model so that suitable candidates could be taken forward for testing in cattle.

Rationale for the guinea pig model

The need to define protective immunity to M. bovis and to develop alternative models to evaluate the efficacy of new-generation vaccines is well recognised. The suitability of the animal model for screening candidate vaccines is central to any programme of vaccine development. The low-dose aerosol challenge guinea pig model is the model of choice for testing vaccines against tuberculosis because: the guinea pig is highly susceptible to M. bovis infection and can therefore be used to evaluate the level of

attenuation of auxotrophic vaccines based on virulent M. bovis or M. tuberculosis (guinea pigs were used by Calmette and Guerin to show that BCG was avirulent);

the degree of protection induced by vaccination with BCG is excellent; due to their exquisite sensitivity to M. bovis infection, guinea pigs may be used to discriminate between

various degrees of protection; an aerosol challenge model of guinea-pig infection is currently being used by the NIH vaccine screening

programme for testing the efficacy of new vaccines against M. tuberculosis.

The model was developed for M. bovis by Health Protection Agency, Porton Down (formerly CAMR, Porton Down) and VLA and is described in: Chambers MA, Williams A, Gavier-Widen D, Whelan A, Hughes C, Hall G, Lever MS, Marsh PD, Hewinson RG. A guinea pig model of low-dose Mycobacterium bovis aerogenic infection. Vet Microbiol. 2001 80:213-26.

Objective 03 Correlates of protection suitable for high-throughput vaccine screening determined in the guinea pig using BCG Pasteur.

Whilst the Animal Models Task Force of IMMYC have recommended that infection of the lung and haematogenous seeding of the spleen with challenge organisms should be used as indicators of vaccine efficacy,

SID 5 (2/05) of 31

a number of other potential readouts are available from the guinea-pig low-dose aerosol challenge model. These include reduction in histopathology and long-term survival. At the beginning of this project we envisaged that not all read outs would be instructive or permit high-throughput screening of vaccine candidates and that we might need to include additional or alternative parameters to allow identification of promising vaccine candidates that might be missed in the short term model that defines protection in terms of bacterial burden 5 weeks after challenge.

In our experiments to optimise the model the following readouts were assessed: temperature, weight, Karnofsky scale of pain/distress, root index of virulence, histopathology, lung weight and density, bacteriology in spleen and kidney, number of lung lesions, size of tracheobronchial lymph nodes and survival.

In general we found that either bacterial burdens in lung and spleen 5 weeks after challenge or survival time post-challenge were the most useful readout systems with which to judge vaccine efficacy. We found that survival time after challenge was the most suitable readout system for identifying candidates that protect guinea pigs against M. bovis challenge better than BCG. Unfortunately this method for assessing vaccine efficacy precludes the use of bacterial burden as a read out system because at time of death most guinea pigs have a similar bacterial burden.

This is in line with the readout systems used for testing vaccine efficacy against M. tuberculosis in guinea pigs. At the time of writing the EU TB Cluster have switched from using a short term challenge model that uses colony counts as a read out to a long term challenge model that uses survival as a read out in order to identify candidates that protect better than BCG (Williams et al, 2005 Tuberculosis 2005 85:29-38). In addition, we found that loss of body weight was a good indicator of disease progression. This parameter was used to determine the time at which guinea pigs should be killed to determine survival time (i.e. when the weight of the animals was 20% less than their initial body weight).

Objectives 06, 07, 08, 09 Evaluation of Vaccine Candidates

Live, attenuated vaccines: Homologous boosting with auxotrophic BCG.

We have previously reported that a leucine auxotroph of BCG conferred significant protection of the lungs and spleens of guinea pigs infected with M. bovis in the absence of a cutaneous hypersensitivity reaction to tuberculin (Chambers et al., 2000). This suggested that protective immunity to tuberculosis might, at least in part, be achieved without sensitisation to the tuberculin skin test. We therefore tested whether boosting with the leucine auxotroph could improve protection against infection to that obtained for BCG. Guinea pigs received one, two, or three doses of leuD auxotrophic vaccine and were compared with animals given a single dose of BCG Pasteur. There was a three-week interval between boosts and all animals were challenged five weeks after the last immunisation in each group. The animals were killed 10 weeks later (or in extemis) and the spleen and lungs cultured for M. bovis. There was no evidence that boosting gave an additive effect on protection. However, there was a slight survival advantage seen in the group that received three doses of leuD, compared with one or two doses. This difference was not significant statistically.

Local and systemic persistence of the leucine auxotroph. Guinea pigs were inoculated subcutaneously with 5 x 104 cfu BCG Pasteur or 5 x 104 cfu of the leucine auxotrophic strain of BCG Pasteur. Vaccinated guinea pigs were killed one day after inoculation and then at seven day intervals until day 28. Spleens and the lymph node draining the site of inoculation were taken for culture. BCG Pasteur was cultured from the deep inguinal draining lymph node (DLN) at all time points up to day 28. In contrast, the leucine auxotroph was only cultured from the DLN on days one and seven. This strain was never cultured from the spleen at any time point. In contrast, low numbers (<100 CFU) of BCG Pasteur were cultured from the spleen at time points up to and including day 21. By day 28, neither strain was cultured from the spleen. These data are consistent with persistence data generated by others in mice using these strains (McAdam et al. Infect Immun 1995 63:1004-12) and support the contention that persistence and ability to sensitise for a tuberculin reaction are related (see below).

Relationship between inoculation dose and DTH sensitisation. As mentioned above, we have previously demonstrated that a leucine auxotroph of BCG conferred significant protection of the lungs and spleens of guinea pigs infected with M. bovis in the absence of a cutaneous hypersensitivity reaction to tuberculin. However, protection was not as great as that observed for BCG and we postulated that the auxotrophic strain was cleared before it could stimulate sufficient circulating memory T cells to give rise to a cutaneous hypersensitivity reaction. We therefore performed an experiment to establish the relationship between dose of live vaccine and sensitisation to the tuberculin skin test. Guinea pigs were sensitised with BCG Pasteur or the leucine auxotroph over a range of inoculation doses. For the leuD auxotroph this was seven doses ranging from 3x10 3 CFU to 3x106

CFU. For BCG Pasteur this was five doses ranging from 1 CFU to 1x104 CFU. Five weeks later the animals were tuberculin tested and the size of the DTH reaction measured 24 and 48 hours later. For the leuD auxotroph, guinea pigs receiving 6x104 CFU and higher were sensitised to tuberculin. No DTH reaction to tuberculin was seen in guinea pigs receiving 3x104 CFU or lower. In contrast, guinea pigs receiving BCG Pasteur at 10 CFU or

SID 5 (2/05) of 31

higher were sensitised to tuberculin. Only the animals receiving a calculated 1 CFU were not sensitised. The same observations were made at both the 24 and 48 hours time points. Interpolation of data points suggested that the DTH reaction size observed at 48 hours in guinea pigs receiving 103 CFU BCG Pasteur would equate to that seen with 6x105 CFU leuD auxotroph. These data (and the persistence data also generated by this project - see above) suggest that either persistence of the organism, replication of the organism or high doses of an organism that is subsequently cleared are required to stimulate a strong enough T cell response to sensitise animals to the tuberculin skin test and provide good protective immunity.

Evaluation of other auxotrophic strains.

The strains examined in the vaccine-testing programme thus far all have gene disruptions marked with an antibiotic resistance gene and some can revert to wild-type. These characteristics make the current strains undesirable. Therefore we next determined the protective efficacy of a lysine auxotroph of BCG substrain Pasteur. This mutants has a non-revertible and non-suppressible deletion in the lysA gene, encoding diaminopimelate decarboxylase, the last enzyme in the lysine biosynthetic pathway (Pavelka and Jacobs, J Bacteriol. 1999 181:4780-9.). Moreover, this strain does not carry any antibiotic resistance markers. The lysine mutant does not replicate in the mouse but has been shown to enter a physiological state that prevents the cells from losing viability and results in the gradual loss of cells by slower immune clearance mechanisms (Pavelka et al., Infect Immun. 2003 71:4190-2 and personal communication). This persistence gives an improved immune response to vaccination over the leuD mutant that results in sensitisation to the tuberculin skin test.

It has been suggested that the inability to boost BCG in most animals may be due to the inability of BCG vaccines to replicate in an immunized host and that this could be overcome by using a booster vaccine that does not depend upon replication. Since the lysA auxotrophic mutant does not replicate in the host, it is possible that it could be used as a booster vaccine. Studies in mice had shown that that a single immunization with the lysine auxotroph did not generate an immune response capable of significantly restricting the growth of virulent M. tuberculosis Erdman following an aerogenic challenge. However, administration of a second or a third dose of this vaccine increased protection substantially, as measured by the number of viable bacteria per organ, to a level similar to that achieved with a single dose of BCG-Pasteur (Pavelka et al., Infect Immun. 2003 71:4190-2 and personal communication). This level of protection did not seem to be greatly increased by a third dose of vaccine.

We therefore compared the protective efficacy of the lysA mutant with that of the leuD mutant against M. bovis challenge in guinea pigs. We also tested whether vaccination with a second dose of vaccine could confer increased protection over a single vaccine dose. Guinea pigs (8 per group) were vaccinated with 5 x 104 cfu of BCG Pasteur, BCG Pasteur lysA or BCG Pasteur leuD. Relevant groups were boosted 5 weeks after the first immunisation. All groups were challenged with 10 cfu M. bovis 5 weeks after the final immunisation and killed 10 weeks after challenge. An end point of ten weeks post challenge was chosen in this experiment in an attempt to increase the window with which to detect a boosting effect. We therefore used numbers of lung lesions and bacterial burden in spleen as parameters of protection as it was not feasible to take bacterial counts and look at detailed lung pathology in one experiment. This approach was consistent with guinea pig work being performed as part of the NIH human TB vaccine testing contract which had shown that differences in lung pathology at later time points in the infection process might be a more useful predictor of vaccine efficacy than bacterial burden in the lung at 5 weeks post-infection (Baldwin et al 1999 Infect. Immun. 66: 2951-2959).Results showed that neither auxotroph conferred better protection than BCG and that, unlike previous experiments, there was no boosting effect observed for any of the vaccine candidates. The results also confirmed that it is not possible to boost BCG-mediated protection with a second dose of BCG. Although not statistically significant there was a trend towards less protection, as measured by bacterial burden in the spleen and number of lesions in the lung, in the groups that received a second dose of vaccine (Figures 1 & 2 below).

SID 5 (2/05) of 31

SID 5 (2/05) of 31

Evaluation of the virulence of live vaccine candidates in the guinea pig. Having demonstrated that auxotrophic strains of BCG could not confer protection equal to that of BCG in the absence of a skin test response we next sought to develop attenuated strains of M. bovis that might provide better protection than BCG. A number of vaccines were produced in pursuit of this goal and tested for virulence in guinea pigs before the efficacy of suitably attenuated vaccine candidates was assessed in the aerosol challenge model.

Rationale for live vaccines tested: Most of the gene targets were chosen on the basis of genetic differences between M. bovis and M. tuberculosis identified by comparative genomics or transcriptomics (SE3030) on the basis that genetic differences would indicate genes that were key to the survival of M. bovis in its natural host since M. tuberculosis H37Rv does not cause pathology and is attenuated in cattle (see SE3030).

M. bovis mpb83. One of the major phenotypic differences between M. tuberculosis and M. bovis is that M. bovis constitutively expresses high levels of a 25 kDa protein of unknown function, MPB83, whereas M. tuberculosis does not. MPB83 is a major antigenic target in cattle and badgers infected with M. bovis. We therefore postulated that MPB83 might be a virulence factor for M. bovis. Moreover, removal of this gene would mean that animals vaccinated with a vaccine lacking MPB83 could be differentiated from those infected with M. bovis if MPB83 were used as a diagnostic antigen. The mpb83 knockout mutant of M. bovis was generated as part of SE3208 and is described in the final report for this project.

To assess whether the mutant was attenuated, 100 colony-forming units (CFU) of M. bovis BCG Pasteur, M. bovis 2122/97 and M. bovis ∆mpb83 were introduced by intramuscular injection in guinea pigs. After 6 weeks the guinea pigs were sacrificed, the spleens and lungs removed, homogenized, and plated onto modified 7H11 plates to calculate the CFU present in each tissue sample. Results showed that that M. bovis ∆mpb83 was attenuated for growth in the guinea pig, with low CFU counts (Table 1) that were backed up by poor pathology seen in animals infected with the mutant (Figure 3). This mutant was therefore brought forward for testing as a live vaccine candidate (see below).

Figure 3: In vivo growth of M. bovis mpb83 mutant and wild-type in guinea pig model.

The attenuation of the recombinant M. bovis mpb83 was also supported by pathology scores at post mortem.

Table 1: Guinea pig gross pathology scores for BCG Pasteur, M. bovis AF2122/97 and M. bovis mpb83

Strain Local LNs Spleen scores Liver score Lung score

BCG 0 0 0 0

BCG 0 0 0 0

BCG 0 0 0 0

Average 0 0 0 0

SID 5 (2/05) of 31

2122/97 3 3 1 0

2122/97 3 3 0 0

2122/97 3 2 0 0

Average 3 2.7 0.3 0

mpb83 1 1 0 0

mpb83 1 0 0 0

mpb83 1 0 0 0

mpb83 1 0 0 0

mpb83 0 0 0 0

mpb83 0 0 0 0

Average 0.7 0 0 0

Spleen scoring system:0 = No lesions1 = <10 F2 = >10 F or <10 RN3 = >10 RNF = flat lesions, R = raised nodular lesions.

Lymph node scoring system: Liver and lung scoring system:0 = Not enlarged. 0 = No visible lesions1 = enlarged 1 = <10 lesions2 = grossly enlarged – not purulent. 2 = >10 lesions3 = grossly enlarged – purulent.

M. bovis Mb3909c As an initial focus for comparative genomics we focussed on the RD1 region which is deleted in all BCG strains and linked to the attenuation of the vaccine strain. One gene in RD1, Rv3879c/ Mb3909c, showed sequence divergence between M. tuberculosis H37Rv, CSU#93 and M. bovis AF2122/97. Hence, we chose to determine whether mutation of Mb3909c could effect the virulence of the mutant. The mutant was constructed as part of project SE3208 and the knockout strain was designated M. bovis ∆Mb3909c. To assess whether the mutant was attenuated, 100 colony-forming units (CFU) of M. bovis BCG Pasteur, M. bovis 2122/97 and M. bovis ∆Mb3909c and controls were introduced by intramuscular injection in guinea pigs. After 6 weeks the guinea pigs were sacrificed, the spleens and lungs removed, homogenized, and plated onto modified 7H11 plates to calculate the CFU present in each tissue sample. However, the CFU counts showed that the M. bovis ∆Mb3909c mutant was not attenuated in the guinea pig model; therefore progressing the mutant to vaccine efficacy studies was not warranted.

This work was published in Inwald J, Jahans K, Hewinson RG, Gordon SV. Inactivation of the Mycobacterium bovis homologue of the polymorphic RD1 gene Rv3879c (Mb3909c) does not affect virulence. Tuberculosis (Edinb). 2003;83:387-93. Further details may be found in this publication.

M. bovis RD1.A single region (RD1), which is absent in all BCG substrains, and linked to the attenuation of BCG was deleted from virulent M. bovis by our collaborator Prof WR Jacobs, Albert Einstein College of Medicine, Bronx, USA, who also demonstrated the resulting knock out mutant was significantly attenuated for virulence in both immunocompromised and immunocompetent mice (Hsu T, et al., Proc Natl Acad Sci U S A. 2003 100:12420-5.). This mutant was supplied to us and tested for attenuation in the guinea pig. A group of 6 guinea pigs was immunised sub-cutaneously with 5 x 104 cfu of the RD1 knock-out mutant. No recombinant BCG could be cultured from the lungs or spleen of guinea pigs 25 weeks after vaccination. Furthermore, there was no gross pathology in the organs of these animals at the time of post mortem. This strain was therefore e brought forward for testing as a live vaccine candidate (see below).

SID 5 (2/05) of 31

M. tuberculosis PlcABCD: Phospholipases C play a role in the pathogenesis of several bacteria. Mycobacterium tuberculosis possesses four genes encoding putative phospholipases C, plcA, plcB, plcC and plcD. M. bovis is a natural mutant lacking plcA, B and C but has retained the gene encoding plcD. The growth kinetics of the triple and quadruple mutants in a mouse model of infection has revealed that both mutants are attenuated in the late phase of the infection (Raynaud C, et al., Mol Microbiol. 2002 45: 203-17). The aim of testing these mutants in guinea pigs was to determine if plcD was a virulence factor for the M tb complex and, if so, whether the plcABCD mutant would be a suitable vaccine candidate against M. bovis infection. However, in contrast to the results in mice, the plcABC and plcABCD mutants were not attenuated in guinea pigs.

M. bovis UmaA1. Glickman and colleagues have shown that the inactivation of the umaA2 gene, which is adjacent to umaA1, effects such diverse phenotypes as cording, virulence and persistence in M. tuberculosis (Glickman MS et al., Mol Cell. 2000 5:717-27). We postulated that an M. bovis mutant that had been disrupted in the umaA1 gene by transposon mutagenesis as part of SE3208 would also effect such phenotypes in M. bovis. Using the same protocol as that described for the M. bovis mpb83, we found that the umaA1 mutant was also significantly attenuated in guinea pigs but to date it has not been tested as a vaccine candidate.

Evaluation of the efficacy of live attenuated vaccine candidates in the guinea pig.

M. bovis mpb83 (SM55).

The protective efficacy of the M. bovis mpb83 was assessed according to the following schedule. Animals (groups of 8) were immunized subcutaneously with 5x104 CFU of the either M. bovis mpb83 (designated SM55) or BCG Pasteur. A control group of animals was also vaccinated subcutaneously with PBS. Eight weeks after immunization, animals were challenged with 10 cfu M. bovis AF2122/97 using our standardised aerosol challenge. Animals were weighed weekly and those that lost 20% of their maximal body weight were killed. All surviving animals were killed thirty-five days after challenge. The spleens were taken for CFU determination and the left apical & caudal lung lobes and right apical and cardiac lung lobes taken for CFU determination. The remaining lung lobes were placed in formalin for histology.

Results

The M. bovis mpb83 mutant conferred protection against M. bovis challenge that was at least as good as BCG as measured by survival time (Table 2), or bacterial buden in the lung and spleen (Figure 4).

Table 2: Survival of guinea pigs to day 35 (termination of experiment)

Group Alive Dead Significance (Fisher’s Exact Test)

BCG

SM55

PBS

7

8

1

1

0

7

P < 0.02

P < 0.002

NA

Figure 4. Colony-forming units (CFU) of M. bovis harvested from the lung and spleen of individual guinea pigs vaccinated with (PBS), live (BCG) Pasteur, or M. bovis mpb83 (SM55). Bacteriology was performed on the organs of animals killed 35 days after aerogenic challenge with M. bovis 2122/97 or earlier if the animal had lost 20% of its maximal body weight

SID 5 (2/05) of 31

SID 5 (2/05) of 31

M. bovis RD1.

The protective efficacy of the M. bovis RD1 was assessed according to the following schedule. Animals (groups of 24) were immunized subcutaneously with 5x104 CFU of the either M. bovis RD1 or BCG Pasteur. A control group of 16 animals was also vaccinated subcutaneously with PBS. Twenty weeks after immunization, animals were challenged with 10 cfu M. bovis AF2122/97 using our standardised aerosol challenge. Five weeks later, 16 animals from the vaccinated groups and 8 animals from the PBS control group were killed. The spleens were taken for CFU determination and the left apical & caudal lung lobes and right apical and cardiac lung lobes taken for CFU determination. The remaining lung lobes were placed in formalin for histology.

The remaining animals (eight per group) were left to generate survival data. They were weighed weekly and animals that lost 20% of their maximal body weight were killed.

Results

Protection at 5 weeksBased on bacterial load in the lungs and spleen 5 weeks after challenge, both BCG and the RD1 KO conferred significant protection compared with the saline control. There was no difference in the protection between BCG and the RD1 KO in either organ.

Longer-term protectionWe used a 20 week vaccination-challenge interval in an attempt to skew the M. bovis aerosol challenge model to diminish the protection obtained with BCG, and so provide an opportunity for the RD1 KO vaccine to demonstrate superior protection.

Bacterial loads in the lungs and spleen of guinea pigs at the time of death were similar, with no significant difference between the three groups.

BCG did not prolong survival compared with the saline control (by the Log-Rank test), although the median survival time was significantly extended. In contrast, survival obtained with the RD1 KO was better than BCG, both in terms of comparison against unvaccinated animals, and in comparison with BCG itself – extending median survival time by a further 24% (20 days) (Figure 5).

Figure 5: Results of guinea pig survival comparing BCG with the RD1 KO strain against aerosol challenge with M. bovis

Median survival: saline = 55 daysBCG = 84 daysRD1 = 104 days

ConclusionThe RD1 KO vaccine conferred equivalent protection to BCG in the short term and showed some evidence of improved long-term survival compared with BCG although this was not the result of reduced bacterial loads at the time of death. Vaccination of cattle with this recombinant strain would also allow differential diagnosis using RD1

SID 5 (2/05) of 31

0 14 28 42 56 70 84 98 1120

20

40

60

80

100 salineBCGRD1

Days

Perc

ent s

urvi

val

antigens that include ESAT6 and CFP10. We recommend that this vaccine should be taken forward for testing in cattle.

Killed Vaccines

Vaccines based on killed whole cell preparations of mycobacteria should be safe and would present a wide repertoire of antigens to the recipient. However, such killed vaccines have conventionally conferred little to no specific protection to subsequent challenge with virulent mycobacteria, either because the important protective antigens are only expressed when the bacteria are metabolically active or grown under appropriate conditions, or possibly through failure to induce T-cells with effector function. There are notable instances of protection against TB using killed vaccines (e.g. H. Spahlinger, Lancet 1 (1922), pp. 5–8.) and it has been proposed that, in terms of generating protective immunity, the particular antigens that are presented may be less important than the way in which they are presented. The majority of vaccination studies on killed preparations of mycobacteria have used heat as the method of killing. However, this treatment may significantly denature important antigens and could account for the disappointing results generally seen with such vaccines. An alternative to heat inactivation is treatment with formalin. Formalin treatment of whole mycobacteria has the advantage of killing the organism whilst retaining the antigenic integrity of many of the proteins present.

We therefore revisited the potential of killed whole-cell vaccines by comparing their efficacy with live BCG Pasteur in a guinea pig challenge model. BCG Pasteur was inactivated with a low concentration of formalin and showed to be non-viable in culture or severe combined immunodeficient (SCID) mice. Formalin-inactivated BCG was mixed with non-phospholipid liposome adjuvants (Novasomes) and administered to guinea pigs as a single subcutaneous inoculation. All formulations were well tolerated and one conferred a significant survival advantage against lethal aerogenic challenge with M. bovis.

Formulation of Novasomes with adjuvantsFormalin-treated BCG were mixed with the following negatively-charged Novasome™ (non-phospholipid liposome) proprietary adjuvants (Novavax, Inc., MD: http://www.adjuvants.net/): NAX 57 (fusogenic; made with polyoxyethylene-2-cetyl ether (Brij 52)); NAX M57 (fusogenic; made with polyoxyethylene-2-cetyl ether plus monophosphoryl lipid A (MPL)); NAX M77 (fusogenic; made with polyoxyethylene-2-stearyl ether (Brij 72) plus MPL); NAX M687 (non-fusogenic; made with glycerol monostearate and batyl alcohol plus MPL). The cells were resuspended in the adjuvant using a tuberculin syringe fitted with an 18 gauge needle and then fully homogenized by passing the suspension back and forth between two tuberculin syringes forty times. A control preparation of killed BCG was treated in the same way but using sterile water in place of the adjuvant.

Vaccination and challenge of guinea pigsFemale Dunkin–Hartley guinea pigs weighing between 350 and 450 g and free of intercurrent infection were obtained from David Hall & Partners, Burton-on-Trent, UK. Groups of six guinea pigs were immunized with 100 ml of each formalin-treated vaccine/adjuvant formulation subcutaneously in the nape. Two control groups of six guinea pigs each were vaccinated with 5x104 CFU live BCG Pasteur or PBS alone. Five weeks after vaccination all guinea pigs were challenged aerogenically with a live suspension of M. bovis AF2122/97 to achieve an inhaled retained dose in the lungs of approximately 10 organisms. Guinea pigs were exposed to M. bovis aerosol for exactly five minutes in batches of eight, randomized for the vaccine treatment they had received.

ResultsNovasomes™ of different composition were chosen with the addition of MPL in some and MPL plus batyl alcohol in another. MPL was chosen for its ability to activate macrophages in the context of liposomes, as well as its adjuvant properties with mycobacterial vaccines. Batyl alcohol is a mono-octadecyl ether of glycerol isolated from shark liver and has been shown to have macrophage activating properties. No adverse reactions were observed at the vaccination site of any animal throughout the experiment. One guinea pig which received formalin-killed BCG in NAX M687 adjuvant died 5 days prior to challenge but there was no suggestion this was due to vaccination. Following challenge with M. bovis, the weights of individual animals were recorded twice weekly. Varying numbers of guinea pigs in each group had to be killed before the end of the experiment because they had reached the humane endpoint (20% loss in maximal body weight). Figure 6 shows the survival curves for each group. The first guinea pigs from the PBS control group were killed 39 days post-challenge and all animals in this group had been killed by day 63. By comparison, all animals vaccinated with formalin-killed BCG in NAX M687 adjuvant survived to the end of the experiment (70 days post-challenge) and only one animal vaccinated with live BCG had to be killed before then (at day 66). Table 3 summarizes the number of animals in each treatment group that survived to the end of the experiment. The proportion of animals surviving in each group were significantly different from one another (P<0.02, Chi-square test), so the significance for each group was tested against the PBS control group. To take into account the number of pairwise analyses, the P-value was set at 0.008 (0.05/6). Using this stringent level of significance only killed BCG in NAX M687 conferred a significant survival advantage against lethal challenge when expressed as percentage survival. Table 3 also shows the median survival time (MST) for each group. This differed significantly between groups (P=0.003, Kruskal–Wallis test). Comparing the

SID 5 (2/05) of 31

MST for each vaccine group with that of the PBS control group revealed a significant extension of the median survival in those animals vaccinated with live BCG or killed BCG in NAX M687 (P<0.01 and P<0.001, respectively by Dunn’s multiple comparison test).

Figure 6. Survival of guinea pigs following aerogenic challenge with M. bovis 2122/97 on day 0. Guinea pigs were inoculated 5 weeks earlier with vaccines comprised of formalin-killed BCG Pasteur formulated with sterile water (none) or different Novasome™ non-phospholipid liposome adjuvants (NAX), as indicated (open squares). For comparison, each graph shows the survival of challenged guinea pigs vaccinated with PBS (open circles) or live BCG Pasteur (closed circles). Animals were killed at the humane endpoint or at the end of the experiment (day 70).

SID 5 (2/05) of 31

Table 3. Number and percentage of guinea pigs surviving to the end of the experiment and the median survival time per group

Homogenates of lung and spleen from each guinea pig at the time of death were plated for the enumeration of M. bovis. These data are summarized in Figure 7 and Table 4. As for percentage survival, the level of significant protection was set at P=0.008 for pairwise comparisons with the PBS group. Vaccination with live BCG reduced the bacterial load in the lung and spleen compared with the PBS group by an average of 1.21 and 1.55log10 CFU, respectively (Table 4). At the level set, this was only considered significant for the lung, although the P-value for the spleen was 0.022. Killed BCG in water did not reduce the bacterial load in the lungs but when formulated with adjuvant, lower lung CFU counts were observed. However, these were not statistically significant reductions compared with the PBS control (Table 4). In contrast to the lung data, vaccination with killed BCG in water did reduce the bacterial load in the spleen, and the reduction was increased further by the inclusion of NAX M77 and NAX M687 adjuvants. However, this reduction was only statistically significant in the case of killed BCG with NAX M77 ( Table 4). The CFU data for individual animals is shown in Figure 7. Some of the samples were contaminated, thereby reducing the number of individual data points. Nonetheless, within each group those animals surviving to the end of the experiment (filled symbols) generally had the lower organ CFU, especially in the case of the lung. The spleen CFU counts for the NAX M687 group were almost identical to those of NAX M77, with the exception of one animal ( Figure 7). This was responsible for the failure to achieve statistical significance in this group. Despite the range of organ CFU seen in the NAX M687 group, all these animals survived to the end of the experiment ( Figure 6; Table 3), suggesting the enhanced survival effect of the vaccine may be reflected in the pathology.

SID 5 (2/05) of 31

Figure 7. Colony-forming units (CFU) of M. bovis harvested from the lung (A) and spleen (B) of individual guinea pigs vaccinated with (PBS), live (BCG) Pasteur, or formalin-killed BCG Pasteur formulated with sterile water (none) or different Novasome™ non-phospholipid liposome adjuvants (NAX), as indicated. Bacteriology was performed on the organs of animals killed 10 weeks after aerogenic challenge with M. bovis 2122/97 (filled circles) or earlier if the animal had lost 20% of its maximal body weight (the humane end-point) (open circles). The bar shows the mean of the individual values. Data for some animals were lost due to contamination of the samples.

Table 4. Reduction in CFU compared with the PBS group in the lungs and spleens of vaccinated guinea pigs

The extent of gross pulmonary tuberculosis in the right lung of each animal was assessed by weight (pre-fixation), by counting the number of lesions visible on the surface of the lung and assigning a score based on lesion size and severity (post-fixation). By these criteria, no vaccine influenced gross pulmonary tuberculosis in a statistically significant way, although live BCG reduced the number of gross pulmonary lesions by 32% (data not shown), which was consistent with the influence of vaccination reported previously using this model.

SID 5 (2/05) of 31

The experiments were repeated using formalin-killed M. bovis and BCG Tokyo. The results are given in Table 5 below with those for BCG Pasteur for comparison.

Table 5. Protection of guinea pigs by killed vaccines in combination with adjuvants, compared with live BCG vaccines and PBS.

PROTECTION AGAINST DEATH

LUNG CFU SPLEEN CFU

VACCINE ADJUVANT NUMBERSURVIVED

LOGPROTECTION

LOG PROTECTION

PBS none 3/18

Live BCG Pasteur none 15/18* 1.34# 1.92#

Live BCG Tokyo none 5/6* 1.42# 1.43#

Dead M. bovis water 1/8 0.58# 0.63

Dead M. bovis 57 1/6 0.64 0.11

Dead M. bovis M57 1/6 0.83# 0.02

Dead M. bovis M77 1/6 0.82# 0.27

Dead M. bovis M687 4/6* 1.37# 1.03

Dead BCG Pasteur water 2/6 -0.16 0.57

Dead BCG Pasteur 57 2/6 0.27 -0.33

Dead BCG Pasteur M57 3/6 0.29 0.56

Dead BCG Pasteur M77 3/6 0.60 1.45#Dead BCG Pasteur M687 5/6* 0.66 1.11#

Dead BCG Tokyo water 1/6 0.73 0.79

Dead BCG Tokyo 57 5/6* 0.43 0.65

Dead BCG Tokyo M57 4/6* 0.76 1.49#Dead BCG Tokyo M77 0/6 0.02 0.35

Dead BCG Tokyo M687 1/6 0.01 1.34#

*Significance against PBS (p = <0.05), Fishers exact test#Significance against PBS (p = <0.05), Un-paired t-test

Novasome formulation M687 boosted the protective efficacy of both formalin-killed M. bovis and BCG Pasteur and was the only adjuvant to improve the protective efficacy of each killed vaccine tested. The protective efficacy of BCG Tokyo was improved by formulation with Novasome M57 and Novasome 57, which also improved the protective efficacy of killed M. bovis. Interestingly, BCG Tokyo is more closely related to M. bovis (see below) than BCG Pasteur and shares more common antigens – especially MPB70, MPB83 and MPB64, which are all immunogenic during M. bovis infection.

Discussion and Future Work

Although significance was only achieved in terms of survival with NAX M687 (Table 3) and spleen CFU with NAX M77 ( Table 4), encouragingly, these effects were achieved with a single dose of vaccine given via a realistic route, i.e. subcutaneously. As reported for M. tuberculosis (Orme IM. Infect. Immun. 1988 56:3310–3312), killed BCG without adjuvant conferred no significant protection against challenge with M. bovis. The protection we observed with live BCG Pasteur was equivalent to that seen by us previously in this model using a 10-week challenge-necropsy interval (Chambers et al., 2002).

In conclusion, we provide preliminary evidence that it is possible to achieve protection against TB using a killed whole mycobacterial cell vaccine administered once in an appropriate adjuvant. These vaccine formulations are cheap and safe to produce (being already based on BCG) and had no reactogenicity in the naive guinea pig, although the consequences of administering them in situations of pre-existing tuberculosis needs to be investigated with regard to the potentially detrimental ‘Koch phenomenon’. Future studies should address titration of the vaccine dose, identification of the optimum Novasome™ composition and the effect of boosting. Given that the profile of expressed mycobacterial proteins alters during their culture in liquid media, there is scope to optimize the conditions in which the BCG are grown prior to harvest and formalin inactivation and in this way

SID 5 (2/05) of 31

maximize the number and concentration of protective antigens in the final preparation. Finally we recommend that the lead candidate from this study should be taken forward for assessment and optimisation in cattle.

Are there differences in the protective efficacy against M. bovis challenge imparted by different strains of BCG?

BCG vaccines are mutated forms of the causative agent of bovine TB, Mycobacterium bovis, which became attenuated between 1908 and 1921 while being cultivated in Calmette’s laboratory at the Pasteur Institute. However, freezing the bacteria for storage, or lyophilization, only became possible decades later, exposing BCG daughter strains to a further half-century of in vitro evolution. In various laboratories throughout the world, these vaccines had each been propagated through some 1000 additional passages under the same laboratory conditions responsible for their original attenuation, recently demonstrated to coincide with the genomic loss of RD1. Further adaptation to these laboratory selective pressures are speculated to have rendered BCG strains ‘over-attenuated’ and unable to adequately mimic M. tuberculosis when exposed to a human host. Below s a phlogeny of BCG strains elucidated by Marcel Behr on the basis of historical data and recent comparative genomic analysis using an M. tuberculosis whole genome microarray (Behr MA. Lancet Infect Dis. 2002 2:86-92.)

Figure 8. Historical phylogeny of BCG with the results of recent genomic comparisons overlaid. This figure shows that all currently available BCG strains differ from that first used in 1921. Taken from Behr MA. BCG--different strains, different vaccines? Lancet Infect Dis. 2002 2:86-92.

From the phylogeny shown in Figure 8, it can be seen that BCG Tokyo is closer to the progenitor strain of M. bovis than either BCG Pasteur or the only BCG licensed for human use in GB, BCG Danish. However, from the results shown in Table 5 it can be seen that BCG Tokyo and BCG Pasteur gave similar levels of protection against aerosol challenge with M. bovis in guinea pigs.

Subunit Vaccine Candidates

Among several strategies to replace BCG with novel TB vaccine candidates, e.g., vaccination with a subunit protein, naked DNA, and improved whole bacterial vaccines, vaccination with a subunit protein is perhaps the most attractive from a regulatory perspective both in terms of safety and standardisation.

The results obtained with formalin-killed BCG and M. bovis in Novasomes described above suggested that the use of subunit vaccines might be a feasible approach as long as the protective antigens can be identified and delivered in a suitable formulation to stimulate protective immunity.

The choice of antigens to be included in an experimental subunit vaccine largely relies on the fact that live, as opposed to killed, BCG vaccine displays a high level of protection in animal models. The hypothesis that the response to antigens secreted from live mycobacteria is essential for host protection has been put forward, and

SID 5 (2/05) of 31

several lines of evidence support an important role for these substances in eliciting a protective response against a subsequent TB challenge. An important feature of immune responses against extracellular mycobacterial antigens is their early onset in the course of infection and their ability to rapidly stabilise mycobacterial loads in the parenchymal organs of infected animals at levels that are significantly lower than those in nonvaccinated controls.

However, none of the TB vaccines studied so far have been able to effectively prevent the development of pulmonary TB in animals infected with Mycobacterium tuberculosis. Even when the early mycobacterial multiplication was substantially inhibited due to vaccination, later in the infection course, the mycobacteria recommenced their progressive growth in the lungs, leading to increasing tissue damage and death. Given that the majority of TB cases in humans are thought to be cases of reactivated disease, it has been suggested that a novel subunit TB vaccine should include antigenic components that accumulate in latent (or dormant) mycobacteria during their long survival in the host. Theoretically, an immune response against such dormant-state antigens should protect against late-phase mycobacteria; however, there is no convincing evidence that these substances leave the dormant bacterial cells and become available for recognition by the immune system while metabolically passive microorganisms remain within macrophage phagosomes.

That said, it is attractive to speculate that an optimised subunit vaccine should be able to elicit an immune response against both early-secreted mycobacterial antigens and antigens that are expressed later in the infection process. We therefore sought to determine the protective efficacy of a number of M. bovis antigens that are secreted at different times of infection and that are either immunogenic in cattle during M. bovis infection (including culture filtrate proteins (CFP), MPB83, MPB70, ESAT-6, MPB63, Rv3019c) or that had shown promise as vaccines against M. tuberculosis (Ag85, Rpf proteins, HSP65, HSP70, Apa).

DNA Vaccines

Despite the promise shown in mouse models, only one of a number of DNA vaccines that we tested was found to have any significant protective effect against M. bovis challenge in guinea pigs.

Recent reports of successful DNA vaccination against Mycobacterium tuberculosis in small-animal models had suggested that DNA vaccines act by reducing lung pathology without sensitizing animals to tuberculin testing. We therefore evaluated the ability of vaccines consisting of DNA encoding the mycobacterial antigens MPB83 and 85A to reduce lung pathology and prevent hematogenous spread in guinea pigs challenged with a low dose of aerosolized M. bovis. Vaccination with MPB83 DNA reduced the severity of pulmonary lesions, as assessed by histopathology, and resembled M. bovis BCG vaccination in this respect. However, unlike BCG vaccination, MPB83 DNA vaccination did not protect challenged guinea pigs from hematogenous spread of organisms to the spleen. In contrast, vaccination with antigen 85A DNA, a promising DNA vaccine for human tuberculosis, had no measurable protective effect against infection with M. bovis. This was also the case for a DNA vaccine encoding the HSP65 antigen, which has also shown promise in the mouse model of M. tuberculosis.

The results of the MPB83 study were published in: Chambers MA, Williams A, Hatch G, Gavier-Widen D, Hall G, Huygen K, Lowrie D, Marsh PD, Hewinson RG. Vaccination of guinea pigs with DNA encoding the mycobacterial antigen MPB83 influences pulmonary pathology but not hematogenous spread following aerogenic infection with Mycobacterium bovis. Infect Immun. 2002 70:2159-65. The paper is attached to this report and detailed results and methodologies can be found therein.

Protein Subunit vaccines

Optimisation of suitable adjuvants for protein vaccination in guinea pigs.

A number of protein subunit vaccines against tuberculosis have shown promise but require administration with adjuvants to stimulate relevant immune responses for protection. Although guinea pigs are the model of choice for evaluating protective immunity to aerogenic challenge with virulent mycobacteria, few studies have been undertaken to identify suitable adjuvants for vaccine screening in this species. The work described above with formalin-killed BCG and M. bovis had identified a number of promising adjuvant formulations based on Novasome technology. However, it proved difficult to secure this technology for the testing of protein subunit vaccines. We therefore compared the efficacy of several adjuvants to induce T cell responses to culture filtrate protein in guinea pigs.

Preparation of antigen and adjuvants: Purified CFP from M. bovis AN5 strain was produced from a 10 week culture of M. bovis AN5, grown as a surface pellicle on BAI medium. Freunds incomplete adjuvant (FIA) (Sigma, Poole, UK) was mixed 1:1 with CFP and emulsified by sonication on ice for 3 x 5 s using a Sonics vibracell with CV33 probe at 20% amplitude (Sonics & Materials Inc., CT, USA). Monophosphoryl lipid A-trehalose

SID 5 (2/05) of 31

dicorynomycolate (MPL–TDM, Sigma) was mixed with CFP according to manufacturers instructions. MPL–TDM/dimethyl dioctadecyl ammoniumbromide (DDA) was formulated as MPL–TDM with the addition of 1 mg/ml DDA (Sigma) and subsequent sonication as before. Tomatine (Sigma) was formulated as a molecular aggregate according to the method of Rajananthanan et al. (Vaccine 1999 17:715–730). CpG oligodeoxynucleotide (ODN) sequence 2007 (TCGTCGTTGTCGTTTTGTCGTT) was synthesised with a nuclease-resistant phosphorothioate backbone (QIAGEN, Hilden, Germany). 50 mg of CpG was formulated in 30% Emulsigen (MVP Laboratories, Ralston, USA) and mixed 1:1 with 150 mg CFP. CFP was mixed with each adjuvant to a final concentration of 600 mg/ml.

Results: Of the several adjuvants tested, the most promising was CpG ODN formulated in an aqueous emulsion. This adjuvant induced type 1 T cell responses of a similar magnitude to that of FIA, as measured by delayed-type hypersensitivity reactions (DTH) (data not shown), antigen-specific T cell proliferation (shown below in Figure 9) and antigen-specific IgG1 and IgG2 responses (data not shown). These data demonstrate the potential for CpG motif based adjuvants for use in TB vaccine screening in guinea pigs where a type 1 T cell response is required.

Figure 9. Antigen-specific spleen cell proliferative responses from individual guinea pigs, 3.5 weeks following final immunisation with the indicated antigen–adjuvant formulations. Spleen cells were cultured in the presence of antigen (CFP, 10 mg/ml) and proliferation assessed by 3H thymidine incorporation. Results are expressed as SI. Bars represent mean group values ± S.E. *P<0.03 compared with saline immunised guinea pigs.

The results of this study were published in: Hogarth PJ, Jahans KJ, Hecker R, Hewinson RG, Chambers MA. Evaluation of adjuvants for protein vaccines against tuberculosis in guinea pigs. Vaccine. 2003 21:977-82. The paper is attached to this report and detailed results and methodologies can be found therein.

Testing protein subunit vaccinesHaving identified a suitable adjuvant for testing secreted proteins as vaccine candidates in guinea pigs, we next tested a number of vaccine candidates for their ability to protect against M. bovis challenge. Culture filtrate proteins (CFP) has been shown to induce protection against M. bovis in cattle (Wedlock et al Infect. Immun. 2000, 68:5809–5815) and so we chose to test CFP and the MPB70, which is the predominant protein found in the culture filtrate of M. bovis. In addition, because our ultimate goal was to produce a vaccine comprising a protein cocktail which is able to elicit a response against both early-secreted mycobacterial antigens and antigens that are expressed later in the infection process, we also tested proteins that are expressed in the later stages of infection. Mycobacterium tuberculosis expresses five proteins that are homologous to Rpf (resuscitation promoting factor), which is secreted by growing cells of Micrococcus luteus. Rpf is required to resuscitate the growth of dormant Micrococcus luteus organisms, and its homologues may be involved in mycobacterial reactivation. Mycobacterial Rpf-like products are secreted proteins, which makes them candidates for recognition by the host immune system during reactivated tuberculosis. Recently it was shown that vaccination of mice with Rpf-like proteins results in a significant level of protection against a subsequent high-dose challenge with virulent M. tuberculosis H37Rv, both in terms of survival times and mycobacterial multiplication in lungs and spleens (Yeremeev VV et al., Infect Immun. 2003 71:4789-94).

SID 5 (2/05) of 31

Immunization and challenge. Guinea pigs (5 per group) were immunised with either CFP, MPB70, or with one of two Rpf proteins that had shown promise in mice i.e. Rv2389 or Rv2450. Proteins were produced as described previously (Yeremeev VV et al., Infect Immun. 2003 71:4789-94; Carr et al., J Biol Chem. 2003 278:43736-43). Each animal was vaccinated with 150g protein in CpG immunEasy adjuvant (1:1 antigen:adjuvant) (which induced similar immune responses with CFP to the CpG-Emulsigen adjuvant described above) in a total volume of 0.5 ml subcutaneously at the nape of the neck at weeks 0, 3 and 6. Four weeks later the guinea pigs were challenged with 10 cfu M. bovis using the standard aerosol challenge model and animals were killed 5 weeks later and the lungs and spleens removed and used to determine the bacterial burden.

Results: None of the protein vaccines tested gave protection against M. bovis challenge as measured by CFU in lungs (Figure 10) or spleens (results not shown).

Figure 10: Colony-forming units (CFU) of M. bovis harvested from the lung (A) of individual guinea pigs vaccinated with (PBS), live (BCG) Pasteur, Rv2389, Rv2450, MPB70, culture filtrate proteins or adjuvant (CpG immunEasy). Bacteriology was performed on the organs of animals killed 5 weeks after aerogenic challenge with M. bovis 2122/97. The bar shows the mean of the individual values.

Heterologous Prime Boost Regimes

Since neither DNA vaccination or protein subunit vaccination was efficacious in guinea pigs we next evaluated a heterologous prime-boost approach to vaccination which had shown some promise in the development of vaccines against human tuberculosis. In human TB vaccine development, as in bovine TB vaccine development, no subunit vaccine has yet been shown to confer protection superior to BCG in the mouse or even equivalent to BCG in the more sensitive guinea pig model. One possible reason for this may be that although these subunit vaccines are effective at inducing cellular immune response, the level of responses they induce tends to be low. Heterologous prime-boost immunisation strategies have been developed in which two different vaccines are given, each encoding the same antigen, several weeks apart. Such strategies have been shown to induce higher levels of interferon gamma secreting CD8+ cytotoxic T lymphocytes, and Th-1 type gamma interferon secreting CD4+ T cells in animal models of HIV, malaria and TB than homologous regimes. We therefore investigated the utility of this approach for the development of vaccines against M. bovis using the experimental approaches described below.

A DNA-MVA heterologous prime-boost regime

In this experiment we investigated the protective efficacy of a prime-boost immunisation regime using a DNA vaccine and recombinant MVA expressing two mycobacterial antigens, ESAT-6 and MPT63. ESAT-6 is a key antigenic target in murine infection and is a human CTL target. It is present in M. tuberculosis and M. bovis, but not BCG. MPT63 is a small, secreted antigen that is immunogenic in guinea pig models [Manca, 1997 Infect Immun 65:16-23.] and present in some strains of BCG. Use of these constructs in a prime-boost approach has been reported to be protective in a murine challenge model of M. tuberculosis [McShane, 2001Infect Immun 69: 681-6]. The aim of the present study was to investigate the protective efficacy of a prime-boost regime using these constructs in a guinea pig model of M. bovis infection.

Materials and MethodsConstruction and propagation of the DNA plasmid pSG2.TEMPk and recombinant MVA expressing the esat-6 and mpt63 genes from Mycobacterium tuberculosis (MVA-E6/63) have been described previously [McShane,

SID 5 (2/05) of 31

2001, Infect Immun 69:681-6]. A recombinant MVA expressing beta-galactosidase (MVA-gal) was used as a control.

The vaccination schedule for each group is shown in Table 6. Guinea pigs immunised with DNA received 100g of plasmid DNA into the biceps femoris muscle of both hind limbs on each occasion. Where appropriate, guinea pigs received 5x106 PFU of MVA intradermally in each ear. For the BCG group, the concentration of inoculum was adjusted to 2x105 CFU/ml immediately prior to injection of 5x104 CFU subcutaneously in the nape. A sham-vaccinated control group received 250l of PBS in the nape.

Table 6: Vaccination schedule, group name and size

Group name (n) Week 1 Week 4 Week 7 Week 10

PBS (6) - - - PBS

DDDD (6) pSG2.TEMPk pSG2.TEMPk pSG2.TEMPk pSG2.TEMPk

DDDcM (5) pSG2.TEMPk pSG2.TEMPk pSG2.TEMPk MVA-gal

DDDrM (6) pSG2.TEMPk pSG2.TEMPk pSG2.TEMPk MVA-E6/63

BCG (6) - - - BCG Pasteur

All guinea pigs were challenged with 10 cfu M. bovis strain AF2122/97 via the aerosol route six weeks after the last vaccination Each animal was weighed weekly and killed by peritoneal overdose of sodium pentobarbitone when it had lost 20% of its maximal body weight (the humane endpoint). The lungs and spleen were removed aseptically immediately after death. The whole spleen and half of the lungs were used for bacteriology. The remaining lung was placed in 10% formal buffered saline for histopathology.

ResultsUsing this approach, a significant reduction in bacterial load was seen in the lungs of DNA-MVA immunised animals, and in the BCG immunised positive control group. In both of these groups, the level of lymphocytic infiltration was significantly higher than in the control groups.

Table 7: Protection in lungs and spleens of vaccinated guinea pigs challenged with a low-dose M. bovis aerosol

Immunisation

group n

Mean log10 CFU

(lung) ± SE

Log10 protection

(lungs)

Mean log10 CFU

(spleen) ± SE

Log10 protection

(spleen)

PBS 6 7.05 ± 0.09 0.00 5.95 ± 0.16 0.00

DDDD 6 7.02 ± 0.20 0.03 5.84 ± 0.22 0.11

DDDcM 5 6.86 ± 0.34 0.20 5.41 ± 0.42 0.53

DDDrM 6 6.58 ± 0.16 0.48* 5.49 ± 0.29 0.46

BCG 6 6.44 ± 0.19 0.61* 4.69 ± 0.08 1.26*

*P < 0.05, by Student t-test.

In addition, the median duration of survival in the DNA-MVA immunised guinea pigs was increased by 29% from 32.5 in control animals to 42.0 days, although this was not statistically significant (Mann-Whitney and Logrank Tests). Compared with the PBS control, vaccination with BCG Pasteur increased the median survival time by 57.5 days (P < 0.005, Mann-Whitney Test).

These results showed that although the subunit vaccines used in this experiment did not confer protection to the level of BCG, the use of heterologous immunisation regimes did improve upon the protection conferred by single subunit vaccines and highlighted a clear strategy for future TB vaccine development.

SID 5 (2/05) of 31

This experiment also resulted in an important collaboration with Prof. Adrian Hill’s Group at Oxford University who are performing the first human clinical trials of a new TB vaccine (MVA expressing Ag85). Our collaboration with Prof. Hill’s Group has given us access to this vaccine, which will be tested in cattle as part of SE3224.

A prime-boost vaccination strategy aimed at improving BCG vaccination

In our final approach we combined the use of a live vaccine, BCG, with a cocktail of DNA vaccines. By using this strategy we sought to improve the efficacy of BCG by boosting BCG-induced immunity with subunit vaccines. Two advances in our vaccine development programme at VLA had allowed us to pursue this strategy. First, we had developed specific diagnostic tests that differentiated between cattle that had been vaccinated with BCG and animals that were infected with M. bovis using antigens that are present in M. bovis but absent in BCG (see SE3028 & SE3212). Second, we had demonstrated that a cocktail of DNA vaccines could enhance the protection afforded by BCG in cattle (SE3212; Skinner et al., Infect Immun. 2003 71:4901-7). The purpose of this experiment was to determine whether the guinea pig was a suitable model with which to predict the efficacy of promising heterologous prime-boost vaccination regimes aimed at improving BCG for cattle. Therefore the vaccines that we used to boost BCG immunity were the same cocktail of DNA vaccines that had proved efficacious in cattle i.e. DNA vaccines encoding HSP65, HSP70 and Apa (Skinner et al., Infect Immun. 2003 71:4901-7).

SID 5 (2/05) of 31

Materials and MethodsGroups of guinea pigs were immunized according to the protocol given in the Table below:

GROUP

(n)

Wk0 Wk3 Wk6 Wk9 Wk16

A (8) Vector DNA2 Vector DNA Vector DNA 10 CFU aerosol

bovis

B (8) Expt DNA1 Expt DNA Expt DNA 10 CFU aerosol

bovis

C (16) Expt DNA Expt DNA Expt DNA BCG

Pasteur

10 CFU aerosol

bovis

D (8) Vector DNA2 Vector DNA Vector DNA BCG

Pasteur

10 CFU aerosol

bovis

E (16) BCG

Pasteur

10 CFU aerosol

bovis

Total = 56 animals1Expt DNA: An equal mix of pCMV-apa, pCMV-hsp70, and pCMV-hsp65. 400g total of DNA was injected i.m. on each occasion (half at each site).2Vector DNA: 400g total of pCMV-link DNA was injected i.m. on each occasion (half at each site). Groups C to E received 250l s.c. of BCG Pasteur (dose = 5x104 CFU at 2x105 CFU/ml).

All guinea pigs were challenged with 10 cfu M. bovis strain AF2122/97 via the aerosol route six weeks after the last vaccination Each animal was weighed weekly and killed by peritoneal overdose of sodium pentobarbitone 5 weeks after challenge. The lungs and spleen were removed aseptically immediately after death. The whole spleen and half of the lungs were used for bacteriology. The remaining lung was placed in 10% formal buffered saline for histopathology.

Results

The results showed that the cocktail of DNA vaccines used alone had no protective effect against M. bovis challenge (Figure 11). Furthermore the guinea pig model was unable to predict the results obtained in cattle (and mice: Hogarth et al., Vaccine. 2006 24:95-101) i.e. that priming with DNA vaccine could improve the protective efficacy of BCG in an heterologous prime-boost vaccination strategy. For this reason we decided that the mouse model and not the guinea pig model should be used to screen vaccines for use in prime-boost strategies to improve the protective efficacy of BCG. This is now the strategy taken forward in SE3224.

SID 5 (2/05) of 31

Figure 11: Colony-forming units (CFU) of M. bovis harvested from the lung and spleenof individual guinea pigs vaccinated with (PBS), live (BCG) Pasteur, 3xDNA encoding HSP70, HSP65 and Apa, 3xDNA encoding HSP70, HSP65 and Apa + BCG, 3xDNA vector +BCG, as indicated. Bacteriology was performed on the organs of animals killed 5 weeks after aerogenic challenge with M. bovis 2122/97. The bar shows the mean of the individual values.

Objective 04 and 05 Use of luminescent assay for evaluating efficacy of vaccine candidates.

The Animal Models Task Force of IMMYC recommended that infection of the lung and haematogenous seeding of the spleen with challenge organisms should be used as an indicator of vaccine efficacy 5 weeks post challenge. This read out is also used to judge vaccine efficacy as part of the human TB vaccine-testing contract

SID 5 (2/05) of 31

for the NIH. Therefore we used this readout as a correlate of protective immunity induced by a vaccine in the initial phase of our vaccine-testing programme. However, the procedures involved namely, tissue homogenisation, serial dilutions, plating and counting, together with the slow doubling time of M. bovis on solid medium (up to 6 weeks to obtain sizeable colonies), make this a time-consuming and expensive bottle-neck in the screening of vaccine candidates. Recent work in Professor Young’s group at St Mary’s Hospital has shown that culture of M. tuberculosis from an infected animal can be replaced by a reporter system based on the firefly luciferase gene. Prof. Young’s group have made a recombinant M. tuberculosis that produces light whilst viable. This recombinant has the same growth characteristics and virulence as the parental strain. Tissue homogenates from an infected animal can simply be assayed in a luminometer for a measurement of the number of viable organisms present in the sample. Using this approach, measurement of luminescence was shown to provide a rapid and simple alternative to the counting of CFU as a means of monitoring mycobacterial viability (Snewin et al., Infect Immun. 1999 67:4586-93).

The recombinant plasmid consisted of the luxA and luxB genes of V. harveyi cloned under the control of the BCG hsp60 promoter in a shuttle vector capable of replication in E. coli and mycobacteria and employing hygromycin as a selectable marker (Snewin et al., Infect Immun. 1999 67:4586-93). Reporter plasmid DNA was prepared from recombinant E. coli DH5 by using Qiagen columns according to manufacturer's recommendations and was used to electroporate M. bovis AF2122/97. Recombinant colonies were tested for luminescence and were maintained at 80°C.

Groups of 5 guinea pigs were challenged with an aerosol of 10 cfu of either: recombinant M. bovis AF2122/97 expressing the luciferase reporter strain (M. bovis pSMT1) or wild-type M. bovis AF2122/97. Guinea pigs were killed at 5 weeks post-challenge. Relative virulence was assessed by post-mortem analysis and by the bacterial load observed in lung and spleen of infected animals. The recombinant M. bovis strain expressing bacterial luciferase caused pathological sign that were typical of M. bovis infection in the lungs and spleens of guinea pigs 5 weeks after aerosol challenge. However the typical lesion pattern associated with M. bovis infection was not observed in the animals infected with wild type M. bovis (positive control). In addition, bacteria could not be isolated from the lung or spleen of these animals although small lesions could be identified on the lung at post mortem. Mouse studies subsequently confirmed the low virulence of this batch of wild type challenge strain suggesting that, for this test batch, the organism had become attenuated on passage.

By the time we had obtained these results the project only had two years to run. Moreover, survival had been shown to be the most useful parameter for determining vaccines that were more effective than BCG in this project and by our sub-contactors at Health Protection Agency, Porton Down, as part of their activities in evaluating human TB vaccines within the the EU Vth Framework TB Vaccine Cluster (Williams, A., Tuberculosis 2005 85:29-38.). We therefore decided not to repeat this experiment but to use the limited resources available to test more vaccine candidates using survival and colony counts as read outs, parameters which are well established and recognised in human TB vaccine testing centres.

Objective 02 Virulence of M. bovis challenge strain compared with two other isolates in the guinea pig.

Introduction

For a number of years there have been anecdotal reports from the SVS and John Gallagher of differences in the pathology caused by M. bovis infection in badgers from Cornwall and Gloucester. In SE3020, analysis designed to look at associations between spoligotypes and the epidemiological features of an outbreak was performed. Of particular interest was the potential relationship between spoligotype and epidemiological features of the disease. These features included:

1) IRs as a proportion of IRs & reactors 2) The proportion of unconfirmed incidents in the same 10 x 10 km square3) Proportion of incidents detected first in the slaughterhouse4) Duration of restrictions 5) Percentage of reactors with visible lesions6) Percentage of samples from which M. bovis isolated.

Preliminary analysis revealed some intriguing differences across different molecular types of M. bovis (Goodchild et al., Proceeding of the Society for Veterinary Epidemiology and Preventive Medicine, Warwick 31 March - 2 April 2003.). This analysis showed differences in numbers of inconclusive reactors across spoligotypes, and revealed that particular spoligotypes are more frequently detected on repeat testing. This suggests that clonal groups of M. bovis may have distinct phenotypes that are relevant to the control strategy for bovine TB in GB.

The standard M. bovis challenge strain used in this project was that from which the M. bovis genome sequence was obtained. This strain (AF2212/97) was isolated from a cow in the south-west of England and a matched isolate from a badger caught in the cattle shed is available (spoligotype 9; the most common British spoligotype). However, work performed under SE0130 using the low-dose guinea pig aerosol challenge model revealed

SID 5 (2/05) of 31

differences in the virulence of at least three different spoligotypes of M. bovis. The isolates differed notably in their ability to cause secondary infection in both the lung and other organs. In particular, differences in ability to colonise the spleen were noted. In these experiments, an isolate with the VNTR pattern 7555*33.1 showed low virulence for guinea pigs compared with the other isolates. Clearly, the choice of M. bovis challenge strain used for vaccine screening is important.

We therefore compared the virulence of four recent isolates of M. bovis. Three of the isolates were of spoligotype 9 but one isolate differed in VNTR pattern and one isolate of spoligotype 17 had a matched VNTR pattern with one of the spoligotype 9 strains. The strains used were AF449/97 (spoligotype 9, VNTR type 8555*33.1), AF1692/96 (spoligotype 17, VNTR type 7555*33.1), AF2122/97 (spoligotype 9, VNTR type 8555*33.1) and AF1280/95 (spoligotype 7555*33.1). Comparison of the virulence of AF1692/96 with AF1280/95 should establish therefore whether VNTR pattern 7555*33.1 is associated with low virulence. Similarly, comparison of AF2122/97 with AF1280/95 should indicate whether isolates of the same spoligotype have the same virulence despite having different VNTR patterns.

Guinea pigs (8 per group) were challenged with 10 cfu of each of the four M. bovis isolates. Animals were weighed every other day and killed four weeks after challenge. Half of the lung, spleen and kidney of each animal was used for bacteriology and the other half was formalin-fixed and kept for histopathology. Organisms isolated from spleens and kidneys of each animal were spoligotyped and VNTR typed to ensure that no cross-contamination had occurred between animals.

Results. In these studies strain AF449/97 was significantly more virulent than the other three strains tested both in terms of the weight loss it induced (Figure 12) and in the numbers of M. bovis detected in the lungs (Figure 12) and spleens. However there was no correlation between spoligotype, VNTR type and virulence and therefore no evidence that VNTR type 7555*33.1 is associated with low virulence.

Figure 12: Comparison of the virulence of M. bovis isolates in guinea pigs.

Appendix: Materials and methodsBacteria and media. Lyophilized M. bovis BCG Pasteur strain (obtained from the Statens Serum Institut, Copenhagen, Denmark) was cultured in 10 ml of M-ADC-TW broth for 7 days and stored at -80°C in seed lots. The strain of M. bovis used in this study (AF2122/97) was isolated from a tuberculin test reactor cow in 1997 and was cultured at Veterinary Laboratories Agency Weybridge. For enumeration, BCG Pasteur was plated on Middlebrook 7H10 agar containing 0.2% (vol/vol) glycerol and 10% (vol/vol) Middlebrook oleic acid-albumin-dextrose-catalase (OADC) enrichment. M. bovis strain AF2122/97 was plated on Middlebrook 7H10 agar containing 4.16 mg of sodium pyruvate per ml and 10% (vol/vol) Middlebrook OADC enrichment. When necessary, serial dilutions of bacterial suspensions were prepared in water containing 0.05% (vol/vol) Tween 80 to maintain dispersion.

Preparation of bacteria for aerosol or intramuscular infection of guinea pigsM. bovis strain AF2122/97 was grown in 7H11 broth and harvested at mid-log phase of growth. For the highest dose aerosol infection in this study, 0.4 g wet mass of M. bovis was resuspended in 15 ml water and sonicated for 1 min using a sonicator and ultrasonic processor W-385 (Heat Systems-Ultrasonics, Inc., New York, USA) set at

SID 5 (2/05) of 31

Exp 2J - Average guinea pig weight post challenge

Time (days)

0 5 10 15 20 25 30

Wei

ght (

g)

300

320

340

360

380

400

420

440

460

480

500

A - 449/97B - 1692/96C - 2122/97D - 1280/95

Viable M. bovis isolated from lung

M. bovis strain

449/97 1692/96 2122/97 1280/95

Via

ble

M.b

ovis

(Log

10 c

fu/m

l)

3

4

5

6

7

--------

-------- --------

--------

20% power. For i.m. infection, the wet mass was diluted to 0.2 mg/ml in saline (representing approximately 200 CFU/ml). The actual concentration of viable bacteria in the suspensions was determined by plating on Middlebrook 7H10 agar (Difco Laboratories, Detroit, MI) containing 4.16 mg/ml sodium pyruvate and 10% (v/v) Middlebrook ADC enrichment. Where necessary, serial dilutions of the suspension were made in water containing 0.05% (v/v) Tween 80 to maintain dispersion.

Aerosol infection of guinea pigs

Female Dunkin–Hartley guinea pigs weighing between 350 and 450 g and free of intercurrent infection were obtained from Charles River UK Ltd., Margate, UK. Guinea pigs were exposed for 5 min to bacterial aerosols containing particles mostly below 5 mm diameter (diameter range 0.5–7 mm, mean 2 mm). The aerosol was generated from the suspension of M. bovis AF2122/97 with a three-jet Collison nebulizer in conjunction with a modified version of the mobile Henderson apparatus described by Druett (Druett, H.A., 1969. A mobile form of the Henderson apparatus. J. Hyg. 67, pp. 437–448.). The apparatus, housed within an appropriate Containment Level 3 biohazard facility, allows controlled delivery of aerosols directly to the snouts of the animals without contamination of fur or eyes. Previous observations using mycobacterial species in the Henderson apparatus have shown that the dose retained in the lungs by animals immediately following a 5 min exposure to aerosol is 5 log10 less than the concentration of the suspension used (A. Williams and M.S. Lever, unpublished data). Thus, three doses of M. bovis were administered to groups of eight guinea pigs from suspensions containing approximately 106, 107 or 108 CFU/ml in order to obtain inhaled retained doses in the lungs of approximately 10, 100 and 1000 organisms, respectively. A control group of eight animals were exposed to a saline aerosol for 5 min.

The Bovine Standard purified protein derivative

The Bovine Standard purified protein derivative (PPD) used in these studies was the Biological First International Standard antigen — tuberculin, purified protein derivative, bovine — currently held and distributed by NIBSC, Potters Bar, UK. Each ampoule contains 58,500 International Units (approximately 1.8 mg) of PPD derived from cultures of M. bovis strain AN5.

Intramuscular infection of guinea pigs

Dunkin–Hartley guinea pigs weighing between 350 and 450 g were injected with 0.1 mg wet mass of live M. bovis AN5 (approximately 100 CFU) in a 500 ml volume of saline in the flexor muscles of the crus region of the right hind leg.

Cutaneous delayed-type hypersensitivity assay

All guinea pigs were assayed for cutaneous DTH 5 weeks after both i.m. and aerosol infection with M. bovis in accordance with the recommended procedures documented in the European Pharmocopoeia for potency evaluation of bovine tuberculin PPD. An area of approximately 10–13 cm2 on the flank of each animal was shaved to remove the fur. The Bovine Standard PPD was diluted in isotonic PBS plus Tween to give a working concentration of 250 International Units (IU)/ml. 0.2 ml (50 IU) of this PPD solution was injected into the shaved dermis of each animal using a 1 ml syringe fitted with a 25G needle. The extent of the hypersensitivity reaction was measured 24 and 48 h later using digital callipers. Two measurements were taken at right angles to one another in order to calculate the induration size in square millimetre.

Post-mortem examination of guinea pigs

Animals infected via the i.m. or aerosol route were killed humanely 5 weeks post-infection (after the final DTH reading) by intraperitoneal overdose of sodium pentobarbitone. Examination was carried out immediately after death when external assessment of body condition was followed by gross internal examination of the musculoskeletal system, neck region, thoracic and abdominal cavities. Retained bacterial dose in the lungs was estimated by counting the number of primary tubercle lesions visible at post-mortem. The whole spleen and lungs were removed aseptically and placed separately into 5 ml sterile distilled water for bacteriology.

Bacterial enumeration

Spleens and lungs were homogenised in 5 ml sterile distilled water using a rotating blade macerator system (MSE homogenizer). Viable counts were performed on the macerate by serial dilution in sterile distilled water and plating of dilutions onto Middlebrook 7H10 agar (Difco Laboratories) containing 4.16 mg/ml sodium pyruvate and 10% (v/v) Middlebrook ADC enrichment. Plates were incubated at 37±2°C and examined after 5 weeks for growth of M. bovis. The number of colonies on each plate containing between 30 and 300 colonies was counted and recorded.

SID 5 (2/05) of 31

Determination of pulmonary disease by using formalin-fixed tissue.

The trachea, bronchi, and heart were dissected away prior to detailed examination of the fixed lungs from each animal. The number of visible lesions on the dorsal surface of all lobes was recorded on a diagram of the lungs along with the positions and sizes of the lesions and information about whether necrosis or consolidation was present. Sections were then cut from the lungs of each animal and used for histopathological evaluation. A section was prepared from the base of the left apical lobe and from the right diaphragmatic lobe of each lung; we made sure that each lobe was sectioned at the same position for every animal. Duplicate sections were stained with hematoxylin and eosin stain and with van Giesson stain in order to aid visualization of fibrous tissue. The two sections were scored in a blinded fashion for the following features: area of the section occupied by granulomatous inflammation (granuloma area); fibrosis; calcification; and necrosis (coagulative or caseous). Necrosis was described as coagulative when the architecture of the tissue was preserved. In contrast, if all tissue architecture was lost, the necrosis was termed caseous. The extent of lymphocytic infiltration in each section was also assessed, along with the distribution of the lymphocytes (perivascular and/or peribronchiolar, peripheral to the granuloma, or inside granulomas). The extent or severity of each feature was assigned a score ranging from 0 to 4.

Statistical analyses.

Appropriate statistical tests were used, and the significance of observed differences tested, using the InStat software package (version 3.00, GraphPad, San Diego, CA). The Mann-Whitney nonparametric test was used when the data were not sampled from Gaussian distributions. One-way analysis of variance was used for some of the data.

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 (2/05) of 31

Publications Arising from this project.*

1. Chambers MA, Williams A, Gavier-Widen D, Whelan A, Hall G, Marsh PD, Bloom BR, Jacobs WR, Hewinson RG. Identification of a Mycobacterium bovis BCG auxotrophic mutant that protects guinea pigs against M. bovis and hematogenous spread of Mycobacterium tuberculosis without sensitization to tuberculin. Infect Immun. 2000 68:7094-9.

2. Chambers MA, Williams A, Gavier-Widen D, Whelan A, Hughes C, Hall G, Lever MS, Marsh PD, Hewinson RG. A guinea pig model of low-dose Mycobacterium bovis aerogenic infection. Vet Microbiol. 2001 80:213-26.

3. Chambers MA, Williams A, Hatch G, Gavier-Widen D, Hall G, Huygen K, Lowrie D, Marsh PD, Hewinson RG. Vaccination of guinea pigs with DNA encoding the mycobacterial antigen MPB83 influences pulmonary pathology but not hematogenous spread following aerogenic infection with Mycobacterium bovis.

4. Infect Immun. 2002 70:2159-65. 5. Chambers MA, Wright DC, Brisker J, Williams A, Hatch G, Gavier-Widen D, Hall G,

Marsh PD, Glyn Hewinson R. A single dose of killed Mycobacterium bovis BCG in a novel class of adjuvant (Novasome) protects guinea pigs from lethal tuberculosis. Vaccine. 2004 22:1063-71.

6. Hogarth PJ, Jahans KJ, Hecker R, Hewinson RG, Chambers MA. Evaluation of adjuvants for protein vaccines against tuberculosis in guinea pigs. Vaccine. 2003 21:977-82.

7. Inwald J, Jahans K, Hewinson RG, Gordon SV. Inactivation of the Mycobacterium bovis homologue of the polymorphic RD1 gene Rv3879c (Mb3909c) does not affect virulence. Tuberculosis (Edinb). 2003;83:387-93.

* The efficacy of the vaccine candidates was assessed in small animal models and cattle as part of a co-ordinated programme of vaccine development at VLA (proposals SE3208, SE3209, SE3028 and SE3212). Therefore publications may appear in more than one report.

SID 5 (2/05) of 31

General enquiries on this form should be made...

Documents

Transcript of General enquiries on this form should be made...