CORDIS

Executive Summary:
NEWTBVAC proposed the discovery and the preclinical development of a new generation of tuberculosis (TB) vaccines. The project was divided into seven work packages (WP) that coordinated the different topics, antigen discovery (WP1), subunit and antigen-delivered vaccines (WP2), live vaccines (WP3), prime boost strategies (WP4), preclinical model platforms (WP5), TB biomarkers (WP6), and coordination including translational product development support (WP7). In WP1, a new class of lipoid antigens, including Ac2SGL, were identified as well new subdominant antigens both of which could complement the current antigens in future vaccine formulations. Novel adjuvants, including CAF09, were identified and developed. In WP2 and WP4, novel viral vector based vaccines, including LCMV and ChAdOx1.85A were developed, which can be deployed as booster vaccines but can also complement other viral vector-based subunit vaccines. New HBHA-based protein adjuvant formulations were developed with a focus on methylated and non-methylated HBHA peptides. In WP3, a broad number of novel viable BCG or M. tuberculosis derived vaccines for BCG replacement have been developed These included an improved rBCG VPM1002, a novel BCGΔZMP1, and the live attenuated MTBVAC. The animal model platforms of WP5 proved to be an instrumental core service which provided the relevant information needed for decisions for next steps of individual candidates. The models also included head-to-head testing and direct comparison of vaccine candidates which contributes to the future gating of novel vaccine candidates. WP6 successfully developed a platform of new biomarkers and biomarker assays based on a broad array of immunological, genomic, transcriptomic and metabolomic read-outs. The design of decision-making matrices developed in WP6 will help in the selection and gating of future vaccine candidates as well as in the stratification of most relevant study populations of future vaccine trials. The activities of WP1 to WP6 as well as the decision making process as to which vaccine candidates to move forward, was coordinated in WP7, where several experienced and respected vaccine developers contributed as members of the PDT and CDT structures.
Overall, NEWTBVAC has been instrumental in bringing the complete TB vaccine pipeline forward at all stages. Novel antigens which are now tested in preclinical and clinical vaccine evaluations had been discovered within NEWTBVAC. Novel viral carriers and adjuvants have been developed for next-generation vaccine candidates. Novel live vaccine candidates, have been translated to preclinical and clinical R&D. Several highly promising biomarkers have been developed and are being validated for use of monitoring in clinical trials. PDT and CDT members have enabled the translation and further development of numerous vaccine candidates discovered in NEWTBVAC. These activities were instrumental for bringing candidates such as ChAdOx1.85A MTBVAC, VPM1002, HBHA and BCGΔZMP1 to their current preclinical or clinical stages of development.
TB vaccine development remains a complex process comprising discovery research, preclinical development and clinical assessment. At all stages NEWTBVAC has successfully completed its major goals and thus contributed to an impressive global pipeline of novel vaccine candidates at different development stages. NEWTBVAC was the critical driver of European TB vaccine development as well as the constructive coordinator of European activities with international partners, notably the Aeras Foundation supported by the Bill & Melinda Gates Foundation in the US. With these other stakeholders NEWTBVAC leadership has a stringent gating strategy based on rational and broadly accepted criteria that allows selection/down-selection of candidates of the TB vaccine portfolio, in order to develop only the most promising candidates. This approach is the basis for the establishment of the Global TB Vaccine Partnership (GTBVP) which aims to manage a global portfolio by advising the support only the most promising preclinical and clinical stage candidates through this well-defined selection process.

Project Context and Objectives:
NEWTBVAC proposed the discovery and the preclinical development of a new generation of tuberculosis (TB) vaccines. The proposal comprised:
(1) the sustainable development of the European pipeline of new vaccine discoveries and the advancement of promising candidates into clinical trial assessment;
(2) the development of new second-generation vaccines based on novel prime–boost strategies and/or new combinations of promising subunit vaccines with the aim to reduce active TB disease in latently infected TB-infected individuals (LTBI);
(3) the discovery, evaluation and application of new biomarkers relevant for monitoring of clinical TB vaccine trials.
The final report of NEWTBVAC demonstrates the great success of this proposal in all of these areas. The project was divided into seven work packages (WP) that coordinated the different topics. WP1 focused on the identification of important novel bacterial antigens. WP1 thereby could introduce relevant antigens into WP5 and WP6 where their value for vaccine design was evaluated in immunologic assays or animal models, respectively. WP2 focused on the development of novel subunit vaccines and antigen delivery systems, thereby closely interacting with WP1 in an attempt to design effective antigen adjuvant formulations, as well as WP5 and WP6, for the evaluation of novel antigen delivery systems in animal models or immunologic assays, respectively. WP3 focused on the development of live vaccines comprising both recombinant BCG and Mycobacterium tuberculosis (Mtb) strains. Several effective vaccine candidates could be developed that proved highly efficient in animal models (WP5) and thus some of these could be brought forward into the product and clinical development pipeline with stringent supervision of the product development team (PDT) and the clinical development team (CDT) as panels being part of WP7. WP5 provided the experimental animal platform for preclinical efficacy and safety testing and developed novel improved animal models. In WP6, novel biomarkers were identified and platforms for validation of most relevant biomarkers for vaccine development established. WP7 provides a link between preclinical research and development (R&D) and product development as well as the link to early clinical testing. It thus coordinated the equally important and demanding task of translating research findings into clinical development.

Overall, NEWTBVAC has been instrumental in bringing the complete TB vaccine pipeline forward at all stages. Novel antigens which are now tested in clinical vaccine trials had been discovered within NEWTBVAC. Novel viral carriers and adjuvants have been developed for next-generation vaccine candidates. Novel live vaccine candidates, notably, the Mtb-based double deletion mutant MTBVAC have been translated from preclinical R&D into clinical trial assessment of safety and immunogenicity. Finally, several highly promising biomarkers have been developed and validated for use of monitoring in clinical trials.

Thus, NEWTBVAC has proven its strength by providing a full pipeline from early discovery in the wet lab, to early clinical trial testing resulting in the development of novel TB vaccine candidates, and also laying the basis for the development of next-generation vaccines. NEWTBVAC therefore has created a successful and sustainable European collaborative network comprised of basic researchers, applied developers of vaccine formulations, experienced coordinators and managers of preclinical and clinical development pathways as well as clinical study performers.
Although the success of numerous projects in all WPs is overwhelming, in the following some of the most innovative highlights are being selected.

In WP1, lipoid antigens were identified which could provide a totally new class of vaccine antigens in adjunct to canonical protein antigens. Similarly important could be the characterization of subdominant antigens which could complement dominant antigens in future vaccine formulations.
In WP2 and WP4, a novel next-generation viral vector was developed, namely, Chimpanzee adenovirus, which can be not only used as booster on top of BCG or other novel BCG replacement vaccines, but also combined with other viral vector-based subunit vaccines. WP2 also focused on the development of HBHA-based protein adjuvant formulations with a focus on methylated and nonmethylated HBHA peptides.

The comprehensive list of numerous novel viable vaccines for BCG replacement strategies in WP3 is overwhelming. One vaccine candidate, MTBVAC, developed previously in this WP has now entered clinical testing. Nevertheless this vaccine candidate will be further improved. Similarly, the recombinant BCG vaccine VPM1002, already in phase II clinical trial and originally developed by NEWTBVAC consortium members, has been modified by genetic means to further improve its vaccine efficacy and safety. This rich panel of novel viable vaccines is complemented by totally new constructs which in the near future will be ready for product development and subsequently for clinical trial testing.

The animal model platforms provided in WP5 proved to be an instrumental core service for vaccine development. Not only did WP5 provide the relevant information needed for decisions about the future fate of vaccine candidates within NEWTBVAC, it also provided a platform for head-to-head testing of several vaccine candidates, further supporting development of next-generation vaccine candidates. In addition, WP5 developed models which could be of great help for the future gating of novel vaccine candidates.

WP6 successfully developed a platform of biomarker assays that also will provide essential information for vaccine development and selection of most promising candidates. The biomarker field is still in an emerging and developing stage, and hence the design of decision-making matrices as developed in WP6 will help in the gating of future vaccine candidates as well as in the stratification of most relevant study populations of future vaccine trials. The activities of WP1 to WP6 as well as the decision making process as to which vaccine candidates to move forward, was coordinated in WP7, where several experienced and respected vaccine developers contributed as members of the PDT and CDT structures.

Because of the great success of NEWTBVAC and other organizations in bringing forward highly promising vaccine candidates, a stringent gating strategy needs to be developed to bring forward the most promising candidates based on rational and broadly accepted gating criteria. By including different stakeholders from Europe and the US, this process has been brought forward significantly under the leadership of NEWTBVAC. In addition, the PDT and CDT members of WP7 have monitored the progress of numerous vaccine candidates as developed in NEWTBVAC. These activities were instrumental for bringing MTBVAC into phase I clinical trial and the further transition of VPM1002 from phase I into phase II trial. It has also been instrumental in coordinating preclinical development stages for the HBHA-based vaccines and the recombinant BCG zmp1.

In sum, TB vaccine development is a highly complex process comprising basic research, preclinical development and clinical assessment. At all stages of this complex matrix NEWTBVAC has successfully completed its major goals and thus established an impressive pipeline of novel vaccine candidates at different development stages. Importantly, NEWTBVAC was the critical driver of European TB vaccine development as well as the constructive coordinator of European activities with international partners, notably the Aeras Foundation supported by the Bill & Melinda Gates Foundation in the US. Importantly, NEWTBVAC has not restricted itself to R&D aspects but also included into its strategy, management structures to make is possible to bring forward the most promising candidates through well-defined selection processes.

Project Results:
3. Description of the main S&T results/foregrounds

WP 1

Background:
The goal of WP1 was the identification of novel mycobacterial antigens and their evaluation in immunological assays (WP6) and animal models (WP5). Three approaches were taken to enhance the existing pipeline of vaccine antigens:

1. identification of subdominant epitopes
2. stage specific antigen discovery
3. identification of lipid antigens
ad1. In the past, identification of Mtb epitopes has primarily been performed by analysis of the antigen specificity of isolated Mtb-specific T cells. However, from studies in other model systems we know that epitopes that are dominant immunogens may not be expressed at high levels on infected cells and vice versa, epitopes that are highly expressed on infected cells are not invariably immunodominant. In addition, it might well be that subdominant epitopes (not inducing strong immune responses during natural infection) turn out to be the better vaccine components. The objective was therefore to examine the protective capacity of subdominant epitopes and develop methods and vaccine constructs that induce them.
ad2. Emerging evidence indicates that latent Mtb includes different stages such as non-replicating bacteria, resuscitating bacteria, and bacteria returning to a non-replicating stage due to the immune response against them. Each of these stages is associated with expression of stage specific proteins. From a vaccine point of view, it is therefore important to study Mtb at these different stages, and ideally produce vaccines against all Mtb stages.
ad 3. The immunogenicity of mycobacterial lipids in vivo remains poorly defined. In order to understand the role of lipid-specific T cells in the immune response during mycobacterial infection, it is important to identify which lipid molecules induce specific immunity and whether they are part of the target of protective immunity. The proposed studies had the aim to systematically isolate immunogenic lipids from virulent Mtb and evaluate their Immunogenicity in patients with active or latent tuberculosis and in CD1b transgenic mice. The possibility that new lipid antigens modulate innate immune responses and influence immunity to Mtb was also considered within this work package.
Highlights
In accordance to our initial deliverables and milestones our work identified a panel of new mycobacterial antigens which were evaluated for immunogenicity in vitro and tested for protective efficacy in vivo. All subworkpackages- subdominant epitope discovery, stage specific antigen discovery and lipid antigens successfully contributed to filling the pipeline with promising vaccine candidates:

- TB10.4 peptides lead to strong protection in CB6F1 mice
- Novel immunogenic epitopes were eluted from MHC class I and MHC class II molecules of macrophages infected with Mtb or with MVA vectors encoding mycobacterial antigens
- Dimers of the Esx secretion system (H65) were as protective as BCG in mice
- Rv2034 shows comparable efficacy to BCG after challenge of mice with Mtb
- Immunization with diacylated sulfoglycolipid reduces the bacterial load in the spleen of guinea pigs challenged with Mtb and induces T cell responses in CD1b transgenic mice.

Key findings
1. Prophylactic vaccination using a pool of overlapping TB10.4 peptides induces both dominant and subdominant T cell responses, and leads to improved protection in CB6F1 mice (SSI).

2. A truncated ESAT-6 (D15-E6) induced subdominant T-cell responses and provided significant protection upon Mtb-challenge. Dual vaccination with ESAT-6 and D15-E6 provided more protection than vaccination with individual peptides. This additive/synergistic effect is possibly due to the induction of distinct CD4+ T cell responses in vivo regarding differentiation and polyfunctionality Therefore the simultaneous induction of dominant and subdominant T cell responses is a promising approach to improve vaccine efficacy. (SSI).

3. CD4 T cells targeting subdominant epitopes against ESAT-6 are resistant to infection-driven terminal differentiation and can enhance the control of Mtb infection in both prophylactic and post-exposure models of Mtb infection (SSI).

4. Eleven mycobacterial HLA class I ligands were identified, validated and reproduced from B-LCL cells infected with MVA vectors containing Ag85A and TB9.8. Two of these ligands induced IFN- release from T cells obtained from primed human donors. Optimized infection strategies yielded four MHC class II restricted peptides. 2 MHC class I ligands (Ag85A and TB9.8) induced IFN- response in T cells and were thus verified as epitopes. Four MHC class I restricted epitopes were eluted from Mtb-infected macrophages, two of which induced Th1-skewed T cell responses in human T lymphocytes. These results establish the elution of mycobacterial peptides from infected macrophages as a novel approach to identify immunogenic epitopes from Mtb (UTUB, UULM)

5. A fusion protein with Rv1284 and other prototype late antigens including Rv2659, Rv2660, and Rv2661 was constructed. As a preventive vaccine, this construct (designated H66) gave rise to significant protection. The fusion protein Ag85B-ESAT-6-TB18.2 (H67) shows strong and specific immune activation in a post-exposure vaccination model. In addition, an Esx dimer substrate for each of the five ESX secretion systems was selected and combined into a fusion protein with one of the main criteria being their expression throughout Mtb. infection. The EsxD-EsxC, ExsG-EsxH, and ExsW-EsxV were selected and combined into a fusion protein, H65. Prophylactic vaccination with H65 gave protection at the level of BCG and the fusion protein exhibited high-predicted population coverage in high endemic regions. (SSI).

6. Three in vivo expressed tuberculosis antigens (IVET) were identified and are immunogenic in HLA-DR3tg mice. One of these antigens (Rv2034) induced Th1-biased T-cell responses in HLA-DR3 transgenic mice. Analysis of a RV2034-specific T cell clone demonstrated polyfunctionality and antimicrobial activity against intracellular Mtb suggesting an association with protective host responses. Studies focusing on the protective value of Rv2034 combined with CpG and CAF09 were performed showing significant reduction of CFU in the lungs of HLA-DR3 tg mice following Mtb challenge. TB challenge experiments in guinea pig using Ag85B-ESAT6-Rv2034 showed protection comparable to BCG in the spleen and marginally comparable to BCG in the lungs. (LUMC).

7. Two plasmid DNA injections followed by a protein boost of three DosR encoded antigens (Rv2626c, Rv2627c, Rv 2628) induced strong Th1 responses in the lungs of H-2d mice but failed to protect from intratracheal challenge with wild type Mtb or Mtb constitutively expressing the DosR regulon (WIV).

8. Lipoarabinomannan is the most immunogenic lipid antigen for human T cells and natural infection with Mtb triggers long lasting memory responses. The strongest response to mycobacterial lipids was found in donors that have completed successful treatment for tuberculosis. To improve the immunogenicity, LAM was incorporated into hydrophobic liposomes termed “LIPLAM”. LIPLAM induces T-cell responses in primed individuals (UULM).

9. Cationic liposomes were developed for optimized delivery of lipid antigens in guinea pig experiments. Guinea pigs vaccinated with Ac2SGL/PIM packed into liposomes had a lower bacterial burden in the spleens than control animals (CNRS).

10. Immunization of CD1b transgenic mice with Ac2SGL induces lipid-specific, Th1-skewed- and CD1b-restricted T cell responses (UNIBAS)

11. Several mycobacterial lipid fractions and importantly the newly identified immunogenic diacylated sulfoglycolipid (Ac2SGL) interfere with the differentiation of monocytes to dendritic cells (ISS).

Summary and Perspective: All projects within WP1 progressed as planned and finished successfully at the end of February 2014. Several promising antigens that induce protective T cell subsets (e.g. polyfunctional- and polycytotoxic T cells) were identified. The majority of vaccine candidates was already tested for immunogenicity in humans and/or animal models and some have already advanced to efficacy studies against Mtb-infection. Another major outcome of this work package was the establishment of new tools to refine antigen discovery. The elution of peptides from MHC-molecules, improved delivery systems for hydrophobic antigens and MHC/CD1- transgenic mice for evaluating vaccine efficacy were developed and tested successfully. These findings provide a profound basis for accelerating antigen discovery in the near future. The increased portfolio of newly discovered antigens and the availability of state of the art technology will be exploited in future projects submitted within the Horizon2020 call of the European Union.

WP 2 & 4
Background:
This workpackage is focussed on development and optimisation of subunit vaccines as candidate TB vaccines. There are four main components to this workpackage: (1) Viral vectored TB vaccines; (2) Optimisation of protein/adjuvant formulations; (3) Optimisation and immunogenicity studies with HBHA; (4) Immune modulation.

Key findings from this workpackage:
(1) Viral vectored TB vaccines
1. A simian adenovirus expressing antigen 85A, ChAdOx1.85A is immunogenic and can confer some limited protection against M.tb challenge in mice.
2. An optimised regime of BCG prime, followed sequentially by simian adenovirus expressing Ag85A administered intranasally, followed by MVA85A administered intranasally or intradermally, is reproducibly significantly more protective than BCG alone. This promising BCG – ChAdOx1.85A – MVA85A vaccination regime is now being evaluated in guinea pigs and non-human primates. There is also a phase I first time in man clinical trial ongoing in Oxford with ChAdOx1.85A administered to BCG primed healthy adults alone, or in a prime – boost combination with MVA85A.
3. Four potentially promising antigens identified from studies in patients with LTBI, were inserted into the simian adenoviral backbone (ChAdOx1). These 4 constructs were immunogenic, and in an aerosol M.tb challenge experiment, conferred significant protection when used alone. The 4 promising new antigens are now being inserted into MVA vectors and will be evaluated in collaboration with the ChAdOx1 constructs to determine the most protective antigen / vector combination.
4. A range of first and second generation vaccine vectors against M.tb based on replication-deficient LCMV were generated. Based on immunogenicity for both CD4 and CD8 T cells, these were down selected to rLCMV expressing Ag85B and TB10.4 (rLCMV/Ag85B-TB10.4). This vector is well tolerated and highly immunogenic, in neonatal and adult mice.
5. A first M.tb challenge experiment in C57BL/6 mice vaccinated with rLCMV/Ag85B-TB10.4 has demonstrated protection against pulmonary TB.
6. EPI interference immunogenicity experiments at UNIGE have not found any interference of EPI vaccination with simultaneous or subsequent vaccine responses to neonatal immunization with rLCMV/Ag85B-TB10.4.

(2) Optimisation of protein/adjuvant formulations
1. The Mingle ligand and PolyI:C are synergistic for the induction of CD4+ and CD8+ T cell responses. The combination of MMG and PolyI:C induces high levels of CD8+ T cell responses.
2. The combination of MMG and PolyI:C was highly stable over time.
3. When TDB was combined with the CpG in DDA liposomes, there was potent CD8+ T-cell induction directed towards a model antigen after s.c. administration but this was not reproducible with a TB antigen. Interestingly, the combination of TDB and CpG in DDA liposomes was found to enhance IL-17 and IL-2 production by CD4+ T cells irrespective of the antigen.
4. Using pegylated CAF09 (DDA, MMG and Poly(I:C)) leads to increased levels of antigen-specific CD8+ T cells. However, although CD8+ T cell induction has been optimized, this has not resulted in improvements in protection against M.tb challenge in vivo. This data suggest that vaccine promoted CD8+ T cells may have a limited role in protection against M.tb.
5. Evaluating the signaling pathway for MMG, an absolute requirement for the Mincle-FcRgamma signaling pathway was identified, but not the TLR-Myd88 pathway, in macrophage activation, indicating that MMG activates murine macrophage via the same mechanism as the mycobacterial cord factor Trehalose-dimycolate. The data suggest that MMG has two faces: on one hand it can bind to CD1b and trigger non-classical T cells in the human system, on the other it acts as a PAMP by activating the PRR Mincle in the mouse.
6. Neither mouse nor human Mincle-complementation by retroviral gene transfer is sufficient to confer responsiveness to MMG. In contrast, the Mincle of both species complements the induction of G-CSF expression by TDB.
7. Neutrophils and inflammatory monocytes were among the first cell types being recruited upon immunization with MMG. Neutrophil recruitment was equally dependent on recognition via Mincle and MyD88 signalling, whereas influx of inflammatory monocytes depended on Mincle to a large and on MyD88 to a lesser extent. Neutrophil depletion did not impair the capacity of CAF01 to induce Th1 and Th17 response, whereas blockade of CCR2 by antibodies strongly reduced the response. The use of CCR2-/- mice showed that this chemokine receptor is pivotal for the adjuvanticity of TDB/CAF01. Given the dependence of monocyte mobilization from the bone marrow and immigration into tissues on CCR2, these data together suggest a major contribution of monocytes, but not neutrophils, to the Th1/Th17-inducing activity of TDB/CAF01.
8. The induction of CD4+ T cell responses requires strong physical interaction of the antigen and the adjuvant to allow their co-ordinated delivery to the immune system.
9. There is a synergy between non-TLR (MINCLE) and TLR ligands leading to the induction of CD8+ and CD4+ T cell responses in adult and neonatal primed mice.
10. Non-TLR and TLR ligands can be effectively co-formulated, and after immunisation the ligands remain co-located at the injection site with the cationic liposomes. In terms of immune response, the addition of CpG was shown to enhance the efficacy of DDA/TDB liposomes; however immune responses were not further enhanced by the addition of polyI:C to the formulation.
11. The cationic component of the liposome delivery system facilitates increased antigen delivery to antigen presenting cells and the recruitment of monocytes to the site of injection. The monocyte infiltration to the site of injection and the production of IFN-γ upon antigen recall was markedly higher for liposomes formulated from DDA and DC-Chol-based liposomes which exhibited a longer retention profile at the site of injection compared with lower transition temperature lipids. Therefore high-transition, quaternary ammonium lipid structures should be employed in the design of cationic lipid adjuvants as they offer long term retention of the antigen and a slow release of liposome and vaccine antigen from the injection site.
12. Using low levels of polyethylene glycol (PEG), we can promote earlier antibody responses without impacting on antigen retention of the formulation. High levels of PEG, reduced antigen retention by the formulation, therefore reduced stability, resulted in lower Th1 activity and pushed the response to a more Th2 bias. TLR3 and TLR9 could be added to the liposome formulation without altering stability. Cationic charge is an important factor for the retention of the liposomal component at the site of injection - a moderate to high (>50%) level of antigen adsorption to the cationic vesicle surface is required for efficient antigen retention at the injection site.

(3) Optimisation and immunogenicity studies with HBHA
1. Results obtained by the screening of 8 LTBI subjects allow us to definitively conclude that non-methylated peptides are not recognized.
2. No class II HBHA peptide was identified by the elution of naturally processed peptides from HBHA-loaded APC. This difficulty is in line with the intrinsic properties of the HBHA protein, and the probably poor abundance of HBHA peptides among all the eluted peptides present at the cell surface of APC.
3. So far, it has not yet been possible to produce recombinant methylated HBHA in E. coli by co-expression of HBHA and the relevant methyltransferases within the same E. coli cell, although numerous candidate genes have been tested. A bioinformatics approach is now being pursued.
4. Several immunogenicity and prime-boost studies with HBHA, alone and in prime-boost combination have been carried out with some promising results; several experiments will continue beyond NEWTBVAC. Aeras have constructed a recombinant strain of BCG expressing HBHA which is currently being evaluated in immunogenicity and challenge experiments.
5. Peptide elution from methylated HBHA continues.

(4) Immune modulation.
1. PD-1 expression was found to be up-regulated on CD4+ and CD8+ T cells from active tuberculosis patients compared to latent infected and healthy individuals. PD-L1 is highly expressed on myeloid DCs ex-vivo isolated from tuberculosis patients compared to latent infected and healthy individuals. PD-L1 expression is also up-regulated on monocyte derived-DC cells infected with rBCG. Results show that PD-1/PD-L1 blockade abrogated the suppression of T cell responses induced by ManLAM, suggesting that inhibition of PD-1/PD-L1 pathway rescued ManLAM specific-T cell responses and amplified cytokine secretion in chronic pulmonary tuberculosis patients.
2. Recent findings showed that Mtb-specific T cells expressing PD-1 do not present an exhausted phenotype; on the contrary, they have high proliferative capacity and cytokine secretion. In addition, adoptive transfer experiments revealed that CD4+ T cells exacerbated disease when released from PD-1 inhibition. Based on the above mentioned findings, we considered blocking of PD-1/PD-L1 pathway not a valuable strategy to be pursued to improve T cell responses after vaccination. However, we still consider blocking of negative signaling pathways a promising strategy to improve endogenous T-cell responses and vaccine efficacy. Therefore we suggest exploiting other inhibitory pathways as targets for immunomodulation in the future.
3. Despite published evidence of TIM-3 upregulation in C57BL6 mice following aerosol infection with MTB, MPIIB observed no phenotype during infection of TIM-3-/- mice in a similar system. Redundancies in biological negative feedback systems for activated T cells may be able to compensate for TIM-3 knockout in our mouse model of mycobacterial infection.
4. rBCG increased the CD4 and CD8 T cell response compared to SSI vaccinated controls. This was true in the lymph nodes, spleens and lung post vaccination. rBCG also led to an increase in the proportion of central memory (TCM) CD4 T cells that remained in the months following vaccination, suggesting these cells may underlie protection upon challenge. Adoptive transfer of small numbers of endogenous or transgenic M.tb specific TCM was sufficient to inhibit pulmonary tuberculosis. The number of classical T effector memory (TEM) cells were similar after both vaccines. rBCG also drove an increase in antigen specific TFH cells, which was accompanied by an enhanced antibody response.

Summary and perspective
The work conducted in this WP has gone very well and there have been many significant findings, as highlighted above. Work on viral vector delivery systems, optimising adjuvant formulation and work on HBHA as a potential candidate antigen for a new TB vaccine will continue and will be the subject of further EC grant applications within the H2020 programme.


WP 3
Overall objectives
Using molecular mycobacterial genetics, the overall objectives of WP3 were to:
• Improve already existing TB live vaccine candidates (e.g. MTBVAC, VPM 1002) generated in previous projects;
• Generate novel TB live vaccine candidates based on genetically modified M. tuberculosis and other species of the TB complex (M. microti, M. bovis and M. bovis BCG);
• Understand the molecular bases of protection conferred by live vaccine candidates;
• Test safety, immunogenicity and protective efficacy of live vaccine candidates in animal models (in collaboration with other NEWTBVAC partners), and propose vaccine candidates for phase I clinical trial.
Main results & achievements
Our main results are summarized in the Table below, that shows safety (in SCID mice), immunogenicity in human isolated immune cells (mostly dendritic cells, DCs) in vitro and in mice in vivo, protection efficacy in mice, and protection efficacy in guinea pigs (GP) of all the live vaccine candidates generated during the 4 years of the project. Some strains, generated but not tested so far in animal models, are not shown in the Table below.
Strain Safety Immunogenicity in human cells (DCs)1 Immunogenicity in mice Protection efficacy (mice) Protection efficacy (GP) Comments
Recombinant BCG strains
Man°LAM2 =BCG N.D.3 =BCG =BCG =BCG
+Ac2SGL4 =BCG =BCG N.D. N.D. =BCG
+ESX1(mod)5 =BCG >BCG >BCG >BCG ≥BCG
Moreau ESX16 =BCG N.D. >BCG >BCG N.D.
VPM-ΔnuoG7 >BCG
=VPM N.D. >BCG
=VPM >BCG
>VPM =BCG
=VPM
VPM-Δpdx18 >BCG
=VPM N.D. >BCG >BCG N.D. Better clearance than BCG & VPM
VPM-ΔsecA29 >BCG
=VPM N.D. N.D. =BCG
VPM-IL7/IL18/IFNγ10 N.D. N.D. >BCG
>VPM >BCG
=VPM N.D.
Δzmp111 =BCG N.D. >BCG =BCG >BCG Currently tested in primates
Recombinant M. tuberculosis, M. microti & M. bovis strains
GC1237-ΔRv1503c12 =BCG
BCG =BCG ≥BCG14
M. microti >BCG N.D. N.D. =BCG M. microti ESX115 =BCG N.D. >BCG >BCG =BCG
ΔESX516 =BCG N.D. >BCG >BCG N.D.
MTBVAC- Δerp17 >BCG
>MTBVAC N.D. =BCG =BCG
MTBVAC- Δzmp118 >BCG
=MTBVAC N.D. =BCG =BCG
ΔsigE19 =BCG >BCG >BCG >BCG ≥BCG
ΔsigE-ΔfadD2620 =BCG N.D. N.D. >BCG
≥ΔsigE ≤BCG21
M. bovis-Δmce222 M. bovis-ΔlprG23 BCG =BCG
1. In human DCs, candidates were tested for their ability to induce DC maturation and cytokine production (e.g. IL12).
2. This strain expresses a ManLAM lipoglycan that is void of mannose caps, thought to be involved in immune suppression
3. N.D. not done
4. This strain is genetically modified to express the lipid antigen diacylated sulfoglycolipid (Ac2SGL) that is not present in wild-type BCG, and is a potent CD1-restricted antigen recognized by T cells during natural TB infection in human. Because only GP, but not mice, express the CD1b molecule involved in Ac2SGL presentation to T cells, this candidate was tested for protective efficacy in GP only in a prime/boost regimen where animals were primed with the candidate and boosted with purified Ac2SGL.
5. This strain contains a modified version of the ESX1 locus, normally absent in all BCG strains and present in M. tuberculosis.
6. This BCG moreau strain is naturally void of immunosuppressive PDIM lipids and contains the ESX1 locus, normally absent in all BCG strains and present in M. tuberculosis.
7. This strain is a modified version of VPM 1002 (BCG-ΔureC::hly) in which the anti-apoptotic gene nuoG is deleted.
8. This strain is a modified version of VPM 1002 in which the gene pdx1, involved in vitamin B6 biosynthesis, is deleted.
9. This strain is a modified version of VPM 1002 in which the anti-apoptotic gene secA2 is deleted.
10. This strain is a modified version of VPM 1002 expressing the indicated cytokines.
11. This strain is a modified BCG in which the anti-inflammatory gene zmp1 is deleted.
12. This M. tuberculosis strain of the W-Beijing lineage (GC1237) is deleted for Rv1503c, a gene involved in phagosome maturation arrest.
13. The mutant is as safe as BCG in a low-dose intranasal challenge model, and less safe than BCG in a high-dose intravenous challenge model in SCID mice.
14. The mutant is as protective as BCG in the lungs and more protective than BCG in the spleen in GP.
15. This M. microti-based candidate contains the ESX1 locus.
16. This H37Rv strain is deleted for the ESX5 locus.
17. This strain is a modified version of MTBVAC (M. tuberculosis deleted for phoP and fadD26), further deleted for the virulence gene erp.
18. This strain is a modified version of MTBVAC, deleted for the anti-inflammatory gene zmp1.
19. This H37Rv strain is deleted for the transcriptional regulator SigE.
20. This H37Rv strain is deleted for the transcriptional regulator SigE and FadD26, involved in PDIM synthesis.
21. This mutant is as protective as BCG in the lungs and less protective in the spleen in GP.
22. This M. bovis strain is deleted for the virulence gene mce2.
23. This M. bovis strain is deleted for the lipoprotein LprG.
Conclusions
Overall, our WP generated over 30 novel live vaccine candidates. Based on our results we were able to classify these vaccine candidates according to their priority for further development. Following are our conclusions:
1. First and foremost, we were very happy that one of these candidates (BCG-Δzmp1, generated by Drs. P. Sander & E. Boettger) showed the most promising results in terms of safety in immune-compromised animals and protective efficacy in GP. This candidate is currently being tested in non-human primates, and will enter clinical development for first-in-man phase I trial soon.

2. Based on our results, other candidates are selected for further development:
• GC1237-ΔRv1503c
• BCG Moreau::ESX1
• H37Rv-ΔESX5
• MTBVAC-Δerp
• VPM-ΔnuoG
• VPM-Δpdx1
• H37Rv ΔsigE-ΔfadD26

All the other candidates were terminated, as they did not show any promise, as compared to BCG, in terms of safety and/or protective efficacy.
3. One of the major conclusions of our work is that combining two promising strategies, when considered individually, does generally not provide improvement over one strategy or the other (e.g. the VPM-Δzmp1 or the MTBVAC-Δzmp1 are not better than VPM, MTBVAC or BCG-Δzmp1).

4. Importantly, our WP generated over 30 publications in international peer-reviewed journals, in which NEWTBVAC and FP7 support is acknowledged. In a number of these publications, two or more partners of the WP are involved as authors, reflecting the excellent communication and collaboration level within the WP.

 
WP5
WP5 in Numbers
With the ultimate objectives of stimulating TB vaccine research and discovery, and advancing promising candidates into clinical stages of vaccine evaluation, NEWTBVAC comprised a dedicated workpackage – WP5 - to exploit animal models supporting preclinical development. More established models for vaccine evaluation as well as exploratory models – a total number of 9 preclinical platforms - were offering along the programme capacity to the consortium and in particular to the workpackages 1, 2 and 3 on antigen discovery, subunit vaccines and antigen delivery systems, and live vaccines, respectively. Availability of investigational slots was expressed by open calls to all members and through the coordinating office, TBVI. Applications were collected, discussed and ranked according to priority by a User Selection Panel (USP), composed of experts that were independent of any of the developers labs, and complemented by the principle investigator of the respective animal model facility.

About 50 candidate vaccines or regimes have been considered for preclinical testing in either of the respective models, and in 18 USP meetings over 4 years highest priority proposals have been identified. Ultimately, 35 investigational regimes evolving from the NEWTBVAC consortium have been tested in one or more of the models, and 7 additional control treatments were included next to non-vaccinated and standard BCG vaccinated control groups (see table below). Of these 35 candates, 21 were so-called live vaccine candidates evolving from WP1 and based on modifications of M.bovis BCG or other genetically engineered mycobacterial species (including M.tuberculosis M.bovis (wild-type) and M.microti). The remaining 14 candidates were subunit vaccines either delivered in specific adjuvant formulation or by vector (i.c. plasmid DNA). Four out of those 35, were defined as heterologous strategies vaccinating in combination with standard BCG simultaneously or as a BCG prime-boost treatment. During the life time of the NEWTBVAC programme no virus vector/particle-based vaccines have been considered for testing in any of the dedicated animal models of WP5. Overall, the capacity planning for WP5 was found to be well balanced, although in a few cases selectable candidates were not tested due to capacity limits. While all in vivo stages of the animal experiments have been finalised before the end of the programme, a few downstream evaluations of some experiments are still awaited at this point. All deliverables have been provided and specific study results and reports have been sent to the vaccine developers as data became available.

Preclinical safety
For live attenuated mycobacterial vaccines, that are being developed as a possible BCG replacement, addressing vaccine safety is of utmost importance in the vaccine development process. To this end, not only regular immunocompetent models (mostly mouse), but in particular also the highly sensitive, immuno compromised, severe-combined immuno-deficient (SCID) mouse model comprises a relevant tool. Lacking major adapitve immune components, these SCID mice provide a robust and sensitive indicator of residual virulence. In a few runs a total number of 11 candidates have been evaluated, one of which represented a specific parental strain for additional internal control purpose. Considering the 10 novel live candidate vaccines, 1 was found to be more virulent than BCG and will, thus, need to be considered for further engineering to obtain an acceptable level of attenuation. Five out of 10 were similar to BCG. Interestingly, 4 candidates were found to be safer than BCG by readout of SCID mouse survival upon experimental infection. One of those safer than BCG was a makerless derivative of the urease-deficient, listeriolysin-expressing recombinant BCG that is currently advanced in clinical trials as VPM1002. Another candidate performing better than standard BCG was a BCG derivative complemented for the expression of RD1 with specific point mutations. Finally, also 2 independent constructs on different BCG backgrounds and both deficient in the expression of the zmp1 gene product were found safer than BCG in SCID mice. The SCID study results of the latter candidates are supportive of further development and strengthen their possible product portfolio.
As part of the NEWTBVAC programme a safety evaluation of different administration routes using SCID mice has been considered. However, these activities have not been pursued due to logistical issues and time constraints. Upon continued extensive discussion within NEWTBVAC it has been confirmed that harmonisation on intraveneous administration in the SCID mouse model is essential for comparability.

Preclinical efficacy
As there is no conclusive definition of the nature of protective TB immunity, animal models provide essential tools to provide proof-of-concept for the protective efficacy of new vaccine candidates against experimental challenge with M.tuberculosis (M.tb). While mouse models mostly present a first gate for such efficacy evaluation, WP5 harboured specific advanced model capacity for NEWTBVAC’s vaccine developers.
In collaboration with South Korean partners at Yonsei University and supported by local governemental subsidies specific Beijing type K strain challenges were set up in mice to establish a more stringent test modality in comparison to lab strain challenges typically using M.tb H37Rv (or Erdman). It was established that Beijing strain K – a predominant clinical strain in local outbreaks – replicated at least 10 times more rapidly than H37Rv at early stage mouse infection. Concordantly, survival times of infected mice were significantly reduced and histopathological changes enhanced, while cytokine levels (e.g. IFNg, IL12p40) in the lungs were generally lower. These findings underpinned the higher virulence of Beijing K strain well over H37Rv, and K strain was subsequently used for infectious challenge upon investigational vaccination with two vaccine candidates from NEWTBVAC. Of the two a sigE-, fadD-deficient mutant of M.tb was found to convey improved protection by bacterial load as well as histopathology readout at 1 and 2 months post-infection and over standard BCG control vaccination.
A routine model for propagating candidates that demonstrated efficacy in mice is established in high suspceptibility guinea pigs. Typically, groups of N=8 female outbred Dunkin Hartley guinea pigs were undergoing preventive vaccination after which animals were rested and challenged by a standardised and aerosolised dose of 10-50 CFU of M.tb H37Rv (Collison nebuliser, mean aerosol diameter of 2 m) for a relatively short-term follow up post-infection. In total 22 candidates and 3 additional controls, next to standard BCG and saline, have been enroled into guinea pig experiments under NEWTBVAC WP5. Mostly live attenuated vaccines have been tested and only three candidates were evaluated as prime-boost regimes using BCG for primary immunisation.
While few candidates failed to demonstrate protective efficacy by bacterial burden in lung or spleen, remarkably improved protection over standard BCG has also been observed. In particular, a recombinant M.tb strain rendered deficient in Rv1503c gene expression was found to significantly reduce splenic mybacterial counts post-H37Rv infection. Deficiency of the zmp1 gene, knocked out in two independent constructs using either M.bovis BCG Pasteur or BCG Danish as parental strains, resulted in reduced bacterial burden in the lungs of infected guinea pigs with either of these constructs. The reproducibility of its protective capacity in this model strongly supports the portfolio of this candidate.
The latter recombinant zmp1 deficient BCG on the Danish background has also been tested in the final year of NEWTBVAC in a standardised non-human primate (NHP) model using treatment groups of N=6 rhesus macaques and a (high-dose) challenge with M.tb harmonisation strain Erdman K01 for efficacy readout. The zmp1 deficient BCG, applied as a single intradermal injection for primary vaccination, was found more immunogenic in rhesus, with accellerated immune kinetics over BCG in particular. Exploiting a relatively high dose challenge of a targeted 500 CFU of M.tb strain Erdman by intrabronchial instillation as a highly stringent test condition, it was found protective by showing less lung pathology compared to non-vaccinated controls. Reduced necrotic areas in comparison to standard BCG in 3 out of 6, although not statistically signficant, might be considered a positive indicator of improvement beyond BCG. Another two investigational slots in this NHP experiment were used to evaluate the H56 subunit vaccine formulated in a novel proprietary cationic adjuvant formulation, CAF09. A homologous prime-boost regime of H56 in CAF09 - although inducing vaccine antigen specific immunity - did not show any protective efficacy on its own. However, a heterologous regime on top of a standard human dose of BCG Danish 1331 for primng, did provide statistically significant improvement of lung pathology as well as tracheobronchial lymph node involvement better than standard BCG only in this stringent high susceptibility model in support of H56 as a potent booster vaccine. Assessment of clinical measures, histopathology and in particular bacterial load in specific organs is still ongoing and continues beyond the life time of NEWTBVAC and this reporting time point.

Investigational models
As a relevant aspect of clinical vaccination, activities were specified in the WP5 programme to investigate maternal pre-exposure in a mouse model. These activities were enabled with support of South Korean governmental funding at the International Vaccine Institute, Seoul. Neonates were obtained from animals that were previously exposed to live M.bovis BCG, M.avium or heat-killed M.tb strain H37Rv, or from naive control mothers. At one week after birth these cohorts of neonates were either BCG vaccinated or left untreated, after which BCG and PPD specific IFNg responses were measured and protection assessed by experimental infection with M.tb H37Rv intranasally. Whereas BCG was found protective, the maternal immune status to M.tb by previous exposure did not impact neither on immunogenicity nor on the protective cacacity of standard BCG.
As part of a contingency plan, neonatal mice were also offered as a platform for vaccine testing. Plasmid DNA vectored mycobacterial antigen 85A provided enhanced immunogenicity, but no benefits over BCG only control vaccination in this model. Another vaccine is still being evaluated; data are awaited beyond this final reporting time point.
Beside using mice the aspect of vaccination in the face of a developing immune system has also been explored in pigs (Sus scrofa) as a larger vertebrate model. In two runs tools and procedures have been established (ELISPOT, cell cytometry) and exploited to study the immunogenicity profile of a prototype subunit in adjuvant vaccine in neonate (baby) versus adolescent animals. Although only a weak and transient immune response was obtained by this subunit vaccination, neonate pigs over adolescent animals presented with a differential IFNg response.
Further, mini pigs have been exploited in WP5 as a relatively low susceptibility model that displays pathological characteristics of latent TB infection. To this end, groups of N=6 M.tb H37Rv infected animals and without clinical signs of disease were submitted to 2 times homologous post-exposure vaccination with adjuvant formulated subunit vaccine candidates based on Rv1753 and Rv3616, or with proprietal RUTI as a positive or saline as a negative control stimulus. Although bacterial load in the experimental treatment groups was not different, the mean gross lesion volume was significantly reduced in respiratory lymph nodes. Analysis of extensive data including advanced CT images is ongoing and continues beyond the NEWTBVAC life time.
With relevance to the clinical condition of mycobacterial pre-exposure, vaccine candidates have been tested in a modified Cornell model of persistent M.tb infection, installed by low dose aerosol challenge of 20-50 CFU of strain H37Rv. A total of 11 vaccine candidates and a specific (empty plasmid vector) control have been analysed in this model as post-exposure vaccines using two runs with groups N=12 animals each and using saline, standard BCG and the H56 subunit vaccine as internal controls. Next to H56, BCG and the empty plasmid vector control, another five of the candidates demonstrated efficacy by reduction of bacterial loads in lung (4 out of 5) or spleen (2 out of 5). Amongst these candidates were those that had demonstrated efficacy in the guinea pig model for prophylactic vaccination (and/or good safety in the SCID mouse model) previously: zmp1 deficient BCG, ureC deficient and listeriolysin expressing BCG, and sigE deficient M.tb.
Finally, beyond the budget plan, NEWTBVAC has created and carried through the opportunity to cross-connect within the so-called ‘one health’ concept between clinical and veterinary TB vaccine development. In particular, the promising zmp1 deficient BCG vaccine candidate has been successfully tested for its immunogenicity profile in the cattle model.

 
WP6
Original general background and goals
The goal of WP6 is to identify, test and prioritize surrogate-endpoints of protection and disease (“correlates”) in human tuberculosis. Better correlates of protection will help to identify protective antigens and demonstrate immunogenicity and potential efficacy of new vaccines. In addition, they will permit optimization of dose, vehicle, adjuvant, and immunization schedules of new candidate vaccines, and will thus help to reduce the protracted time scale of TB vaccination trials. The new correlates and biomarkers studied in WP6 build upon new discovery research in WP6, as well as upon findings from the FP6 TBVAC project, and will be tested and validated in various M. tuberculosis (Mtb) exposed/infected or M. bovis BCG vaccinated individuals. These will include TB patients with active disease and after cure, latently infected individuals, and BCG vaccinated children and neonates. Additionally, an important activity will be the development of new assays based on newly discovered biomarkers, and to compare the performance of these new assays to that of existing assays, allowing prioritisation of assays for future use. Assays developed or further refined in WP6 will be applied to monitor vaccine-induced immunity and immunological memory in Phase I trials in NEWTBVAC in the EU and Africa. Ultimately, correlates and markers identified in NEWTBVAC WP6 will need to be tested and validated against clinical endpoints. This can only be achieved in large scale field studies in endemic areas, using appropriate biomarker assays that are applicable for the specific vaccine trials.
Discovery of new TB biomarkers is the second major goal of WP6. A solid biomarker discovery-pipeline is needed to obtain optimal sets of biomarkers and biomarker signatures. Recent studies show that no single biomarker is able to provide a sufficient correlate of protection, including IFN. It is becoming clear that multiparameter signatures will be needed to capture the complexity and dynamics of the human response to Mtb infection in relation to protection and protective immunity. This will be achieved through a variety of innovative discovery approaches, including global genomic, transcriptomic and metabolomic strategies, focusing on both the host and the pathogen response during the different phases of infection. As soon as this work yields new biomarkers, these will be evaluated as described under the first goal. As in the past few years, these activities will be coordinated proactively by the WP6 leadership and by joining forces with other initiatives outside NEWTBVAC (including WHO, STOP-TB, FP7 ADITEC, FP7 TRANSVAC, Aeras, BMGF-GCGH programmes and other alliances). In WP6, we will be able to include unique cohorts of healthy volunteers that have participated in several clinical vaccine trials, including trials testing the safety and immunogenicity of leading new vaccine candidates (notably: MVA-Ag85A; Ag85-ESAT6 (H1) + IC31; rBCG::Hly) as well as BCG (UNIPA, UOXF, ULB, LSHTM, UUlm, MPIIB, LUMC).

Original WP6 strategy
Building on the consensus strategy that the TB Biomarker group has formulated over the years in TBVAC and NEWTBVAC, we follow a two-track strategy:
Track 1: to move the most promising findings and assays from TBVAC, NEWTBVAC and related projects activities forward into assay development and validation; and
Track 2: to complement this by the discovery of new biomarkers to fill the pipeline.
Important innovative elements in this project will be a focus on assays that measure the functional capacity of human cells in the innate and adaptive immune response to Mtb, including Mtb killing assays, poly-functional and memory T-cell subsets and innate immune responses (macrophage/dendritic cell responses, human TB granuloma models, and other relevant models).
Activities proposed will be integrated with other international activities in which many NEWTBVAC and TBVI researchers participate. Future biomarker discovery and validation will need studies in larger cohorts at relevant TB endemic sites, both in Africa and Asia.



Specific goals
1) To identify novel biomarkers or biomarker signatures that are associated with protection against TB disease, and/or markers associated with (early) disease activity and lack of protection/ susceptibility.
2) To prioritize and validate already identified biomarkers from TBVAC1 WP4 “correlates of protection”, and new markers from “1”, in small cohorts (at least 20 individuals per group, including TB patients, BCG vaccinees and healthy controls).
3) To move newly identified (from 1) or promising (from 2) biomarkers pro-actively into assays that are simple, robust and reproducible and can preferably be used in TB endemic settings.
4) To harmonize, standardize, validate, and apply assays for immunological monitoring of clinical TB-vaccine trials, building on successful strategies developed in TBVAC1 and NEWTBVAC.
5) To advise on the selection of tailor made sets of biomarkers for clinical trial monitoring based on vaccine type, antigen expression and trial setting. “Decision making matrices” for prioritization of tailor made biomarker sets will be developed for different vaccine or other interventional studies, as well as observational (longitudinal) studies.


Track 1
WP6.1a Evaluation and validation of newly discovered biomarkers
The aim of WP6.1a is to evaluate and validate newly discovered TB biomarkers as biomarkers or biomarker signatures for monitoring protective and vaccine induced immunity against TB. We aim:
1. To determine at the single cell level cytokine production, cellular phenotypes and function by multi-parameter flow cytometry and ICS in the specific response to Mtb.
2. To analyse patterns of Mtb antigen induced multiplex cytokine production.
3. To assess these responses in well defined cohorts of Mtb infected or exposed individuals, BCG vaccinated adolescents, BCG vaccinated infants and healthy naive individuals, in order to correlate them with protective immunity.
4. To adopt and evaluate Mtb killing assays (from STOP-TB) as a novel correlate of protective immunity.

Major progress M50 highlights
- general:
- WP6 organized a flow and Luminex workshop in June 2010 at UOXF (exchange of information on multiplex cytokine measurements)
- WP6 organized an ICS-flow WS -> WP6 consensus long term FACS ICS assay
- WP6 organized a second ICS-flow WS workshop in June 2012 at UOXF, together with TBTEA. Also Mtb killing assays were discussed, focusing on the first promising results at UOXF.
- WP6 organized a third ICS-flow WS workshop in June 2013 at LUMC. Also Mtb killing assays were discussed.
- UUlm:
- Mtb infection induces long-lasting LAM-specific T-cells, which produce IFN- and exert antimicrobial activity against intracellular Mtb. The frequency of LAM-responsive T-cells is higher in susceptible than in protected individuals.
- The frequency of LAM-specific polycytotoxic T-cells is higher in protected than in susceptible individuals
- LUMC/UNIPA:
- multifunctional T cells are associated with disease, not protection
- LUMC:
- HLA-E restricted CD8 T cells identified with specificity for Mtb;
- distinct CD4 and CD8 Tregs identified against BCG which inhibit Th1 and thus may limit TB vaccine efficacy; a new CD8+CD39+ Treg subsets identified
- cohort building in progress
- UOXF:
- flow Ab panels for T cell subsets to assess differences in memory, effector and regulatory T-cell responses and mucosal homing markers in vaccinees established
- phenotyping and mycobacterial inhibition assays ongoing
- strong correlations between early innate immune and effector T-cell responses
- ULB:
- Long term FACS ICS assay on freshly isolated PBMC for CD4+ and CD8+ multifunctional T cell responses to PPD, HBHA, and ESAT-6 was performed on a cohort of 81 adults and 87 children
- For adult patients, for each antigen, 3 major subgroups of CD4+ cytokine-producing cells: IFNγ+ TNFa+/ IFNγ+/ TNFa+. Analysis of LTBI divided in 2 groups according to the QFT-TB-Gold result reveals that in response to HBHA, the profile of QFT+ LTBI resembles that of active TB cases in contrast to QFT- LTBI. Only low level of cytokines among CD8+ cells. Data confirmed on small cohort in short-term assay.
- For the children: TNFa/IFNy double+, TNFa single, IFNy single and IL17 single positive CD4 + subpopulations may help discriminate between active and latent TB and among LTBI subjects the profile of those responding to RD1/11 IGRA resembles that of active cases whereas RD1/11 negative LTBI subjects have a more distinct functional profile.
- For adult patients: HBHA- and ESAT-6 cytokine / chemokine measured by multiplex assays puts forward a number of promising cytokines for discrimination between active TB/ latent TB/ controls: GMCSF, IFNg, TNF-a, IL-13, IL-10, IP-10 and IL17– Validation cohort under analysis
- LSHTM:
- Multiplex assays reveal a broad and BCG vaccine- induced biomarker response following PPD and HBHA stimulation in 7 day whole blood assays
- fewer polyfunctional T-cell responses (IFNγ/TNF/IL2/GMCSF) and more single cytokine positive responders in day-5 compared to day-1 ICS stimulation assays
- Detection of multifunctional, PPD-specific Th-1 CD4+ T-cells in BCG vaccinated but not unvaccinated infants at 3 and 12 month time points.
- Detection of IL-17+ CD4+ T-cells in BCG vaccinated infants
- 13 cytokine/chemokine responses detected by Luminex in overnight ICS supernatants in BCG-vaccinated but not unvaccinated infants.
- Performance of whole blood mycobacteria growth inhibition assays on samples from BCG vaccinated and unvaccinated infants.
-
- UNIPA:
- unexpectedly complex T-cell phenotype patterns in the response to Mtb
- plasma based multiplex measurement of cyto/chemokines, cytotoxic mediators (granulysin, perforin, granzymes)
- IRB:
- flow Ab panel developed and optimized for the enumeration of chemokine receptor expression on T cells from the naïve (TN), stem cell memory (TSCM), central memory (TCM), and effector memory (TEM) cell subsets.


Major Progress M50 description
Collaborative WP6 studies on multifunctional T cells in TB showed that these cells were highly prevalent in TB patients compared to LTBI, contrary to expectations (LUMC and UNIPA). UNIPA found triple (IFNγ/TNF/IL-2) expressing CD4 T-cells in 85–90%TB patients but only in 10–15% of LTBI subjects. UNIPA and LUMC aim to confirm these data in larger cohorts, including patients after curative therapy. Moreover, LUMC found that BCG efficiently induces multiple different Treg populations in humans following vaccination, which down-regulate Th1 responses and thus may limit BCG-vaccine efficacy. CD4 and CD8 Treg populations and their functional activities are being analysed further in extensive multi-parameter flow panels and functional assays. Multifunctional T cell responses against a series of Mtb antigens (secreted, DosR, rpf, in vivo expressed) were observed using these same panels, in collaboration with WP1, revealing prominent CD4 and CD8 mono-, dual and triple (IFNγ/TNF/IL-2) responses. Moreover, new HLA-E restricted CD8 T cells with specificity for Mtb peptides could be identified. Also at LUMC, a follow-up cohort of TST convertors is being built, aiming to assess secondary TB cases, which will be an opportunity to determine biomarker profiles associated with susceptibility vs. protection. Finally, LUMC started to collect and analyse published biomarker data obtained by non-biased methods using pathway analysis and found novel biomarkers based on these analyses. Data have been presented and discussed during the annual meeting in Les Diablerets in Febr 2013 and 2014, as well as during the WP6 meeting in Leiden, June 2013 and the TB-BM meeting in DC (Sept 2013).
UNIPA assessed the phenotype of cytokine-producing cells, using both a short-term and a long-term in vitro stimulation assay, in blood samples from LTBIs and TB patients before and after completion of therapy. Results show unexpectedly complex phenotype patterns of cells which produce different combinations of cytokines, which actually seem to exclude a correlation between phenotype and function. In detail, the phenotype of cytokine-producing cells is extremely heterogeneous and changes depending on the time of in vitro stimulation (short-term versus long-term), the type of antigen (protein versus peptide) and the disease status (LTBI individuals versus patients with active disease before and after therapy). Using plasmas, new data indicate that the low levels of Granzyme A in antigen-stimulated plasma correlate with active disease, while high levels are found in TB patients after completion of therapy and in LTBI individuals. Moreover, a number of cytokines and chemokines are differently upregulated in the plasmas of patients with active TB disease, and these include IFNg, IP-10, IL-1R antagonist, IL-2 and IL-17, suggesting that these cytokines may have a potential impact as candidate biomarkers. In addition, using peptide/HLA-E tetramers Mtb peptide specific, HLA-E restricted CD8 T cells were detected in the circulation of patients and LTBI.
At UOXF, following optimisation, flow cytometry analysis has now completed for the assessment of differences in memory, effector and regulatory T cell responses and also mucosal homing markers in volunteers vaccinated either via the intradermal or intramuscular routes. Two panels have been optimized: panel one: CD4, CD19, CD14, CD45RO, CD27, CCR7, CCR4, CD3, CD8, IL-2, IFN-γ, TNF and IL-17 and panel two: CD4, CD19, CD14, CD45RO, CD25, CD39, CCR7, CCR6, CXCR3, CD3, CD8, IFN-γ, IL-17. Samples were processed beginning of 2012. In addition, qPCR and microarray analysis of samples collected following intradermal or intramuscular vaccination with MVA85A revealed strong correlations between early innate immune responses and effector T-cell responses (IFN-γ ELISPOT), which were confirmed by flow cytometry. Innate immune responses have been measured using DNA microarray analysis. This data was used to identify an RT-PCR panel that could predict immune response to vaccination with MVA85A. This RT-PCR panel was published in Plos One (Matsumiya et al 2013). In addition, using a whole blood assay utilising a MGIT machine for assessing mycobacterial growth it was observed that mycobacterial growth is inhibited in volunteer blood collected 8 weeks following vaccination with BCG. The performance of this assay in frozen and freshly isolated PBMC has been compared to that of whole blood.
At ULB, the optimized long-term FACS ICS multi-parameter protocol (collaborative WP6 effort) was used on freshly isolated PBMC to evaluate multifunctional T cell responses to HBHA, ESAT-6 and PPD. 81 adult patients have been prospectively included comprising 17 patients with active TB, 28 with latent TB (defined on the basis of the tuberculin skin tests results; 10 with a positive QFT-TB-Gold test) and 12 control non infected subjects. The other 24 subjects were unconfirmed aTB suspicions, aTB patients co-infected with HIV, or subjects with unclassifiable TST. In addition, 87 children were included comprising 15 with active TB, 19 with latent TB, 29 healthy exposed children and 24 children were excluded (status unknown/unclear). For adult patients, three major subgroups of cytokine-producing CD4+ T cells (among 7 possible subgroups representing all combinations) were detected both for aTB and for LTBIs subjects, both in response to HBHA and to ESAT-6 and also PPD: double positive IFNγ+ TNFa+, single positive IFNγ+ and single positive TNFa+. This means that the analysis of multifunctional CD4+ T cells in these cohorts of patients did not enhance the discrimination between active and latent TB obtained by bulk analysis of IFNγ secretion. Division of LTBI according to their response to QFT-TB-Gold test was investigated, and revealed that the heterogeneity of the responses to latency antigen HBHA among LTBI is attributed to differences in response to early antigens (QFT). Indeed, the cytokine profile observed for QFT- subjects is characterized by a low frequency of TNF-a-only producing cells and a high frequency of IFN-g-only producing cells, as opposed to QFT+ subjects. Interestingly, the profile of patients with active TB resembles the latter, supporting the notion of a spectrum of how Mtb infection may present. A single positive TNF-a population appears to serve as a biomarker of disease in response to PPD, while in response to HBHA it also seemed to identify those LTBI with an increased risk to develop an ATB.
For CD8+ T cells, ULB only detected low levels of cytokine producing cells with this long-term assay. In a small cohort, we performed a short-term stimulation but this did not increase significantly the frequencies of antigen-specific CD8 cells. Cytokine and chemokine secretions were analyzed in parallel in the 5 days cell culture supernatants, on a bigger cohort including 10 controls, 20 subjects with LTBI and 14 patients with TB. Results confirmed the discriminative power between patients with active /latent TB of HBHA-induced IFNγ secretions. Among the 20 cytokine/ chemokine analyzed, GMCSF, IFN-g, TNF-a, IL-13, IL-10, IP-10 and IL17were identified as being promising for the discrimination between aTB LTBI and uninfected cases. . In conclusion, we found several biomarkers of infection and disease, which are currently evaluated in a second cohort to check for robustness. This information may be useful for the development of a better diagnostic tool, able to (1) identify subjects with LTBI e.g. before immunosuppressive treatment, (2) identify patients with ATB for non-evident clinical cases, or (3) identify surrogate markers of protection during vaccine trials. .
For the children cohorts, the statistical analysis showed that the proportion of TNF-a single CD4+ T cells after stimulation with ESAT-6 is better represented in ATB cases as compared to LTBI (and controls), whereas the IFN-g single+ subpopulation after stimulation with PPD, ESAT-6 or nHBHA is represented better in LTBI subjects as opposed to ATB. The same is true for the IL-17 single + subpopulation after stimulation with nHBHA. When dividing LTBI subjects according to their IGRA response to RD1 antigens, the majority of RD1 positive LT subjects had a profile similar to that of ATB cases, whereas RD1 negative LT cases had a significantly lower proportion of the IFN-g/TNF-a and IFN-g single + subpopulations and a tendency toward a higher proportion of the IL-17 single positive subpopulation (IL-17 significant only in absolute values).Interestingly, part of the exposed subjects with negative TST (“negative controls”) had a positive RD1 IGRA, and this subgroup had a cytokine profile resembling RD1 negative LTBI subjects, whereas RD1 negative control subjects had a much more varied profile, reinforcing the notion of spectrum of disease after infection with Mtb.
Another part of the study, was to implement a technique called FASCIA, for Flow cytometric Assay for Specific Cell-mediated immune responses in Activated whole blood, in response to HBHA, ESAT-6 and PPD. This method is based on the electronic gating strategy in flow cytometry analysis, defining the activated cells (blasts) as cells having a greater forward scatter (FSC) and side scatter (SSC) than resting lymphocytes. For 10 controls and 14 LTBI, we compared FASCIA to the optimized whole blood-IGRA (Deliverable D6.1b.2 reported at M36). We found that FASCIA in response to PPD and HBHA was able to discriminate between LTBI subjects and controls, and that results of FASCIA and WB-IGRA were highly correlated for PPD, HBHA and ESAT-6. Likewise, we measured proliferation of stimulated PBMC in the 5 day assay, and found that proliferation of CD8 cells in response to ESAT-6 is a marker of disease, present in most TB cases but not in subjects with LTBI.
At LSHTM recruitment of BCG vaccinated and unvaccinated infants for WP6 biomarker studies, has taken place. Using WP6 SOPs, fewer polyfunctional T-cell responses (IFNγ/TNF/IL-2/GM-CSF) and more single cytokine positive responders were observed in a day-5 compared to overnight ICS stimulation assays. Also NK cells were shown to participate in the BCG induced response. Overnight ICS on samples from unvaccinated infants or vaccinated infants revealed PPD-specific polyfunctional CD4+ T-cells in vaccinated infants that co-express IFN-γ, TNF-α and IL-2 but are negative for IL-17 and IL-10 These cells are not detectable in unvaccinated infants. When co-stimulatory antibodies are included, some PPD-specific IL-17 expression is observed. Analyses have revealed multi-functional, PPD-specific CD4+ T-cells (IFNg+TNFa+IL-2+) in vaccinated but not unvaccinated infants at both 3 and 12 month time points. Analysis of HBHA-specific cytokine responses in 7 day, whole blood assay supernatants has also shown a vaccine-associated response that is absent in unvaccinated infants. The strongest HBHA-induced responses were chemokines whereas the strongest PPD-induced responses were Th-1 cytokines, MIP1a and GM-CSF. GM-CSF, FGF-2, sCD40L, IL-12p40, IL-1a and sIL-2Ra comprised 6 of the 7 most strongly associated biomarkers for each antigen. We have collected supernatants from overnight ICS assays and a subset of these samples (n=20) have been tested by 42-analyte multiplex assay for secreted cytokines/chemokines. In addition to the ICS differences detected (see above), 13 secreted analytes were also observed to be differentially detectable between vaccinated and unvaccinated infants. Mycobacterial growth inhibition assays have been set up for evaluation in on BCG vaccinated adults and infants using whole blood. Although there was a consistent level of mycobacterial growth in MGIAs carried out on samples from unvaccinated infants, there was a more mixed response in vaccinated infants with some individual samples displaying a degree of growth inhibition in comparison to the unvaccinated group. The overall difference between groups was not significant however, indicating that the assay may require further optimisation.
Studies performed at UULM identified lipoarabinomannan-specific IFN- release in protected individuals (IGRA-positive donors with a history of close contact with tuberculosis patients) and susceptible individuals (tuberculosis patients after completion of therapy). Sorted LAM-specific T-cells recognized Mtb-infected macrophages and killed intracellular bacilli indicating that this subset contributes to protection against tuberculosis. The number of responding donors as well as the average frequency of IFN--positive cells was significantly higher in susceptible than in protected donors. More detailed characterization of LAM-specific T-cells revealed a subset of T-cells co-expressing perforin, granzyme A and granulysin. In protected individuals (n=21) the frequency of “polycytotoxic” T cells (= triple positive cells) was significantly higher (64%) than in susceptible individuals (30%). In summary, the quality (“polycytotoxicity”) of LAM-specific T cells appears to correlate with a favourable outcome of infection after exposure to Mtb.
Finally, previous studies from the IRB showed that in LTBI donors, Mtb-reactive T cells are mainly contained in a subset of memory Th1 cells expressing the chemokine receptors CCR6 and CXCR3. A few Mtb-reactive T cells could be detected in a Th17 subset identified by the expression of CCR6 and CCR4. T cells in both subsets recognized antigens that are contained within or in close proximity to three islands containing Esx protein pairs and PE/PPE proteins. To define at the clonal level the lineage relationship between the Mtb-responding Th1 and Th17 cells we performed TCR Vbeta clonotypic analysis by next generation sequencing on genomic DNA. The analysis performed on two donors showed that on average 240-400 sequences were obtained in the memory subsets, with several present at low copy numbers and a few present at high copy numbers, consistent with overrepresentation of the latter either in the memory population. Several sequences were found only in the Mtb-reactive Th1 or Th17 cells. However, a few were shared between the two subsets. Thus, at the population level Th17 cells appear to have a different clonotypic makeup. The presence of Th1 and Th17 cells expressing the same CDR3 TCRB sequence is consistent with an intraclonal diversification and T cell plasticity.
Finally, several groups in WP6 are using multiplex cytokine technology to assess multiple chemokine and cytokines in serum/plasma as well as in stimulated cell supernatants. A separate workshop organized by WP6 was dedicated to exchanging and discussing technical and operational issues, which was considered very helpful to the participants.


WP6.1b Development of new assays based on validated new TB biomarkers
The aim of WP6.1b is to develop new TB biomarker assays, based on newly discovered and validated TB biomarkers from (NEW)TBVAC. Assays developed should preferably be suitable for use in large-scale monitoring studies, such as in TB endemic areas. Assays moved forward for development are:
1. Whole blood assays.
2. Gene-expression analysis.
3. Dipstick assays detecting specific TB biomarkers.

Major progress M50 highlights
- ULB:
- Validation of the robustness of the analysis of the local production of IFNγ in biological fluids in 24 hr assays in response to HBHA and ESAT-6
- Evaluation and validation of 24hrs HBHA-IGRAs on PBMC and on whole blood
- IPAS:
- Identification of MTB specific genes in infected human cells.
- Analysis of miRNA expression upon MTB infection in human primary cells and identification of miRNA whose expression is controlled by genetic factors (Siddle et al. Genome Research 2014).
- Identification of genes whose expression is modulated by extracellular ATP in MTB infected macrophages (N. Dubois-Colas, L. Petit-Jentreau et al. The Journal of Infectious Disease, in press)
- Metallproteinases MMP1-3-8 and 10, and their inhibitors TIMP-1 and TIMP-2 strongly upregulated in infected macrophages (mRNA, protein).
- MMPs and TIMP-1/2 are secreted in TB patients
- Circulating endothelial cells are mobilized in TB patients
- MPIIB:
- multiplex RT-PCR on Rab33A/LTF/CD64 replaced by single reactions
- cohorts screened using the assay
- LSHTM:
- six biomarkers associated with BCG vaccination associated with IFNγ after stimulation with PPD / HBHA
- supernatant IFNγ by ELISA in WBA was compared with ELISpot and ICS IFNγ assays for reproducibility and transferability between labs
- Mycobacterial growth inhibition assay established and comparisons with ICS and ELISA assays performed


Major Progress M50 description
At ULB, the previously developed protocol to analyse the local production of IFNγ in response to HBHA, ESAT-6 and PPD was standardized and evaluated on 57 biological samples from patients suspected of active extra pulmonary TB. To ensure high sensitivity and specificity, robustness of the method and of the interpretation of the data, a major limitation of the test was the number of lymphocytes required (>60%) in the sample. Therefore, only 26/57 samples reach the end of the process and were analysed. The test allowed the identification of all TB cases with high specificity, whatever the type of fluids, within 24 hours when combining the results obtained in response to the three antigens. This short term assay was therefore of great help for clinicians as definitive TB diagnosis by clinical/microbiology proofs were always confirmed weeks after.. The technique was recently adapted to allow analysis of the IFNγ synthesis even when the lymphocyte number is relatively low. In addition, multifunctional T cells (IFN-γ, TNF-α, IL-2, IL-17) were analysed and their possible contribution in the robustness of this new assay evaluated on 16 biological samples. Five samples had very low lymphocyte numbers and 8 were active TB samples. The results confirmed that the frequency of IFN-γ producing cells after stimulation with Mtb antigens is the most reliable read-out for aTB diagnosis and that it is feasible even when lymphocytes are not the major population of the biological fluids. Preliminary data on multifunctional CD4+ T cells indicate 2 different types of profile among the active TB cases which seem to correlate with the degree of dissemination of the disease. .
ULB also developed and validated robust 24 hrs HBHA-IGRA both on PBMC and on whole blood. Correlation of the results obtained with the classical 96 hrs assay and with the 24 hrs assay on PBMC was excellent and the relative discrimination between patients with active TB or latent TB was maintained. The 24hrs HBHA-IGRA assayon PBMC was validated on 137 patients. It was shown that for both CD4+ and CD8+ T lymphocytes, effector memory (TEM) and central memory (TCM) T cells are the predominant IFNγ producers in response to HBHA whilst the naïve and effector T cells have smaller and variable contributions. However, in the 24 hrs assay, TEM represent 60% of the CD4+ IFNγ producers whilst in the 96 hrs assay, the role of TCM is larger. A manuscript describing the 24hrs IGRA on PBMC has been published in CVI. The HBHA-IGRA on whole blood was developed and evaluated on 62 immuno-competent subjects and 115 patients before or under biotherapy. This short term whole blood assay allowed the detection of LTBI and active TB with high sensitivity and specificity but did not allowed anymore to discriminate latent form active infection. Different hypothesis are currently evaluated to understand these results and find how to retrieve and ameliorate this discrimination..
At IPAS, in order to identify new biomarker candidates, global expression profile has been used to identify (i) MTB specific induced genes, (ii) genes induced by extracellular ATP and (iii) miRNA expression in response to MTB infection. Ad (i) In close collaboration with Yoav Gilad (University of Chicago), we evaluated the specificity of the MTB response. Briefly human primary cells have been infected with different strains of MTB, attenuated strains including BCG, M. smegmatis, Yersinia pseudotuberculosis, Staphylococcus epidermidis and Salmonella typhimurium. Whole genome expression profile has been performed using RNA seq. Analysis is still ongoing. Ad (ii) During TB, the center of the granuloma exhibits necrosis resulting from the dying of infected macrophages. The release of the intracellular pool of nucleotides into the surrounding medium may modulate the response of newly infected macrophages. We showed that extracellular ATP modulates the expression of 272 genes in human macrophages infected with MTB, and that it induces their alternative activation (N. Dubois-Colas, L. Petit-Jentreau et al., The Journal of Infectious Diseases in press). These genes represent good candidate biomarkers as only TB patients have important lung damage and may discriminate TB patients from latently-infected individuals. This hypothesis will be evaluated in TB patients in the future. (iii) We assessed changes in miRNA expression upon MTB infection and mapped expression quantitative trait loci (eQTL) in human cells from a panel of healthy individuals. We found that 155 miRNAs were differentially expressed upon infection (K. Siddle et al., Genome Research 2014). We find that the expression of 3% of miRNAs is controlled by proximate genetic factors, which are enriched in a promoter-specific histone modification associated with active transcription and we show that infection coincides with a marked remodeling of the genome-wide relationships between miRNA and mRNA expression levels. Unfortunately, D6.2.5 could not be completed since the available limited budget did not allow for this.
At MPIIB, the development of a multiplex RT-PCR reaction for 4 genes could not be get to run reliably and consistently such that MPIIB decided to return to single reactions for which primer and probes had been optimized earlier. The evaluation of these PCRs has been run on four different cohorts of TB patients and healthy controls. The expression levels of the targeted genes confirm findings from previous microarray data, with increased FCGR1A levels and decreased GZMA levels in active TB patients. Also, significant expression variations between individuals were observed. Although absolute expression measurements of these three genes show clear differences between groups of patients and controls (either latently infected or uninfected), they do not provide a 100% accurate prediction for each individual to be classified as either active TB diseased or healthy. Results were published and discussed in a series of papers (PNAS 109(20):7853-8; Int J Tuberc Lung Dis 16(9):1140-1148; and: Ann N Y Acad Sci. 1283:22-29).At LSHTM, previous multiplex analysis showed that up to 39 biomarkers are necessary to fully capture the difference in response between the UK and Malawi in BCG vaccinated infants. Six biomarkers (GM-CSF, FGF-2, sCD40L, IL-12p40, IL-1α and sIL-2Rα) were strongly associated with IFN-γ after stimulation with both PPD and HBHA in UK infants suggesting a BCG vaccination associated biomarker profile. Further PPD stimulated and control supernatants from Malawian and Gambian BCG vaccinated infants bled 3 months after BCG vaccination were tested by Luminex for comparison with UK data. Studies on Malawian infants have identified 9 cytokine/chemokines (not including IFNgamma) that are detectable at higher magnitudes in Malawian infants vaccinated “late” (i.e. between 3 and 11 weeks of age) as compared to those vaccinated in the first week of life. The detection of supernatant IFN-γ by ELISA following diluted whole blood assay has been compared with ELISpot and ICS IFN-γ detection assays for reproducibility and transferability between labs (in conjunction with TRANSVAC). The development of these SOPs has complemented the use of assays (particularly ICS) in WP6. Although IL-17 and IL-10 are both detectable following PPD stimulation of whole blood and Luminex analysis, data suggested that, unlike IFN-γ, TNF-α and IL-2, neither IL-17 or IL-10 are detectable in response to PPD alone by ICS. The addition of co-stimulatory antibodies however, allowed the detection of an IL-17 response. A mycobacterial growth inhibition assay protocol developed within WP6 by UOXF has been implemented at LSHTM. BCG stocks have been established and assays performed to compare MGIA responses with other biomarker readouts in BCG vaccinated adults and infants.

Several partners in WP6 (LSHTM, LUMC, UOXF) are also involved with efforts to standardize protocol and data analysis steps in both ELISpot and ICS flow cytometry techniques. LSHTM has taken part in multi-centre proficiency panels to test the effect of alternative freezing media on ELISpot responses and to help improve harmonisation guidelines for the use of the assay across different sites, through the Cancer Immunotherapy Consortium. LUMC is a member of Minimal Information About T-cell Assays (MIATA; designed to provide a framework for reporting and publication data from T-cell immunoassays) project, and LUMC and LSHTM have contributed to MIATA (Immunity, 2012).


Track 2
WP6.2 Discovery of new biomarker signatures
The aim of WP6.2 is to identify novel biomarker signatures by new technologies, including unbiased, global genetic and T cell memory based approaches. We aim:
1. To perform microarray analyses on samples from different clinical trial settings, including clinical trials with a new rBCG::Hly vaccine (VPM1002), to identify specific expression profiles.
2. To identify new T cell subsets that give rise to Mtb specific memory T cell subsets.
3. To perform microarray analyses on Mtb infected human macrophages to identify specific innate immune related expression profiles correlating with Mtb infection.
4. To move promising biomarker signatures resulting from 1-3 towards monitoring of patients and controls in WP6.1a to evaluate and validate the most promising biomarkers.

Major progress M50 highlights
- LUMC:
- TB biomarker profiles identified by MLPA discriminate TB from LTBI by multi-component signatures
- global transcriptomic responses in untreated, and cured TB patients compared to LTBI (PLoS ONE 2012) show clear disease associated gene expression patterns in line with data by Berry et al (Nature 2010) and MPIIB (PNAS 2012)
- IPAS:
- metallproteinases MMP1-3-8 and 10, and their inhibitors TIMP-1 and TIMP-2 strongly upregulated in infected macrophages (mRNA, protein).
- MMPs and TIMP-1/2 are secreted in TB patients
- Circulating endothelial cells are mobilized in TB patients
- MPIIB:
- RNA from participants at different time-points following vaccination with parental BCG or recombinant strain VMP1002 isolated and hybridized
- Bioinformatics analysis identified only minor differences between the groups of parental and recombinant BCG vaccine strain vaccinees.
- IRB:
- T cell library assays showed that Mtb-specific CD4+ T cells are largely contained in a CXCR3+CCR6+ memory subset and highly focused on three broadly immunodominant antigenic islands, all related to bacterial secretion systems



WP 7
The key objective of WP7 is to establish a management processes to progress new TB vaccine candidates along preclinical to clinical development pathway.

Product Development Team (PDT) and Clinical Development Team (CDT) structures were established. Remit for PDT is to facilitate the focused product development programmes and advise on development of preclinical product development plans; production issues, cGMP requirements, product specifications; advise on project management where appropriate; and facilitate the access to informal pathways for expert advice on specific issues in production, quality or regulatory issues. Remit for CDT is to facilitate the focused product development programmes and advise on development of clinical product development plans; current regulatory guidance and requirements where appropriate; and guidance on Phase I and further clinical studies.

The annual meeting of NEWTBVAC in February 2010 highlighted the need for a landscape analysis of vaccine candidates within the consortium, with emphasis on their readiness for entering into preclinical development. PDT performed the landscape analysis of nine potential vaccine candidates and report was submitted to TBVI and NEWTBVAC. Process was in place for ongoing analysis as landscape changes. Out of the nine candidates, six are recombinant live attenuated mycobacteria including four recombinant BCG ‘s, and three are subunit vaccine candidates including two based on recombinant proteins with novel adjuvants. From the analysis, preclinical development occurs in a series of steps. Initially the developer constructs various prototypes and evaluates them in at least one animal species. Then, based on these data in animal models, further development of potential prototypes are re-evaluated in several animal species for robust proof-of-concept. Three key criteria are considered for entering into preclinical development stage and they are (1) the candidates that could enter development are selected; (2) the feasibility of producing the vaccine candidates is demonstrated, and (2) the proof-of-concept is documented in animal models.
As part of the agreed NEWTBVAC strategy for the second half of the project (2012 to 2014), it was agreed that a call for new vaccine candidates that were potentially in a position to benefit from focused preclinical development through to preparation of Phase I be made to the consortium. This support would include additional funding (100K Euro) to aid development, manufacture and testing; and the formation of product specific PDTs and CDTs to focus and assist the preclinical development path. A TBVI Candidate Selection Panel (comprised of 8 members from the consortium) was set up to review the applications and provide recommendations to the NEWTBVAC SC. The method of Prioritise/ Rank/ Score of each application was discussed among the Panel members in terms of passing the objectives of Gate 1 criteria, together with novelty and finance considerations. Applications were submitted using a standard application form and evaluated using a standard form with “Gateway 1 Criteria for GUIDANCE”. There are four successful applications and they are rBCG ureC::Hly Hygs (VPM), ChAd85A (UOXF), rHBHA (IPL), and rBCG zmp1 (UZH).

As the project progress, TBVI management team with the assistance of PDT and CDT set up a TB vaccine candidates Portfolio Management process. A live document that presents an inventory of vaccine candidates under discovery and development, together with their brief description has been prepared. For each vaccine candidate a target product profile (TPP), the status of research or development, and forecast for the coming year is updated every 6 months. Both stage gating and priority setting criteria are the tools in this Portfolio Management process.

Within WP7, apart from the four vaccine candidates also received additional support to aid the product development, PDT/ CDT also monitored the progress of other vaccine candidates including MTBVAC (UNIZAR), inactivated MTBVAC+ (UNIZAR), BCG::RD1-ESAT6-L28/29 (IPAS), M. microti-ATCC35782::ESX-1(IPAS), M. tuberculosisΔsigEΔfad26 (UNIPD), rBCGAc2SGL (CNRS), M.tuberculosisΔRV1503c (CNRS) and Ac2SGL (CNRS). The most advance vaccine candidates with their progress are described in the following sections.

MTBVAC ---- Preclinical studies including long-term protection and immunogenicity studies with MTBVAC were completed by UNIZAR and results were promising. The first-in-human Phase 1 clinical trial of MTBVAC in Lausanne, Switzerland with Principle Investigator Prof. François Spertini was started in late Jan 2013; the first volunteer of Cohort 1 was vaccinated. Vaccination of the first cohort was completed first week of April. Based on the safety data collected up to this date, the DSMB gave the green light to proceed to Cohort 2 vaccination. The last two volunteers of Cohort two were vaccinated on 12 Jun 2013 and following per-protocol DSMB meeting, approval was granted to move to Cohort 3 vaccination started as scheduled, on 24 July 2013 and last volunteer was vaccinated on 6th of Nov 2013. The recruitment vaccination phase of Cohort 3 lasted longer as it coincided with summer holidays and initiation of school year (Aug and Sept). Safety follow-up of cohort 1 ended in Nov 2013 and of Cohort 2 end of Jan 2014. To date MTBVAC continues to show excellent safety profile without signs of adverse events, as compared to BCG SSI. The Phase I trial is expected to complete in Jun 2014. Final safety report will be available in July 2014 and final report of the unblinded immunogenicity data will be available in Sept 2014.

In February 2013, a meeting with TBVI and CDT took place at Biofabri and at Sergas, Santiago, Spain to review progress of Phase 1 and development of Clinical Development Plan and review establishment of large-scale industrial production. Currently a Clinical Development Plan for MTBVAC as a neonatal vaccine is being elaborated with the expertise of TBVI Clinical Development Team and Unizar Industrial partner Biofabri. With this aim, a two-day TBVI CDT meeting took place in Madrid in Nov 2013. A substantial progress was made in terms of decision making and identification of clinical trial sites for the further clinical evaluation of MTBVAC in newborns in high-burden TB-endemic countries. A next CDT meeting is planned on 6th May 2014 with aim to finalize the Clinical Development Plan document and preparation of CTA for permission to enter Phase 1b dose-escalation safety, tolerability and immunogenicity of three doses of MTBVAC as compared to BCG SSI in healthy newborns, born to HIV-negative mothers.

Stability studies of MTBVAC clinical lot are ongoing until M48 at BIOFABRI. Swissmedic advised that for advanced Phase IIb and III trials for immunogenicity and protective efficacy in TB endemic countries, it is important to conduct long-term immunogenicity and protective efficacy studies. Such studies are ongoing in UNIZAR with the clinical lot of MTBVAC in different mouse models at UNIZAR until M48 of the NEWTBVAC project. Different antigens found by proteomics in WP3 (Deliverable 3.2.4) will be tested in mouse immunogenicity studies for potential selection for use in Clinical Trials.

rBCG ureC::Hly Hygs --- The task to remove the hygromycin B resistance marker cassette of the clinical vaccine candidate BCG ureC::hly HmR (VPM1002) to avoid regulatory concerns was completed. Excision of such a cassette was a complex and time-consuming two steps process. Standard protocols failed and therefore had to be optimized for main parameters, such as antibiotic concentration and temperature. The optimized protocol resulted in successful removal of the hygromycin B resistance cassette from VPM1002 which was confirmed by polymerase chain reaction and Southern blot as shown in the figures below.




Fig. 1. PCR analysis of M. bovis BCG ureC::hly K10C2. pVEP2003 refers to the knockout construct, that was employed to generate M. bovis BCG reC::hly HmR. Fig. 2. Southern blot analysis of M. bovis BCG, M. bovis BCG ureC::hly HmR and M. bovis BCG ureC::hly. Probes for the ureC gene, the hygromycin B resistance (HmR) cassette and the hly gene were used.

ChAd85A --- Additional funding from WP7 was put towards GMP manufacture of a simian adenovirus expressing Antigen 85A in UOXF. This was completed satisfactory. GLP toxicology studies have been completed. Subsequently a Clinical Trial Authorisation has been granted from the MHRA, UK for the conduct of a phase 1 first-in-man clinical trial with this candidate vaccine. This trial was commenced screening and enrolment in July 2013; and is currently ongoing.
HBHA ---- A batch of HBHA had been produced by PX Therapeutics from biomass of BCG, using a scalable manufacturing process (compatible with GMP). This batch of HBHA had been characterized and promoted as reference standard. Analytical methods, such as SDS-PAGE (protein identity and purity), SEC-HPLC (purity), OD measurement (protein content), RP-HPLC (protein content), mass spectrometry analysis (integrity), were set up for characterizing the HBHA. SOPs for quality control assays for drug substance were established. Validation plans for SEC-HPLC (purity), RP-HPLC (HBHA content) and OD measurement (protein content) had been set up. The main outcomes of this study are SOPs and validation plans of QC assays of HBHA.
Due to the change of product development plan, a collaborative partnership of IPL with Aeras has been established for the subsequent GMP production of HBHA and the production of clinical lots. A recombinant BCG strain, over-expressing HBHA (rHBHA) was constructed (patent co-owned by IPL and Aeras), and scalable fermentation and purification procedures have been developed. A first generation of rHBHA (rHBHA1) has been purified and was shown to be immunogenic and significantly increase the protection level in BCG-primed mice when administered in combination with GLA-SE (Glucopyranosyl Lipid Adjuvant-Stable Emulsion produced by IDRI) as adjuvant. Several combinations "adjuvant + HBHA" have been evaluated for their immunogenicity in mice. The adjuvant GLA-SE gave the most promising results. A dose-response study was performed and showed that in naive mice, 3 administrations of 5 µg of rHBHA + GLA-SE induce higher levels of IFNg by HBHA-stimulated splenocytes compared to mice which received the standard HBHA+DDA-MPL combination. In parallel, two doses of HBHA+GLA-SE were administered to BCG-primed mice and were shown to induce lower levels of HBHA-specific IFNg compared to 3 administrations in naive animals. In a preliminary experiment, this prime-boost schedule was able to induce a significant reduction in bacterial load in the lungs of mice challenged with M. tuberculosis as compared to control BCG-vaccinated mice non-boosted with HBHA.
Based on these promising results, a mouse-adapted HBHA-IGRA assay was developed to be used for rHBHA lot release. Briefly, mice are infected with BCG and sacrificed 5-6 weeks later. Isolated splenocytes were either non-stimulated or stimulated with rHBHA or with control antigens (PPD, ConA, native HBHA) for 72 hours. IFNg release was measured in cell culture supernatants by ELISA. Surprisingly, rHBHA1 did not give positive results in this mouse-adapted HBHA-IGRA, in contrast to native HBHA. A second process of purification has therefore been set up where the second chromatography column has been changed. The resulting purified rHBHA is called “second generation” rHBHA or rHBHA2. Preliminary results showed that rHBHA2 was able to induce IFNg production as measured both in mouse-adapted and in human HBHA-IGRA. Immunogenicity and protective efficacy study of rHBHA2 + GLA-SE has been initiated. If rHBHA2 administration in naïve or in BCG-primed mice gives similar or better protection level against M. tuberculosis challenge infection compared to rHBHA1, GMP production of rHBHA2 will be initiated for phase 1 trial in BCG-vaccinated volunteers.
rBCG zmp1 --- Mycobacterial mutants deficient in zmp1 are unable to arrest phagosome maturation. Proof of concept studies in mice have shown that BCG zmp1 mutant is more immunogenic than the parental strain in mice. Most importantly, BCG mutants deficient in zmp1 have been shown to confer an improved protection in the low dose aerosol model of guinea pig infection and this protection was independent of the genetic background of the BCG strain used (BCG Pasteur SmR zmp::aph; BCG Denmark zmp1).
BCG Pasteur zmp was constructed in addition to BCG Pasteur SmR zmp::aph and BCG Denmark Δzmp. For construction of BCG Pasteur Δzmp, the same targeting vector was used as for construction of BCG Denmark. Therefore, the strains are considered to be isogenic with respect to the zmp allele. Gene inactivation was demonstrated by Southern blot analyses. Absence of antibiotic markers in the deletion mutant was proven by phenotypic testing, i.e. by streaking massive amounts of bacteria on plates containing hygromycin and kanamycin, respectively. Expectedly, no growth was observed.
Using the acidotropic dye lysotracker, we demonstrated that BCG inactivation of zmp1 relieves phagosome maturation arrest, which is a hallmark of M. tuberculosis and M. bovis BCG, and that this is independent of the background of the strain (BCG Denmark zmp, BCG Pasteur zmp). Infection of BMDM with BCG Pasteur zmp induced a stronger IL-1β secretion (marker of inflammasome activation) than infection with BCG Denmark zmp. In contrast, infection of BMDC with zmp1 mutant strains as compared to BCG wild-type strains resulted in increased MHC-II antigen presentation (measured by IL-2 release) independent of the strain background. Comparison of transcriptome data with cytokine secretion and Western blot analyses indicate that increase in IL-1β secretion is due to inflammasome activation rather than increased transcription of (pro-) IL-1β gene and increased intracellular pro-IL-1β levels. The BCG zmp1 induced inflammasome activity in turn depends on TLR2/MyD88 signalling pathway, but not on the TLR4/TRIF pathway, as demonstrated by comparison of results obtained with wild-type (C57BL/6) BMDCs and BMDC from corresponding knock-out mice.
In order to develop BCG zmp mutant strain for use in humans as a common portfolio of TBVAC and AERAS, BCG zmp1 was tested for its protective efficacy in cattle (in collaboration with M. Vordermeier, AVLA) and in non-human primates (experiment finalized; assembly of data from post-mortem analyses awaited; in collaboration with F. Verreck, BPRC). In these assays BCG Danish vaccinated and non-vaccinated animals served as a control. Cattle were immunized head to head with BCG Pasteur SmR zmp::aph, BCG Denmark zmp and BCG Danish. Immunization with BCG zmp mutant induced a stronger T cell memory response than immunization with BCG and this response was independent of the strains background. The strength of the T cell memory response is a predictive immune-correlate of protection. Moreover, safety of BCG zmp1 derivatives (BCG Denmark zmp, BCG Pasteur zmp) was tested in SCID mice. Weight development of mice indicated that zmp mutant strains are as least as safe as their parental strains and thus have BCG like safety (in collaboration with G. Bancroft, LSHTM). In summary, results from various immunological and protection experiments suggest that BCG zmp1 is an innovative and promising vaccine candidate.
Joint PDT and CDT were established between TBVI and AERAS with UZH. Issues concerning strain characteristics, unpublished results, implementation and back-up strategies, exchange of material, production conditions etc. were discussed. Use of BCG zmp as either prime or boost vaccine is also considered.


WP 8 Project management
This work package had the aim to provide the project management infrastructure:
• To establish the management infrastructure (Project team, Steering Committee, External Advisory committees, project management tools, website)
 Provide Scientific and Technical Coordination for the activities of the project beneficiarys
 Provide Financial and Contractual Management of the Consortium
 Ethical issues

Project coordination and management of NEWTBVAC were organised according at 2 different levels, in order to handle two important and different aspects of the project management, the decision making level (Steering Committee (SC)), and the day-to-day operational level (Project Management Team (PM)). A legal framework for NEWTBVAC was ensured by a consortium agreement that was signed by all beneficiaries. The Consortium agreement constituted the legal framework for all activities carried out by the NEWTBVAC beneficiarys, that was integrated into eight work packages, and for the activities of the Steering Committee, and the General Assembly, and of the Project management team. It also constituted a legal frame work for the management of (new) and existing knowledge within NEWTBVAC.

The Four annual meetings were held in Les Diablerets. All partners attended these meetings. During these annual meetings separate meetings for each WP were scheduled in addition to the general Assembly where all partners were updated by the coordinator. All WP Leaders also organized an own WP-meeting during each year, these took place mostly in September.

During the lifetime of the project the communication activities focussed on optimizing and stimulating interactions between the project partners and with other projects and initiatives in Europe and the rest of the World. The coordinator closely follows and participated in joined activities like TRANSVAC, ADITEC and also with AERAS and SME’s (ao Biofabri, VPM etc.) and industrial parties.

In addition to the periodic reporting requirements of the EC an internal reporting round was held every 6 months to check on progress and results and this has been discussed in SC meetings. Each year the Steering Committee meet 2 times face to face and had in addition 2 TC.

One of the objectives of WP 8 was also to monitor relevant ethical concerns/issues of the project.
One major area of ethical concern was the use of animals. Animal research is an essential element of NEWTBVAC as there are currently no in vitro systems capable of replicating the effects seen in vivo from either immunisation or from challenge to infection. Therefore, the use of animals for this work was essential to validate the immunological responses and protection from challenge with M.tuberculosis by candidate vaccines, and in addition, has provided information of the safety of candidate vaccines. The second area of concern was the use of material of human origin. The NEWTBVAC project has gathered utilised new human samples and existing archived plasma and (T) cell samples for discovery and development of biomarkers. The coordinator has requested documentation from the beneficiaries indicating that these samples have been collected following formal approval from the appropriate Independent Ethics Committees and proper informed consent. The principles of the Declaration of Helsinki as amended by the 29th (Tokyo 1975), 35th (Venice 1983), 41st (Hong Kong, 1989), 48th (Republic of South Africa 1996) and 52nd (Edinburgh, Scotland 2000) World Medical Assembly and the ICH guidelines will be adhered to throughout the duration of NEWTBVAC.
The following actions were implemented to compile with this objective:
• The coordinator has ensured during the duration of the project that each beneficiary implemented their research project in full consideration of European and national legal legislation, ethical requirements and codes of practice. If requested, the Coordinator can provide the European Commission copies of the necessary authorisations as obtained from the relevant ethical bodies. This was checked on regular basis through a standardizes questionnaire on ethical issues that was filled in by each partner.
• A separate ethical advisory group section in WP8 was put in place with specific ethical deliverables ie. on 3R assessments, acquisition of relevant approvals, and preparing an ethical report for the periodic report of NEWTBVAC. This Committee was led by Dr. Frank Verreck (BPRC) and has on regular basis reported to the Steering Committee on specific ethical items. Also during WP meeting of WP 5 ethical concerns related to Animal studies were discussed between partners.
• The ethical reports were compiled by the coordinator based on a standardizes questionnaire on ethical issues that was filled in by each partner. The reports have been reviewed by two independent external reviewers, one with specific background and experience in ethical issues for the animal studies (dr. Jan Langermans), and one with specific background and experience on use of materials from human subjects, including children (Prof. Peter Smith). The advise or questions from the reviewers were taken into account and followed-up

Potential Impact:
4. The potential impact, main dissemination and exploitation results

4.1 Impact on health and prevention of disease
TB remains a major threat to humanity and cannot be defeated without effective vaccines. M.tb is one of the most efficient human pathogens, and one-third of the world's population is currently infected. The vast majority of TB cases occur in developing countries, especially in adolescents and adults between 15 and 45 years of age, the economically most active segment of the population. A major challenge for Europe and the world is the increasing incidence of multidrug-resistant TB (MDR-TB) and of extensively drug-resistant TB (XDR-TB). Fifteen of the 22 countries considered ‘MDR-TB high burden’ belong to the WHO Europe region. Despite the established long term global use of BCG vaccination, and major recent global efforts in more rapid and effective diagnosis, and observed treatment regimens, the impact on control of global transmission of TB and of TB disease has been relatively limited. Modeling studies show that new safer and more effective TB vaccines will have a big impact on TB disease burden while at the same time preventing development of MDR and XDR forms of TB that are associated with drug compliance. Effective TB vaccines would have tremendous impact on the TB associated morbidity and mortality by reducing TB prevalence and possibly infection (e.g. (e.g. Abu-Raddad et al. 2009. PNAS). The development of new TB vaccines candidates until preclinical development was one of the main aims of NEWTBVAC.

4.2 Impact on cost of TB treatment and cost of TB to the global economy
A recent business case “TB Vaccine Research & Development: A Business Case for Investment” estimates that the development of new vaccine by a preclinical and clinical portfolio management approach may cost approximately €600million over the next ten years. The business case proposes public and private sectors to work together globally to develop a new TB vaccine through a cost-effective manner and by selecting and supporting only the most promising TB vaccine candidate(s) at each development stage. These investments seem relatively small related to the estimated costs of TB in Europe or globally over this period, which are estimated to be €5.9 and €58 billion, respectively. Or compared to the cost of TB to the global economy which are estimated at 0.52 % of GNI which amounts to several hundreds of billion euros, annually (Diel et al. 2012. Eur Resp J).
4.3 Project main direct impact:
NEWTBVAC supported 40 TB vaccine approaches during the past 4 years. 22 of them moved from research to discovery, 6 from discovery to preclinical phase and 4 went to phase 1 clinical trial. 17 biomarkers have been further characterized and validated and 18 new biomarkers were identified.
With these results NEWTBVAC enriched the global TB vaccine pipeline. Through their capacity to harness different immune mechanisms, the new generation candidates will develop into vaccines that may prevent latent TB infection, or even actively reduce subsequent reactivation, or re-emerging, of TB disease. Further development is needed and if the outcomes of the clinical trials are successful, the developed vaccines will save millions of lives and euros, reducing the economic health care burden. This would be a step forward in the right direction for the elimination of TB.
Furthermore NEWTBVAC contributed to identify, optimise and evaluate of innovative approaches on correlates of protection and TB. Such correlates aim to facilitate the selection and prioritisation of candidate TB vaccines for human clinical efficacy testing, and help reduce the protracted time scale, large size and expense of human efficacy trials.

4.4 Main dissemination activities and exploitation of results
Dissemination:
Results from the project were published in peer reviewed journals. This was done as largely as possible. They were also submitted and presented in scientific meetings and congresses. Before submission for publication or presentation, care was taken of any relation to possible intellectual property protection before dissemination. The need for new vaccines to fight tuberculosis more efficaciously was presented and argued to decision makers and the boarder public through advocacy and awareness actions, such a participation on public meetings in parliaments and specialised fora, newspapers and other media where appropriate. This was done in close cooperation with Aeras, Stop TB beneficiaryship as well as with HIV/AIDS and Malaria initiatives. TBVI was playing a crucial role in this aspect of the project. TBVI website was updated timely with results of NEWTBVAC.
Developing and Eastern European countries:
TBVI and (members of) the current NEWTBVAC consortium all have strong links to African beneficiarys, including links with EDCTP. TBVI, Aeras and the STOP-TB beneficiaryship coordinated and organized the Global TB Vaccine meeting from September 21-24, 2010 in Tallin, Estonia and the Third Global Forum on TB Vaccines from 25-27 March 2013 in Cape Town, South Africa . A budget of was allocated for Eastern and/or Developing country participants to participate in these meetings. Many NEWTBVAC participants also participated in these meetings. TBVI was also part of the INYVAC project which supports participation of developing country participants to the ADVAC course and also coordinated a networking and capacity building project (TBTEA) funded by EDCTP.
Exploitation:
One of the strategic objectives of TBVI is to make the vaccines and biomarkers which come out of NEWTBVAC and other projects it is coordinating available the quickest possible. Therefore TBVI has an active policy to facilitate collaboration with vaccine producers who comply with European GMP. This policy will assure that promising candidates enter the industrial process at the earliest possible stage. Another important consequence of this policy is that production and registration requirements are incorporated very early in the product development in order to shorten the delay for access to market. Of NEWTBVAC one industrial beneficiary and one SME were part of the project. Others are active participating or collaborating with TBVI including Biofabri, Institut Mérieux.
In collaboration with other players like Stop TB Beneficiaryship, Aeras, TBVI wwas actively preparing the funding and global distribution of new vaccines for TB. It was supporting the policy of Advanced Market Commitments, a process first elaborated to support access to vaccines against streptococcus in developing countries. Together with other beneficiarys, TBVI has strived for maximal affordability from a financial point of view as well as from a logistic point of view and looked for all possible collaborations with existing global vaccine initiatives (UNICEF, WHO, GAVI, GFATM). Particular attention was paid to existing effective access mechanisms such as expanded programm on immunization of WHO, world wide distribution system of UNICEF, GAVI.
NEWTBVAC was the critical driver of European TB vaccine development as well as the constructive coordinator of European activities with international partners, notably the Aeras Foundation supported by the Bill & Melinda Gates Foundation in the US. With these other stakeholders NEWTBVAC leadership has a stringent gating strategy based on rational and broadly accepted criteria that allows selection/down-selection of candidates of the TB vaccine portfolio, in order to develop only the most promising candidates. This approach is the basis for the establishment of the Global TB Vaccine Partnership (GTBVP) which aims to manage a global portfolio by advising the support only the most promising preclinical and clinical stage candidates through this well-defined selection process. This partnership will be further established and implemented in the coming year.

Intellectual property:
Well aware of the need to protect discoveries, NEWTBVAC had a policy to protect discoveries which came out from this collaborative research. Discoveries were subject to patent protection and/ or trademarks. In total 2 patents and 1 trademark resulted from the project. This was done in the light of protecting knowledge and discoveries but aims not to be a hurdle to bring the appropriate vaccines to those who need them most, developing countries in particular. A statement confirming this was and is part of the consortium agreement. Ownership and intellectual property ownership was resided with the (individual) beneficiarys. The consortium agreement further defined these conditions.



List of Websites:
TBVI
Runderweg 6
8219 PK Lelystad
The Netherland
danielle.roordink@tbvi.eu
http://www.tbvi.eu/projects/newtbvac.html