REVIEW

MVA: A promising viral vector platform for Asian Infectious Diseases by

Stephen Lockhart

Senior Vice President, Vaccine Development, at Emergent BioSolutions (EBS)

38

REVIEW

U

ntil recently vaccines have largely fallen into two main categories. There are subunit vaccines in which purified proteins or carbohydrates are used to induce protective immune responses and live attenuated vaccines, which are weakened versions of pathogenic bacteria or viruses. In developing new vaccines both approaches present challenges. To induce an immune response, a vaccine must do two things; • Deliver a specific protective antigen • Deliver a signal to tell the immune system that it must respond to the antigen. This is often called a “danger signal”. Viral vectors do both. To create a viral vector vaccine, the gene for a protein antigen from a pathogen is inserted inside the gene of a harmless virus. When the viral vector is administered to a person it expresses the inserted antigen alongside the viral antigens. The viral vector creates proteins inside human cells—whereas subunit vaccines deliver proteins extracellularly—viral vectors can induce an immune response more like that of a natural infection than many subunit vaccines. However, because it is simply expressing the antigen from the pathogen inside a completely different virus it cannot cause the targeted disease, unlike live attenuated vaccines. MVA (Modified vaccinia Ankara) is an attractive viral vector. It was created from a vaccinia virus isolated in Ankara, Turkey, and was attenuated by hundreds of passages in chicken cells. During this process MVA lost around 15% of its genome and lost the ability to replicate in mammalian cells, while remaining capable of replicating in bird cells. MVA is a DNA virus, so genes for protective antigens can be inserted into a relatively stable genome; the large number of deletions provides several potential insertion sites for foreign genes to be inserted. Genes can be inserted by a process of homologous recombination using synthesised genes inserted into plasmid DNA, creating a recombinant MVA (rMVA). Importantly, the MVA virus DNA is not incorporated into the host genome, so that infection is purely transient. This is quite the opposite from the requirement for gene therapy, where long term expression and gene integration are desirable features.

After being taken up by mammalian cells rMVA is able to go through most stages of infection, including protein synthesis of recombinant genes. However, it is unable to repack into new viruses so when the host cells dies, antigens are released without further spread of virus. As a result, MVA has little risk of causing systemic disease, whereas the parent vaccinia virus has a risk of causing systemic infection. On injection, MVA induces production of IL-12, interferon-α, interferon-γ (IFNγ) by dendritic cell which drive a cell-mediated immune response. As the viral proteins, including the recombinant protein, are released they are thus in an environment where a range of immune responses can be induced. These include antibody responses induced by the protein as well as T-cell responses induced by a process of cross-protection, in which the proteins are taken up by antigen presenting cells and processed into fragments before being presented to activate cognate T-cells. As a result, rMVA can induce a broad immune response to the protein coded for by the inserted foreign antigen. A range of assays are used to assess the cellular immune response to a vaccine. The IFNγ ELISpot quantifies peripheral blood mononuclear cells (PBMC) expressing a single cytokine marker. In brief, cells are stimulated with the recombinant antigen and then are exposed to antibodies identifying IFNγ. The bound IFNγ antibodies are identified by secondary antibodies carrying a colorimetric marker. Other methods look at multiple cytokine markers (intracellular cytokine staining) and using flow cytometry cell populations expressing not only multiple cytokines but also other markers such as CD4, CD8 and so on can be assessed. Newer methods include transcriptional profiling of RNA expressed in selected peripheral blood mononuclear cells. Each assay has particular uses. The ELISpot is readily standardised and can be performed in relatively large numbers, particularly when automated. If used with frozen samples it can be fully validated. However, it only looks at a single dimension of the response, whereas newer assays create a broader picture. Antibody responses should not be forgotten. And for any pathogen there may be specific

functional assays, both cellular and humoral, that are required to characterise the response to a vaccine. rMVA-induced responses include strong T-cell responses, particularly CD4+ but also CD8+ responses, especially when given after priming by an antigen related to the foreign antigen present in the rMVA, so called “heterologous prime-boost”. Essentially, the same or an immunologically similar antigen is used to immunize the host on two occasions, a ‘prime’ and a ‘boost’ where the antigen is presented in different (heterologous) contexts, such as different viruses, protein antigen followed by a virus, etc. rMVA induced T-cell responses include polyfunctional T-cell responses in which two or more cytokines such as IL-12, TNFα, IFNγ and IL-17 are expressed by individual cells. Heterologous priming may also occur after exposure to a natural infection or more deliberately after a different vaccine expressing the same antigen, in a socalled heterologous “prime-boost” regimen. Although antibodies can also be produced by heterologous prime boost regimens, repeated administration of rMVA, so-called homologous prime-boost, is particularly good at inducing antibody responses. However, other factors influence the pattern of the immune response, including the antigen and the host immunological status. A rMVA tuberculosis vaccine is one of the more advanced constructs in clinical trials. MVA-85A expresses the gene for the mycobacterial protein 85A. This immunodominant antigen is a component of the mycolyl transferase complex and this essential function may explain why it is well conserved, not just within M. tuberculosis, but also in other pathogenic mycobacteria such as M. bovis, environmental mycobacteria and BCG vaccine, which is an attenuated form of M. bovis. Studies to date indicate that MVA-85A induces a T-cell response in naïve individuals, but is particularly potent in inducing a T-cell response in subjects primed with BCG or exposure to other mycobacteria, including M. tuberculosis in subjects with latent tuberculosis infection. The T-cell response induced by MVA85A is CD4+ biased, although antigen-specific CD8+ T-cells are also induced. The CD4+ bias may be useful in prevention of tuberculosis, as CD4+ cells may be particularly important 39

REVIEW Table 1. Some of the infectious diseases for which MVA based vaccines have been studied in humans.

Disease

Use

Stage

Tuberculosis Boost BCG induced immunity Boost immunity induced by exposure to Mycobacteria

In efficacy studies in infants and HIV infected adults

HIV

Immunogenicity studies completed. Related fowlpox construct used as prime followed by protein boost shows modest efficacy

Boost following DNA vaccine Homologous prime boost

Malaria

Boost immunity following DNA prime vectored vaccines

Effective in human P. falciparum challenge studies

Influenza

Boost natural immunity to conserved antigens

Human immunogenicity studies and a human challenge study completed

HCV

Stand alone vaccine

Phase 1

JEV

Stand alone vaccine

Phase 1

in prevention of tuberculosis; the depletion of CD4+ cells by HIV infection is an important factor in susceptibility of HIV-infected subjects to tuberculosis. The CD4+ T-cells induced are polyfunctional, with cytokine production including interferon-γ, TNF and IL-17. MVA-85A has not raised any safety concerns in trials to date, including studies in infants, adolescents and adults as well as adults with latent tuberculosis, HIV infection or both. MVA85A has advanced into efficacy studies in BCG-exposed, non-HIV infected infants in South Africa as well as HIV-infected adults in South Africa and the Gambia. Enrolment is complete in the infant study and initial results are expected in 2012. The use of rMVA constructs have been studied in a number of infectious diseases (See Table 1). There are some constraints on the type of antigen that can be delivered by MVA. The insertion is of a gene, so the antigen must be the protein expressed from that gene. Large genes may be inserted. It is estimated that up to 25kb may be inserted though there is no clear information on the upper limit. If anything it is the complexity of creating the DNA insert which limits the size of inserts. There are multiple sites in the MVA genome that are available for inserting heterologous genes without significant effect on viral function. Additionally, a range of 40

promoters are available to drive transcription of inserts. Among the infectious diseases for which human rMVA trials have been conducted, a number have significant impact in Asia, including influenza, tuberculosis, HIV, JEV and HCV. However, many more rMVA constructs have been studied in animal models of many other infectious diseases, including those with importance in Asia (Table 2). Moreover, MVA may be a useful vector for a number of significant infections of other mammals, including rabies, equine influenza virus and canine distemper. In oncology, there is a search for ways to induce strong cell-mediated responses to tumour-specific antigens and rMVA have been studied in this context. The field is challenged by the dilemma that immune control of disease is likely to be most effective in early disease, whereas diagnosis is generally late in the progress of cancer and it is difficult to use experimental therapies in previously untreated subjects. Moreover, control of disease with one or even a few antigens is demanding. Nonetheless, cancers that have or are being studied in human trials include renal cell carcinoma, prostate cancer, localised cervical cancer (up to CIN III) and breast cancer. Viral vectors may be of particular interest where a viral infection is associated with cancer. For example, an

Table 2. Some of the infectious diseases for which MVA based vaccines have been studied in preclinical models of human disease, but not yet in humans.

Rabies Severe acute respiratory disease Dengue Influenza Measles Cytomegalovirus Respiratory syncytial virus Leishmaniasis

recombinant MVA against Epstein Barr Virus (EBV) is being studied in nasopharyngeal cancer in Asia, where EBV is considered an important causative agent. The safety profile of MVA either alone or as recombinant constructs is well established. MVA was administered to over 120,000 Germans as a pre-vaccine before vaccinia and no significant issues were identified. MVA causes a local reaction is most subjects, particularly when administered intradermally, with some crusting before healing; this is generally mild. Typical vaccination systemic responses are seen, with a low rate of mild fever. No safety signals of concern have emerged. Before rMVA can be routinely used as a prophylactic vaccine, it must be manufactured consistently, at a suitable scale and cost for distribution across developed and developing country markets. To date, trial supplies of MVA have been made using primary chick embryo fibroblasts. Such a process is difficult to run consistently or at large scale. A number of groups have therefore been working to develop continuous cell lines that might support MVA production. As MVA is largely incapable of replication in mammalian cells it has been necessary to develop avian cell lines, which are permissive for growth. These are now being used to develop larger scale manufacturing processes.

REVIEW As the recombinant insertion forms only a small part of the virus, we believe that once a manufacturing process has been developed, a very similar process could be used for any rMVA. Moreover, using modern single use bioreactors and downstream equipment means it should be relatively straightforward to recreate manufacturing facilities even in areas with relatively limited infrastructure. This lends itself to use by small regional manufacturers for regional priorities. MVA is well-suited to convenient distribution and delivery pathways. The virus is highly stable when frozen and this may allow primary distribution and storage as a frozen product, with secondary distribution and storage in a conventional cold-chain, as is done for oral polio vaccine. The virus is

also stable on lyophilisation and this is an additional option for convenient distribution. Emergent BioSolutions is a biopharmaceutical company which makes vaccines and immune based therapeutics for infectious diseases, autoimmune disease and oncology. It currently markets Biothrax® (Anthrax Vaccine Adsorbed), a vaccine for prevention of anthrax, which is approved in US, India and Singapore. Emergent BioSolutions has an extensive product pipeline, including vaccines based on MVA. MVA85A, a vaccine for tuberculosis, is in development by Oxford Emergent Tuberculosis Consortium (OETC), a joint venture between Emergent BioSolutions and Oxford University. Emergent BioSolutions develops new recombinant MVA candidates at its laboratories in Munich, Germany. These

product candidates are based on MVAtor® (Modified vaccinia virus Ankara vector), a purified strain derived from stocks owned by the Bavarian government. Using MVAtor, Emergent BioSolutions is developing a “universal” vaccine for influenza A, which studies have shown provides protection to mice from a range of strains using conserved influenza antigens. In a Singapore-based joint venture with Temasek Life Sciences Ventures, a pre-pandemic H5N1 vaccine candidate is being developed. By use of haemagglutinin gene inserts, neutralising responses can be produced across a wide range of H5N1 clades. In summary, MVA has a wide range of potential applications for production of candidate vaccines against a number of important Asian infectious diseases.

41

REVIEW

Key terms: • Antigen: An antigen is a part of an infectious organism that can induce an immune response. It is usually a protein or a polysaccharide. • Immune response: These usually fall into two types. Antibodies, which are made by B-lymphocytes, are useful for binding to the surface of pathogens when outside cells, for example when circulating in the blood. Cellular immune responses involve T-lymphocytes and are mainly good at killing host cells already infected by intracellular organisms, hence stopping infection of additional host cells. • Pathogen: Many microbiological organisms such as viruses, bacteria, fungi and parasites are harmless to humans. However, some do cause disease and these are termed pathogens. • Subunit vaccines: contain a purified antigen, but often require an additional component to switch on the immune response. This signal might be an adjuvant or, particularly for polysaccharide antigens, a conjugated protein. This approach is often good for inducing antibody responses but is sometimes limited in producing cellular immune responses. The use of newer adjuvants requires a lot of work to ensure that the non-specific effect on immune responses will not have unwanted effects. There are many successful subunit vaccines including diphtheria and tetanus vaccines, hepatitis B vaccine, pneumococcal conjugate vaccines, human papilloma virus (HPV) and many more. • Live attenuated vaccines: have the advantage of presenting the protective antigens together with a readymade “danger signal”, as the body sees the organism as a reason to mount an immune response. There may be difficulties balancing the risk of disease if organisms are inadequately attenuated and limitation of immune response if organisms are over attenuated. Nonetheless, many successful live attenuated vaccines include smallpox (vaccinia),measles, mumps, rubella, chickenpox (varicella) and rotavirus vaccines.

About the Author Stephen Lockhart is Senior Vice President, Vaccine Development, at Emergent BioSolutions (EBS). EBS develops, manufactures and commercializes vaccines and immune-based therapeutics for infectious diseases, oncology and autoimmune diseases. EBS produces Biothrax(r), the only FDA-approved vaccine for anthrax. Vaccines in clinical development include MVA85A for prevention of TB and Typhella(r) for prevention of typhoid. Stephen is a Director of OETC and the Director and General Manager of Singapore-based EPIC Bio. Stephen studied medicine at Cambridge and Oxford Universities, graduating in 1980. In 1986, he joined the pharmaceutical industry to work in clinical research. Before joining Emergent BioSolutions, he was Head of bacterial vaccine clinical research in Wyeth Research, where he worked on the development of many vaccines, including meningococcal and pneumococcal conjugate vaccines.

42