Mice continue to be a good model for preliminary assessment of tuberculosis vaccine candidates

Keywords:

• mice
• preclinical studies
• tuberculosis
• vaccines

The development of a vaccine involves several phases of study, where the use of animal models in the preclinical stages remains of paramount importance for the study of safety and efficacy in whole organisms rather than using isolated cells, or only bioinformatics predictions.

In the quest for new of tuberculosis vaccines, a number of live models have been used, including different mice, guinea pigs, rabbits, primates, bovine calves, zebrafish, rats, ferrets, mini-pigs, fruit flies, nematodes, flatworms, and even amoebae. Although every one of these species has contributed to emulate certain conditions of human disease, the most widely used continues to be the mouse model, and therefore we will focus this work in describing the preclinical use of this model organism.

Vaccine design

The objective of any vaccine is to provoke a protective and lasting memory immune response against a pathogen of interest, which is required to confer protection. Such information is gathered during the design of the vaccine and preclinical trials. Animal models, despite understandable ethical issues, continues to be useful because they allow to determine various aspects in terms of safety and efficacy of vaccines, which otherwise would be ignored with inherent risks to human beings. Regarding effectiveness, this is evaluated according to various parameters such as i) protection versus infection with the pathogen of interest or signs and symptoms of disease produced by such pathogen, ii) dose and formulation of the vaccine, iii) optimal route of administration, iv) evaluation of the onset, duration, type, and extent of immunity provided by immunization and v) the correlation of immunogenicity with respect to security.¹

Among the criteria for considering an animal model as relevant, we found the degree of similarity of the disease produced in the model in comparison to that in humans, including mode and route of infection, infective dose, access to different immune compartments, post- vaccination immune response, and the ability to correlate the protection and disease progression and pathogenesis. For tuberculosis studies, nowadays is accepted to consider as relevant surrogate markers, the bacterial load, pathological changes in the lung, reducing symptoms and relapse to active disease or damage associated with it.^1-3

As mentioned before, there are different animal models applied for research in the tuberculosis field, and although none of them completely mimics all aspects occurring during infection in human beings, all these models provide valuable information regarding particular stages of infection. Their use allows to study granuloma formation, liquefaction, cavity formation, and hematogenous spread to other organs from the lungs. These models can also reproduce disease symptoms such as fever, weight loss, abnormal radiographs and lung damage, and if the disease is untreated, eventually die, as in the case of humans. However, other than the immunological similarities of the disease between humans and animal models, the cost, availability, space and biosafety requirements for any particular experiment should be taken into account. The differences and complexity of each animal model together with the resources that are available for the tests mentioned above will influence the reproducibility of results, and the value of the preclinical analysis performed.

Murine models

The mouse is one of the most popular experimental models and widely employed having several species that can be used, the most common being C57BL/6, BALB/c, DBA/2 and C3H/HeJ, CBA and 129/SvJ, each one well characterized, but with some differences in immune response and ranging in their level of resistance and innate susceptibility to tuberculosis. The main differences among mice strains are the degree of susceptibility to infection and survival after infection. The CBA, DBA/2, C3H and 129/SvJ strains are classified as highly susceptible, while BALB/c and C57BL/6 are considered highly resistant, given its ability to contain the disease.⁴ Although both groups differ in their capacity to control bacterial replication in the lung, being lower in susceptible animals, in all cases they are able to contain bacterial growth in liver and spleen. In resistant mice, well organized granulomas can be found, while in susceptible mice lesions are often poorly organized and contain few necrotic cells.⁵ Following BCG immunization, mouse resistant strains C57BL/6 and BALB/c showed good protection induced by the vaccine, while BCG parenteral immunization did not confer protective efficacy in susceptible strains CBA/J or DBA/2.⁶

One of the most obvious advantages of murine models is their cost, since they allow the evaluation of many vaccine candidates and use a very large number of mice, with a vast availability of immunological reagents. Mice can be readily infected by different routes, including a low dose aerosol of microorganisms, with the multiplication of bacilli in the lungs and subsequent spread to the liver and spleen. The infection is controlled but not eliminated by cell-mediated immunity, mainly by the response of T cells.

In addition to these species, which are routinely used, some researchers have developed humanized models or deficient in some genes, in order to study some specific stage of the disease or specific mechanism of pathogenesis.⁷ Furthermore, with the use of immunodeficient nude or severely compromised mice, the assessment of safety for live vaccine candidates against TB can be an easy to use and reproducible substitute to evaluate the effect of vaccination in persons infected with HIV.

Of course, the nature of the protection and limited extrapolation to many aspects of human tuberculosis are the main disadvantages or murine models, coupled with differences in clinical manifestations as granuloma formation.⁸ In order to circumvent these limitations, different strategies have been employed. For example, very recently, by using diverse inbred mouse strains consisting of different founders and their recombinant progeny, it was found that genetic susceptibility to infection by M. tuberculosis can be dissociated from the efficacy provided by BCG vaccination to protect from infection,⁹ with infection outcomes resembling the variability found in humans, therefore providing a model system to assess vaccine efficacy in genetically heterogeneous populations. On the other hand, the limited short life of the mouse somehow reduces the feasibility of studies dealing with latency. Nevertheless, in this particular regard, efforts to emulate latency have been developed, where, for example, immunization via the aerosol route with a recombinant BCG strain overexpressing the 30-kDa major secretory protein (rBCG30) followed by aerosol-infection 6 weeks later with virulent M. tuberculosis H37Rv, lead to latent infection of C3HeB/FeJ mice.¹⁰ Likewise, intradermal infection of C57BL/6 mice leads to contained M. tuberculosis infection, which upon CD4+ T cell depletion results in progressive systemic spread of the bacteria and induction of lung pathology.¹¹

Concluding remarks

Animal models are important in different stages of research for the development of vaccines, for determining the mechanism of, route of, disease transmission dynamics, and the host immune response, to immunization and infection, as well as the duration of protection provided by particular vaccine candidates. While the success of a vaccine candidate against tuberculosis can be evaluated in terms of the protection afforded by its ability to reduce symptoms, pathology and bacterial load, not all models allow an effective discrimination of these parameters. Considering the availability of many parental and genetically modified inbred mouse strains, as well as the reagents available to monitor immune responses elicited by vaccine candidates, we think this model system still provide good value for stringent selection of new vaccine candidates against tuberculosis.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.