Tuberculosis (TB), caused by Mycobacterium tuberculosis (MTB), is a leading cause of global infectious disease mortality. Advancement in TB research relies heavily on the use of animal infection models. However, all models have their limitations serving as bottlenecks in research, underlining the need for alternatives. This thesis describes the characterisation of an infection model for the MTB complex (MTBC) in Galleria mellonella (Gm) – the greater wax moth. Time-kill assays of Gm challenged with Mycobacterium bovis BCG (BCG), double auxotrophic MTB (SAMTB), or MTB (H37Rv) demonstrated that Gm can differentiate virulence between members of the MTBC, with establishment of a persistent (BCG) or proliferative (SAMTB and H37Rv) infection. A Gm-MTBC microscopy interaction study revealed internalisation of mycobacteria by phagocytic haemocytes (analogous to mammalian neutrophils and macrophages) as early as 1 hour post-infection. A non-replicative bacilli phenotype (lipid accumulation) was observed only in BCG infection. A histological study of infected Gm identified the formation of granuloma-like structures, containing non-replicative bacilli (loss of Ziehl-Neelsen-staining for SAMTB and H37Rv alone). The utility of Gm as a screen for antimycobacterial drug toxicity/efficacy and comparative MTB virulence was demonstrated using clinically relevant compounds, and isogenic MTB mutants (ΔphoP and ΔdosR), respectively. Proteomic and gene expression analysis of BCG infected Gm, identified immune responses indicative of phagocytosis, formation of granuloma-like structures, and secretion of antimicrobial peptides. Furthermore, proof-of-concept models of ex vivo granuloma-like structures and real-time in vivo drug screening are presented. These results indicate that Gm is a viable alternative infection model for TB being capable of mimicking critical aspects of the disease, and can be used to assess comparative MTB virulence, and as
6
a low-medium throughput antimycobacterial drug screen. The adoption and continued development of this model has the potential to substantially reduce/replace the use of animals in TB research.