Investigating the early interaction between Mycobacterium avium ssp paratuberculosis and the host using a bovine enteroid system

Abstract

Mycobacterium avium ssp paratuberculosis (MAP) is the causative agent of Johne’s disease (JD), a chronic granulomatous enteritis of ruminant’s prevalent world-wide. Infection of calves occurs through the faecal oral route, typically in animals <6 months old. Animals are asymptomatic for 2-5 years before clinical signs begin to show, which typically present as emaciation and chronic diarrhoea. In the subclinical phase, animals will have decreased milk yield, increased susceptibility to other diseases and decreased feed conversion. This has a severe impact on the farming economy and animal welfare, as affected animals are often prematurely culled.

Infected subclinical animals are extremely difficult to identify but can still act as a source of transmission for the rest of the herd by shedding MAP in their faeces. There is no treatment for JD, and the current diagnostic tests are ineffective. By investigating the initial interaction between MAP and the host at the intestinal lining, a greater understanding of MAP pathogenesis can be gained and better diagnostic and therapeutic targets can be identified.

In this work, proteins expressed on the surface of MAP were assessed for their ability to aid attachment, invasion and intracellular survival in epithelial and phagocytic cells when expressed on the membrane of a non-invasive E. coli host strain. The proteins investigated were encoded by mammalian cell entry (mce) genes, mce1A, mce1D, mce3C and mce4A, which have been implicated in attachment and invasion of epithelial cells by other mycobacteria. Interestingly, E. coli expressing Mce1A had enhanced uptake by phagocytic cells and E. coli expressing Mce1D had enhanced attachment and invasion of epithelial cells, but neither protein conferred this phenotype in both eukaryotic cell types investigated. To identify key intestinal cell types involved in MAP pathogenesis, bovine intestinal organoids (enteroids) were assessed for their ability to model a MAP infection in a physiologically representative system. Baso-out 3D enteroids, apical-out 3D enteroids and 2D monolayers were created, and the cell types present were compared to bovine intestinal tissue samples using RT-PCR and immunofluorescence microscopy. The models contained the mature epithelial cell types of the intestine including goblet cells, enteroendocrine cells, Paneth cells and enterocytes. 3D baso-out enteroids and 2D monolayers also contained proliferative cells, but the 3D apical-out enteroids did not and so could not be maintained past 2 weeks of culture.

The models were infected with two strains of MAP over the course of 72 hours, the reference strain K10, and a recent clinical isolate C49. MAP C49 was shown to be present in all three intestinal models in consistently higher numbers than MAP K10, quantified using qPCR of the genomic DNA. This indicates that MAP C49 was better able to infect these models than K10, which may suggest a loss of virulence in MAP K10.

Overall, the data presented has increased our understanding of MAP pathogenesis by emphasising the need for multicellular models which accurately represent the pathogen target cell type/s in vivo and the confirmation of the role of two hypothetical MAP proteins in cellular interactions.