Sydney L Miles: Recent advances in modelling Shigella infection

Shigella is further divided into the species Shigella boydii, Shigella dysenteriae, Shigella flexneri, and Shigella sonnei. S. flexneri and S. sonnei are endemic (see Glossary) in most regions and account for the largest proportion of contemporary infections. By contrast, S. dysenteriae expresses a toxin and causes dysentery epidemics in crowded settings with poor sanitation and hygiene, whilst S. boydii is rare and mainly found in South Asia [1]. Shigellosis is characterised by bacterial invasion of the human gut epithelium, which can result in a range of clinical manifestations, including watery or bloody diarrhoea, fever, and abdominal pain [2]. Shigellosis, although typically self-limiting, still causes ~216 000 deaths each year [1], with the greatest burden of disease falling on children under 5 years old in low- and middle-income countries (LMICs) where mortality is associated with malnutrition [3,4].

 Shigella can spread through ingestion of contaminated food and water, and as such, disease incidence is associated with limited access to water, sanitation, and hygiene (WASH). However, S. flexneri and S. sonnei can also be sexually transmitted, and in high-income countries (HICs) with strong WASH infrastructure, shigellosis is increasingly linked either to travel to regions where Shigella is endemic, or the sustained sexual transmission between men who have sex with men (MSM) [5]. The transmission of Shigella within the MSM community has been associated with the clonal expansion of multidrug and extensively drug resistant Shigella genotypes [6,7], presenting a major public health issue as treatment options become extremely limited.

Since mice are naturally resistant to shigellosis, alternative models have long been sought to fully understand the determinants underlying Shigella pathogenesis (reviewed in [8]). The first animal model of Shigella infection was established as early as 1955, when the Sereny test (a guinea pig keratoconjunctivitis model) was used to differentiate between invasive and non-invasive isolates [9]. Seminal work carried out using this model revealed that Shigella could spontaneously become avirulent, losing the ability to cause keratoconjunctivitis. This work provided the first clues towards understanding that the invasive phenotype of Shigella is dependent on the presence of a virulence plasmid (pINV), which encodes a type three secretion system (T3SS) [10,11] and is often unstable at environmental temperatures [12]. 

More recently, the use of guinea pigs has diverged away from the keratoconjunctivitis model, and more physiologically relevant guinea pig infection models (such as the intra-rectal challenge model) have been implemented for protective efficacy studies [13]. A rabbit ileal loop model, in which a ligated section of the ileum (small intestine) from adult rabbits is inoculated with bacteria, was then used to formulate a detailed model of Shigella pathogenesis in an intestinal context [14]. Work performed using this model revealed the precise route in which Shigella traverses through the intestinal M cells [15], before being engulfed by macrophages, which ultimately succumb to pyroptosis [16,17], triggering the characteristic inflammatory hallmarks of Shigella infection. Both guinea pigs and rabbits are small mammalian models able to overcome some of the barriers posed when modelling human-adapted infections, and their significant contributions enhanced our understanding of Shigella pathogenesis. Both models, however, neglect the natural route of Shigella infection and the complexity and costs of maintaining small mammalian models can be prohibitive to their widespread use.

Read full publication here