Background

Infections of the central nervous system (CNS) are devastating yet remain relatively rare due to the presence of specific protective layers. In particular, the blood–brain barrier (BBB) is a selective physical and chemical filter, and performs its neuroprotective role by controlling molecular import into the CNS and blocking the entry of circulating cells (Daneman and Prat, 2015). In higher vertebrates, brain microvascular endothelial cells, harbouring intercellular tight junctions, form the core structure of the BBB. In addition, pericytes, astrocytes, and an extracellular matrix basal membrane regulate BBB integrity and functions (Obermeier et al., 2013; Daneman and Prat, 2015; Saunders et al., 2016). However, several pathogens have developed strategies to overcome the BBB and invade the CNS (Coureuil et al., 2017). To study interactions between pathogens and the BBB, various cell culture systems have been developed, from simple endothelial monolayers to more complex multicellular cultures, combining endothelial cells and pericytes (Sivandzade and Cucullo, 2018). While these in vitro models have proposed a number of BBB crossing mechanisms, they remain limited in recapitulating the 3D architecture of the BBB and its interactions with the complexity of CNS cells (Sivandzade and Cucullo, 2018; Jackson et al., 2019). Moreover, organoids also present architectural limitations (Bergmann et al., 2018). To bypass this restriction, a number of animal models, in particular rodents and zebrafish, have been used (Dando et al., 2014; Jackson et al., 2019), allowing the investigation of infection mechanisms in vivo. However, in addition to cost and ethical issues, their genetic and cellular complexity prevents systematic, screen-based approaches. Separating the systemic from the local impact of these immune challenges is also difficult in such context, and infections can have lethal consequences for the host. In addition, culturing CNS tissue of these animal models is still difficult.

Drosophila melanogaster represents an original and powerful model to study CNS infection, from identifying novel mechanisms of BBB-pathogen interactions to understanding the extent of consequences on the diversity of CNS cell populations. The fly BBB is composed of two glial layers exhibiting conserved neuroprotective mechanisms with the mammalian BBB. The subperineurial glia (SPG), forming an epithelium-like structure with septate junctions, represent the core structure and physical filter of the BBB (Stork et al., 2008; Hindle and Bainton, 2014; Weiler et al., 2017; Babatz et al., 2018). The perineurial glia, a hemolymph sensor, and the extracellular matrix both cover the SPG (DeSalvo et al., 2014; Parkhurst et al., 2018). Many transporter proteins expressed by the BBB to allow the selective shuttling of diverse solutes and xenobiotics are actually conserved, including solute carriers, ATP-binding cassette transporters, and lipoprotein receptors (DeSalvo et al., 2014; Hindle and Bainton, 2014; Weiler et al., 2017). Moreover, many aspects of mammalian neurogenesis are conserved in flies (Mira and Morante, 2020), enabling us to probe the impact of infection on such process. Drosophila neural stem cells self-renew via asymmetric cell division, switch from quiescence to proliferation, and are present within a specific microenvironment sharing neurogenic features and common constituents with the mammalian niche, including glial populations and the BBB (Homem and Knoblich, 2012; Otsuki and Brand, 2017). Finally, due to Drosophila unrivalled genetics, each of the BBB cell layers and the different CNS populations can be manipulated independently, in parallel with each other, and in a temporally controlled and spatially restricted manner. These features provide a simpler, genetically tractable model to study CNS infection, from pathogen entry to downstream consequences for the host.

We have devised a double, complementary approach to generate CNS infection in Drosophila:

1. An ex vivo protocol that mimics CNS infection. Here, an intact larval CNS explant is cultured in a specific medium inoculated with pathogens (Benmimoun et al., 2020). This explant setup is first designed to allow fast screening of pathogens, mutant strains, and culture conditions for the generation and outcome of CNS infection. In this frame, it has first allowed the identification of a number of prokaryotic and eukaryotic neurotropic pathogens, known to trigger meningitis and/or encephalitis in mammals, that are able to cross the fly BBB. On the prokaryotic side, this included Streptococcus agalactiae (Group B Streptococcus, GBS), Streptococcus pneumoniae, Listeria monocytogenes, and the strictly human Neisseria meningitidis. The eukaryotic Candida albicans and Candida glabrata were also found to reach the Drosophila CNS, but not Cryptococcus neoformans. This ex vivo platform was further used to screen a collection of GBS mutant strains, by which surface lipoproteins were identified—more precisely the lipoprotein Blr—as virulence factors crucial for BBB crossing (Benmimoun et al., 2020). Finally, its combination with Drosophila genetics revealed the Drosophila lipoprotein receptor (LpR2), which is expressed on the surface of the BBB layer on the SPGs, as a host receptor for Blr, and important for CNS invasion by GBS (Benmimoun et al., 2020).
   

Another asset of the explant protocol is to study the effect of pathogenic infection on the different cell populations outside of the systemic context (for example, outside of systemic inflammation, a common hallmark of infection), potentially revealing other layers of response that could be otherwise hidden. For example, the lactic acidosis generated in the milieu by GBS could be quenched by buffer solutions, revealing the contribution of this phenomenon to BBB weakening (Benmimoun et al., 2020).

2. An in vivo setup through direct microinjection of the pathogen into the fly circulatory system (Benmimoun et al., 2020; Winkler et al., 2021). This approach aims to evaluate the impact of infection i) on the CNS, this time including the contribution of systemic components, such as circulating immune cells, and ii) at the level of the whole organism, by assessing a diversity of host–pathogen interactions, in particular virulence. Comparison with the explant system allows to confirm or infirm the importance of molecular and cellular mechanisms in a whole-body context, offering the possibility to discriminate between local and systemic effects, and to identify the hierarchy between and dominance of specific mechanisms.
   

This protocol was first used to determine the conservation of GBS Blr and Drosophila LpR2 interactions in BBB crossing mechanism and CNS invasion, initially identified through the explant setup (Benmimoun et al., 2020). Moreover, assessing survival overtime of the injected larvae in this setup demonstrated that Blr is a virulence factor in Drosophila, a property further validated in a mouse model of GBS hematogenous brain infection (Benmimoun et al., 2020). This whole in vivo CNS infection protocol also participated in revealing a new inflammatory mechanism, in which glial cells direct CNS infiltration by macrophages in response to GBS infection (Winkler et al., 2021).

Here, we detail these two complementary platforms to elicit infection of the Drosophila developing CNS by mammalian neurotropic pathogens. By taking advantage of the Swiss knife of Drosophila genetics, the architectural preservation of the CNS tissue, and the similarity of many fundamental cellular and molecular mechanisms with mammals, these approaches allow more systematic screening of wild-type and mutant pathogen strains, as well as targeted genetic manipulations, to infer precise causal relationships between cellular layers or between molecular pathways during infection. Ultimately, these setups can be used to study all steps of the CNS infection, from crossing of the BBB to downstream consequences on the host, including neuroinflammation and impact on neurogenesis.

Although these methods are described for prokaryotic and eukaryotic pathogens, they could be adapted for other pathogens (virus, parasites) as well as for studying systemic extrinsic stresses, and for investigating the entry of therapeutic molecules into the CNS.