Materials and Methods

Worm host and bacteria system

As a bacteriovore, Caenorhabditis elegans interacts constantly with a variety of bacteria either by feeding or hosting them (Cabreiro and Gems, 2013; Garsin et al., 2001; Schulenburg and Ewbank, 2004). Consequently, C. elegans is an established model for studying innate immunity (Gravato-Nobre and Hodgkin, 2005), as it can be infected with its natural (Jansson, 1994; Schulenburg and Ewbank, 2004) as well as opportunistic pathogens (Garsin et al., 2001; Tan et al., 1999). Most pathogens are taken up orally by the worm (Marsh and May, 2012), and some can proliferate and colonize the worm gut (King et al., 2016; Rafaluk-Mohr et al., 2018).
Naturally, C. elegans is a self-fertilising hermaphrodite (Brenner, 1974), but in this experiment obligate outcrossing worm populations (line EEVD00) with males and females (hermaphrodites that carry the fog-2(q71) mutation) were used (Theologidis et al., 2014). This lineage was generated by Henrique Teotonio (ENS Paris) and encompasses the genetic diversity of 16 natural worm isolates (Theologidis et al., 2014). Worms were kept on Nematode Growth Medium (NGM), inoculated with Salmonella , hereafter referred to as food. Worms were infected with the pathogenic S. aureus (MSSA476) (Holden et al., 2004), which is virulent and kills worm hosts by lysing the intestinal cells lining the gut wall (Sifri et al., 2003). Worms were exposed to E. faecalis (OG1RF) (Garsin et al., 2001), which was isolated from the human digestive system, but has been previously shown to colonize and proliferate in the host gut (Ford et al., 2017; King et al., 2016; Rafaluk-Mohr et al., 2018), where it provides protection.

Experimental evolution - Design

Six single clones of E. faecalis (one for each of the six replicate populations) and a single population of C. elegans were the ancestors (hereafter referred to as the Ancestor) for all evolving populations. To account for potential differences in virulence, a stock of four clones of S. aureus was used for pathogen infections. Both C. elegans and colonising E. faecalis were allowed to evolve in presence of each other, while S. aureus was kept evolutionarily static. Infection withS. aureus was varied over host evolutionary time (indicated by purple in Table 1) to represent temporal heterogeneity in pathogen infection, including a range from always to every 2ndgeneration, every 5th generation, and never (Table 1). Moreover, we included differences in whether pathogens were present at the initial formation of the symbiotic interaction or later (2.1. vs. 2.2., and 5.1. vs. 5.2. in Table 1). Controls for lab adaptation were maintained for the host (No Protective-Microbe control, NPM in Table 1) and E. faecalis(No Host Control, NHC in Table 1).