The dynamic response of organisms exposed to environmental pathogens determines their survival or demise, and the outcome of this interaction depends on the host’s susceptibility and pathogen-dependent virulence factors. The transmission of acquired information about the nature of a pathogen to progeny may ensure effective defensive strategies for the progeny’s survival in adverse environments. Environmental RNA interference (RNAi) is a systemic and heritable mechanism and has recently been linked to antibacterial and antifungal defenses in both plants and animals. Here, we report that the second generation of Caenorhabditis elegans living on pathogenic bacteria can avoid bacterial infection by entering diapause in an RNAi pathway-dependent mechanism. Furthermore, we demonstrate that the information encoding this survival strategy is transgenerationally transmitted to the progeny via the maternal germ line.

The interaction and communication between bacteria and their hosts modulate many aspects of animal physiology and behavior. Dauer entry as a response to chronic exposure to pathogenic bacteria in Caenorhabditis elegans is an example of a dramatic survival response. This response is dependent on the RNA interference (RNAi) machinery, suggesting the involvement of small RNAs (sRNAs) as effectors. Interestingly, dauer formation occurs after two generations of interaction with two unrelated moderately pathogenic bacteria. Therefore, we sought to discover the identity of C. elegans RNAs involved in pathogen-induced diapause. Using transcriptomics and differential expression analysis of coding and long and small noncoding RNAs, we found that mir-243-3p (the mature form of mir-243) is the only transcript continuously upregulated in animals exposed to both Pseudomonas aeruginosa and Salmonella enterica for two generations. Phenotypic analysis of mutants showed that mir-243 is required for dauer formation under pathogenesis but not under starvation. Moreover, DAF-16, a master regulator of defensive responses in the animal and required for dauer formation was found to be necessary for mir-243 expression. This work highlights the role of a small noncoding RNA in the intergenerational defensive response against pathogenic bacteria and interkingdom communication. IMPORTANCE Persistent infection of the bacterivore nematode C. elegans with bacteria such as P. aeruginosa and S. enterica makes the worm diapause or hibernate. By doing this, the worm closes its mouth, avoiding infection. This response takes two generations to be implemented. In this work, we looked for genes expressed upon infection that could mediate the worm diapause triggered by pathogens. We identify mir-243-3p as the only transcript commonly upregulated when animals feed on P. aeruginosa and S. enterica for two consecutive generations. Moreover, we demonstrate that mir-243-3p is required for pathogen-induced dauer formation, a new function that has not been previously described for this microRNA (miRNA). We also find that the transcriptional activators DAF-16, PQM-1, and CRH-2 are necessary for the expression of mir-243 under pathogenesis. Here we establish a relationship between a small RNA and a developmental change that ensures the survival of a percentage of the progeny.

No abstract

15The interaction and communication between bacteria and their hosts modulate many 16 aspects of animal physiology and behavior. Dauer entry as a response to chronic exposure 17 to pathogenic bacteria in Caenorhabditis elegans is an example of a dramatic behavioral 18 decision to survive. This response depends on the RNAi machinery suggesting the 19 involvement of sRNAs as effectors. Interestingly, dauer formation occurs after two 20 87 we will first refer to mRNAs. Differential gene expression of animals feeding on 88 pathogenic bacteria was made in reference to non-pathogenic E. coli OP50 using DeSeq 89 and EdgeR (Dataset 1). We considered differentially expressed genes those found to be 90 statistically significant by either analysis (Table S1 and 2). Differentially expressed mRNA 91 genes included coding and non-coding transcripts. We found that transcriptional changes 92 in mRNAs (coding and non-coding) were larger in the first generation (F1) than in the 93 second (F2) of animals fed with either pathogen ( Fig. 1A and B). Fig. 1C to E illustrates 94 the distribution of non-coding RNA expression in numbers (C and D) and types (E). 95Among non-coding transcripts we found piRNAs, 7k ncRNA, pseudogenes, tRNAs, 96 lincRNAs, asRNAs, snoRNA, snRNA and rRNAs (Fig. 1E). The F1 generation was more 97 abundant and diverse in differential expression than the F2s when feeding on either 98 pathogen. This difference between generations could suggest that changes in the F1 are 99 most dramatic because they reflect changes in bacterial diet composition, as well as 100 exposure to a pathogen. In this context, the subsequent generation may respond more 101 specifically to each new bacterium. 102Most mRNA genes were overexpressed or repressed in a generation or pathogen 103 specific manner (Table S3). There were no shared upregulated mRNA genes in both 104 pathogens and generations ( Fig. 1F). However, a number of genes were commonly 105 expressed in the F1, independently of whether animals were feeding on P. aeruginosa or 106 S. Typhimurium. These genes grouped in Metabolic Processes by gene ontology (Fig. S1A) 107 suggesting there is a common expression pattern that likely responds early to the change in 108 diet from E. coli OP50 and it is not maintained after the F1 (see below). Interestingly, 109 downregulated genes shared many more coincidences between pathogenic conditions ( Fig. 110 1G). Three genes, acdh-1, hpdp-1 and F54D5.12 were downregulated in both bacteria and 111 generations. To distinguish the response to pathogens from changes in gene expression 112 caused by the switch from E. coli OP50 to other bacteria we compared the expression 113 changes in animals feeding P. aeruginosa PAO1 and S. Typhimurium in the F1 to those 114 reported for Comamonas. aquatica (15). There were no coincidences in the upregulated 115 genes ( Fig. S1B), but several genes were repressed in animals feeding on all three bacteria 116 ( Fig. S1C), showing the change itself from E. coli OP50 triggers a characteristic common 11...