We defined the host transcriptional response to viral infection in C. elegans and C. briggsae. From our statistical analysis of Orsay virus infections in N2 and rde-1 mutant strains, we identified a total of 320 DEGs in C. elegans, of which 108 DEGs were shared. In the rde-1 Orsay virus infection, there were more DEGs compared to infection of N2. In addition, the magnitude of the transcriptional changes in rde-1 was generally greater. One possible explanation for this observation is that the higher levels of viral infection in rde-1 may have created a more significant perturbation from the basal state, leading to a more robust transcriptional response. Alternatively, the lack of competent RNAi in rde-1 may have resulted in induction of a distinct, compensatory host response. One potential limitation of these studies is that Orsay, Santeuil, and N. parisii infection is thought to be limited to at most the 20 intestinal cells present in Caenorhabditis nematodes. Because our transcriptional profiling used RNA extracted from populations of entire animals (each C. elegans has 959 somatic cells), some transcriptional responses may have been masked by the basal level of transcription in the uninfected cells, and thus our results are likely an underestimate of the transcriptional changes occurring in the intestinal cells.
Strikingly, 108 of the N2 DEGs were also differentially expressed following infection with the microsporidium, N. parisii (Fig. 1). Orsay virus is a small single stranded RNA virus with a bipartite genome of 3.6 Kb and 2.6 Kb that is only known to encode three proteins . By contrast, N. parisii has a 4.1 Mb genome and encodes more than 2000 genes [27, 37]. Despite the lack of obvious similarity between these two microbes, the fact that a significant fraction of the transcriptional response to these two pathogens overlapped suggests that C. elegans may have some form of a universal “stress response”. One clear commonality between the two is that they are both intracellular intestinal pathogens of C. elegans; in fact they are the only intracellular pathogens of C. elegans described to date. Thus, the conserved transcriptional response may reflect recognition of some shared intracellular perturbation. Interestingly, although some of these shared response genes are potentially involved in the ubiquitin ligase pathway, the majority of the shared response genes are largely unannotated genes of unknown function. These genes could play important roles in immunity against pathogen infection. Alternatively, it is also possible that these genes are important for pathogen infection, and that the pathogen alters the transcriptional response to facilitate infection and replication.
Many of the characterized genes induced by Orsay virus or N. parisii infection in C. elegans were genes in the ubiquitin ligase pathway. When challenged with either Orsay virus or N. parisii, there were 35 unique F-box related or MATH domain genes up-regulated. In addition, SCF complex genes, such as skr-4, were up-regulated in all C. elegans infections while skr-5 and cul-6 were up-regulated in the rde-1 mutant infected with Orsay virus. Most of the F-box and MATH family members have sites in their substrate binding domains that are under strong positive selection and are greatly expanded in C. elegans in comparison to humans . This suggests a possible role of ubiquitin ligase as part of the C. elegans host-pathogen interaction to restrict pathogen proliferation. Indeed, SCF ubiquitin ligases are demonstrated as a line of defense against infection by Orsay virus and N. parisii in C. elegans . Intriguingly, none of the DEGs in C. briggsae were known F-box or MATH genes, suggesting that these ubiquitin ligase pathways may be a specific C. elegans response.
There are varying degrees of conservation between Orsay virus response genes to other pathogens of C. elegans. We analyzed previously published transcriptional profiling studies of infection by 8 bacterial and fungal pathogens and identified three that have a significant fraction of DEGs shared with Orsay virus infection. The three pathogens, P. aeruginosa, P. luminescens and D. coniospora, all can affect the intestine of the worm, but each does so in unique fashion. P. aeruginosa PA14 primarily kills by excreted toxins, P. luminescens colonizes the intestinal lumen, which is characterized by the appearance of cytosolic crystalline structures of unknown origin , and D. coniospora produces threadlike hyphae that penetrate and eventually kill the infected animal . Other pathogens that also target the intestine such as E. faecalis and S. aureus did not have significant DEGs in common with Orsay virus infection, demonstrating a specificity of the host response. The different responses of C. elegans to various pathogens suggest the existence of distinct sensing and regulatory mechanisms. One potential regulatory element in response to virus infection is drh-1, a RIG-I like protein in C. elegans. Previous studies have determined that drh-1 both acts directly as a effector in the RNAi pathway to restrict virus replication and as a sensor of virus infection critical for downstream host responses [19, 20].
Comparative analysis of the DEGs in virus infected C. elegans and C. briggsae identified 58 C. elegans genes whose C. briggsae orthologs were also differentially expressed. Of those, 29 were shared between the three conditions: Orsay virus [N2], Orsay virus [rde-1], and Santeuil virus [JU1264]. Strikingly, 14 of the 29 genes were members of a single gene family, the C17H1 family genes in C. elegans. Induction of members of this gene family in response to viral infection was conserved in two divergent Caenorhabditis nematode species despite ~18 million years of host evolution. Furthermore, analysis of other published transcriptomes identified induction of C17H1 family genes by bacterial and fungal pathogens. The upregulation of a subset of these genes by disparate microbes such as virus, bacteria, and fungi raises the possibility that this gene family may form the core of a pan-microbial stress response. To date, there has been no reported function associated with these family members. The large number of paralogs induced following Orsay virus or N. parisii infection suggests the possibility of functional redundancy, which would provide a challenge in experimental testing of the functions of these genes.