The microsporidian Nematocida parisii is the first characterized intracellular pathogen of Caenorhabditis elegans; and a second Nematocida species, called N. sp. 1, was also reported from an Indian C. briggsae isolate (Troemel et al. 2008). In our lab we further established a collection of wild rhabditid nematodes suspected with microsporidian infection. Here, we characterize molecularly these wild microsporidia and study their specificity of infection.First, the potentially infected rhabditid isolates were characterized using newly designed PCR primers specific for 18S and beta-tubulin genes of microsporidia. In addition to C. elegans, we found wild C. briggsae and Oscheius tipulae as natural hosts for N. parisii; in addition to C. briggsae, we found C. elegans and C. remanei as natural hosts of N. sp. 1. We further found three new microsporidian species. Two of them are close to the Nematocida species in microsporidian clade II and preliminarily named N.-like sp. 2 and N.-like sp. 3. The third species (or group of species), infecting Oscheius, is very distant, and found in microsporidian clade IV, that includes most human pathogens. They are closest to Orthosomella operophterae (with 18S), and thus preliminarily named Orthosomella-like spp.. Phylogenetic analyses with the 18S and beta-tubulin genes provide the same relationships of Nematocida spp. and the three putative new microsporidia species. Co-infections of two microsporidian species occurr in two wild O. tipulae strains, of which one infected with N. parisii and O.-like spp., the other infected with N. sp. 1 and O.-like spp. (confirmed by specific FISH probes).Secondly, two transgenic C. elegans strains ERT54 jyIs8[C17H1.6p::gfp;
myo-2p::mCherry] and ERT72 jyIs15[F26F2.1p::gfp;
myo-2::mCherry], which express a constitutive fluorescent Cherry marker and will induce GFP upon infection with N. parisii (Bakowski et al. 2014), were used in infection tests. Our results show that N. parisii, N. sp. 1, N.-like sp. 2 and N.-like sp. 3 can infect ERT54 and ERT72, forming meronts and spores; N. sp. 1 does not induce the GFP reporter (or only rarely), while the other three induce it consistently, suggesting that C. elegans has a different transcriptional response to the infection of N. sp. 1 than to other Nematocida spp. (see also abstract of Luallen et al.). The results also showed that O.-like spp. could not infect the two transgenic C. elegans strains (nor induce the GFP signals).