Fig 1.
Isolation of nematode-infecting microsporidia.
A. Morphological screen for infected worms (here C. elegans JU2816). Compared to uninfected worms, infected adult worms have a paler body color. Note that the pale body color may result from many environmental conditions, and thus these animals were further screened by Nomarski optics for microsporidian infections. Scale bar: 0.2 mm. B. Geographic distribution of our collection of nematode-infecting microsporidia. Sampling locations are represented by differently colored symbols based on microsporidian species. Black symbols were used when different microsporidian species were found in the same location. The world map is cited from http://www.d-maps.com/carte.php?num_car=3267&lang=en. The France map is cited from http://www.d-maps.com/carte.php?num_car=2813&lang=en.
Table 1.
Collection of wild nematode-infecting microsporidia strains
Table 2.
Collection of other microsporidia species infecting wild nematodes
Fig 2.
Bayesian inference SSU rDNA phylogeny of microsporidia species.
SSU rDNA sequences from 45 nematode-infecting microsporidia species and 60 other microsporidia species in the databases were used. The tree was generated using MrBayes v3.2.2 and refined by FigTree v1.4.2. Model Kimura 2-Parameter (K2P) was applied. Branch colors show the posterior probability, with the corresponding color code shown on the left. The light green boxes designate microsporidia infecting terrestrial nematodes and light-blue rectangles designate those infecting marine nematodes. Scale bar indicates expected changes per site. Branches of species with more than one strain were compressed.
Fig 3.
Bayesian inference phylogeny of concatenated SSU rDNA and β–tubulin sequences of 22 microsporidia species, and comparison with the nematode host phylogeny.
Bayesian inference phylogeny (left) based on 49 sequences concatenated from SSU rDNA and β–tubulin genes of 22 microsporidia species. Model General time reversible (GTR) was applied. The branches were colored and annotated as in Fig 2. On the right is a diagram (generated based on phylogenies from [18,21,31–33]) showing the relative position of nematode species found with microsporidia infections. Nematode-infecting microsporidia pathogens and their hosts were colored based on host genus. Correspondent positions of nematode-infecting microsporidia and nematodes on their phylogenies indicate a possible coevolution of nematodes and their natural pathogenic microsporidia.
Table 3.
Molecular distances of microsporidia SSU rDNA.
Fig 4.
Spore morphology of the different Nematocida species by Nomarski optics.
A. Wild Caenorhabditis elegans strain JU1249, with Nematocida parisii infection. B. Wild C. elegans strain JU2520, with Nematocida ausubeli infection. C. Wild C. briggsae stain JU2747, with N. major infection. D. Wild Oscheius tipulae strain JU1510, with N. minor infection. A ~ D, large and small spore classes are indicated by larger and smaller arrows, respectively. Spores in each class are smaller than those in A-C in the corresponding class. E. Wild Rhabditella typhae strain NIC516, with N. homosporus infection. A single class of spore size is observed, often clustered inside vesicles as indicated with arrows. F. Wild Procephalobus sp. strain JU2895, with N. ciargi infection. A single class of spore size is observed, often clustered inside vesicles as indicated with arrows. Scale bar: 10 μm.
Table 4.
Spore sizes of each nematode-infecting microsporidia species, as determined by Nomarski optics.
Fig 5.
Ultrastructural observations of Nematocida ausubeli.
Transmission electron micrographs of N. ausubeli strain JUm2009 after high-pressure freezing/freeze substitution. A. N. ausubeli meront with two nuclei. B. A multinucleated meront. C. Late stage meront. D. Formation of sporoblasts by polysporous sporogony. E. Cluster of sporonts after sporogony; the arrowheads indicate the nascent polar tube and the arrow indicates the dense membrane structure. F. Sporoblast with a maturing anchoring disk and the dense membrane structure on the future posterior side of the spore (large arrow). Four nascent polar tube coil cross-sections (arrows) are visible, suggesting that this sporoblast may form a spore of large size. G. Late stage sporoblast. The arrow indicates the polar tube. H. Mature spore with surrounding additional membrane (arrows). The internal side of this membrane is coated. I. Mature spore with polar tube indicated by arrow. The anchoring disk and the membranes of the polaroplast are visible on the anterior side, chromatin and ribosomes on the posterior side. J. Cross-section of a spore vesicle containing four spores, each showing two polar tube sections (arrowheads). The upper inset shows two membranes around the vesicle (indicated by arrows). The lower inset shows an enlarged multilayered polar tube. K. A large size spore, with two insets showing the posterior vacuole and at least three polar tube coils (three cross-sections on either side of the spore, arrowheads). L. Lower magnification view of several N. ausubeli infection stages in host intestinal cells. Large arrow and small arrow indicate large spore and small spore, respectively. The large spore is that shown in panel K in another plane of section. Arrowheads indicate sporonts. Two multinucleate meronts are indicated. Scale bar is 500 nm, unless indicated otherwise. A, anchoring disk; Chr, chromatin; M, meront; Nu, nucleus; Pa, anterior polaroplast; Pp, posterior polaroplast; Pt, polar tube; Pv, posterior vacuole.
Fig 6.
Cell exit modes of Nematocida ausubeli and Enteropsectra longa.
A. Nematocida ausubeli. The top panel is an electron microscopy image of a Nematocida ausubeli spore (large arrow) exiting from the intestinal cell into the lumen. Arrows indicate the apical membrane of the host intestinal cell. The hypothetical diagram below illustrates the exit of N. ausubeli spores from intestinal cells by exocytosis. As in N. parisii [27,36], spores appear surrounded by a membrane that fuses with the apical membrane of the host intestinal cell, resulting in the release of spores. We also observed apparently mature spores without an additional membrane and do not know whether they will later acquire a membrane or exit in another manner. Earlier stages were omitted here for simplicity. B. Enteropsectra longa. The top panel is an electron micrograph of Enteropsectra longa spores exiting from the intestinal cell into the lumen, with the host intestinal cell membrane folding out around the E. longa spores (arrow). The diagram below illustrates the exit of E. longa spore from the intestinal cell. The host intestinal cell membrane folds out around the spore until the whole spore exits the cell, after which the host membrane around the spore seems to disappear. Meronts and sporoblasts are not represented in either panel. Scale bars: 500 nm.
Fig 7.
Spore morphology of Enterospectra and Pancytospora species by Nomarski optics.
A. Wild Oscheius sp. 3 strain JU408, with Enteropsectra longa infection. Gut lumen was indicated. Arrow indicates spores, here long and thin spores that are aligned along the apical side of intestinal cells. B. Wild O. tipulae strain JU2551, with Enteropsectra breve infection. Arrow indicates spores, here small spores along the apical side of intestinal cells. C. Wild O. tipulae strain JU1505, with Pancytospora philotis infection. Spores are found throughout the intestinal cell. D. Wild Caenorhabditis brenneri strain JU1396, with Pancytospora epiphaga infection. Spores are seen in the epidermal cells in the tail that does not contain any gut tissue (posterior to the rectum). Anterior is to the right. The "fur" on the outside of the cuticle is formed by unidentified bacteria (see [37] for another example). Scale bar: 10 μm in A-D. E. Two-dimensional diagram of Oscheius sp. 3 intestine infected with Enteropsectra longa. The intestine is formed of polarized epithelial cells. Enteropsectra longa starts to form spores along the apical side of the intestinal cells.
Fig 8.
Ultrastructural observations of Enteropsectra longa.
Transmission electron micrographs of E. longa strain JUm408 after high-pressure freezing/freeze substitution. A. E. longa meront. A nucleus is visible in the cytoplasm full of ribosomes. B. Lower magnification with a multinucleated meront. Two meronts are indicated, one with a single nucleus in the plane of section (left) and one with several nuclei (right). Two host nuclei are visible on the right, with a dark nucleolus. Intestinal cells contain two nuclei. C. Early sporonts with an electron-dense coat indicated by arrowhead. D. Sporont undergoing a cell division (big arrow); small arrow indicates junction of host intestinal cells; the arrowhead indicates a host Golgi apparatus. E. Mitotic spindle (arrows designate microtubules) in a sporont; the spindle plaque is indicated by an arrowhead. F. Nascent polar tube (arrow) in a sporoblast. G. Wrinkled sporoblasts (*). Arrows indicate the host rough endoplastic reticulum folding around the microsporidia. H. Late stage sporoblast in the center, mature spore on the top left; arrows indicate polar tubes. I. Mature spore with the anterior part of the polar tube, including the anchoring disk. J. Cross-section of mature spores. The exospore and endospore layers are shown in the inset. Arrowheads indicate polar tubes. K. Two mature spores in the intestinal lumen that do not show an additional membrane around them. Low magnification inset shows the positions of the two spores in the lumen and arrowhead indicates host microvilli. L. Low magnification view of cross-section of host, with the intestinal lumen in the center. E. longa spores (arrowheads) concentrate around the apical membrane of intestinal cell, while meronts and early sporonts are on the basal side. Scale bar is 500 nm, unless indicated otherwise. A, anchoring disk; Chr: chromatin; Ex, exospore; En, endospore; Lu, lumen; M, meront; Mi: host mitochondrion; Mv, host microvilli; Nu, nucleus; HNu host nucleus; Pt, polar tube; RER, rough endoplastic reticulum; St: sporont.
Table 5.
Nematode-infecting microsporidia specificity.
Fig 9.
Responses of C. elegans strains with transcriptional reporters C17H1.6p::GFP and F26F2.1p::GFP to exposure by different microsporidia.
Strains ERT54 carrying C17H1.6p::GFP (A) and ERT72 carrying F26F2.1p::GFP (B) were analyzed for GFP induction at different time points after infection with different microsporidia and the proportion of animals with GFP induction is shown. GFP was reproducibly induced in ERT54 and ERT72 upon infection with N. parisii, N. major and N. homosporus, while GFP signal was rarely observed in ERT54 and ERT72 inoculated with N. ausubeli or E. longa or the negative control. N. ausubeli did infect the C. elegans reporter strains, as monitored by DIC as in Table 5. C. Transcript levels for three genes were measured after 4 hours of infection of N2 C. elegans by N. parisii (ERTm1) and N. ausubeli (ERTm2). The fold increase in transcript level was measured relative to uninfected N2 levels. Infection dose was normalized between Nematocida by successful invasion events counted as intracellular sporoplasms at 4 hpi. To independently compare the microsporidian doses in parallel to the transcript quantification, we also measure the levels of Nematocida SSU rRNA after 4 hours of infection of C. elegans in the same experiment: we found that the rRNA level measured after infection with ERTm2 was 1.25-fold higher than that with ERTm1.
Fig 10.
Summary of the interactions between rhabditid nematodes and microsporidia in the wild and in laboratory.
A mosaic green color means that the corresponding natural infection was found. Plain green means that the infection worked in the laboratory and red means that the infection did not work in the laboratory. White: not determined.