Fig 1.
A new microsporidian species that infects C. elegans.
(a) Infected head region of a live C. elegans animal from strain JU2807 (derived from the wild-isolated P0 animal, see S1 Fig) showing a large group of structures that appear to be meronts (Me) adjacent to the pharynx (Ph) and intestine (In). (b) Infected head region with the area adjacent to the pharynx filled with spores (Sp). (c) The mid-posterior to tail region of N2 C. elegans infected with N. displodere from 1 dpi to 7 dpi at 15°C visualized by FISH to stain parasite rRNA (red), DAPI to stain nuclei (blue), and DY96 to stain the chitin of parasite spore walls (green). Animals were at the L2 larval stage at 1 dpi, L3 stage at 2 dpi, L4 stage at 3 dpi, and adult stage at 4–7 dpi. Sporoplasms (Sppl), meronts (Me), sporonts (Spnt), and spores (Sp) are indicated. Scale bars are 10 μm. (d) Quantification of symptoms of N. displodere infection over time at 15°C with N2 animals infected as starved L1 larvae at T0. Sporoplasms are mononucleated structures, meronts are multinucleated structures, and sporoblasts are rounded, mononucleated structures stained by FISH (see c above). Spores are oblong DY96-stained structures in infected animals. Fifty animals were quantified for each replicate at each time point, and data points indicate the mean and standard deviation (SD) from four replicates across two experiments. Each symptom was fit to a Boltzmann sigmoidal curve (R square > 0.99 for each curve), and the time to 50% of the animals exhibiting symptoms (T50) is shown.
Fig 2.
Transmission electron micrographs of N. displodere-infected C. elegans.
(a) Cross-section of an uninfected adult with the epidermis (Ep) and intestine (In) shown, separated by the pseudocoelom (Pc). The intestinal lumen (Lu) is indicated. (b) Cross-section of an N. displodere-infected adult at 6 dpi with large multinucleate meronts (Me) presumably in the epidermis, adjacent to two flanks of the body wall muscle (Mu). (c) Large associated cluster of N. displodere sporonts with nascent polar tube coils (PT) in an infected animal at 8 dpi. (d-e) Groups of nascent spores, presumably sporoblasts, in an 8 dpi animal. (f-g) Longitudinal and cross-sectional views of spores in an 8 dpi animal, with five polar tube coils. Scale bars are 1 μm (a-d) and 0.5 μm (e-g).
Fig 3.
N. displodere infects multiple tissues but shows preferential proliferation in non-intestinal tissues.
(a) The anterior region of a C. elegans animal co-infected with N. displodere (green) and N. parisii (red), visualized by FISH using species-specific rRNA probes and DAPI (blue). This image was captured by confocal microscopy with a single z-plane represented in the main inset, and orthogonal views of the x- and y-planes on the top and right insets, respectively, which show a cross-sectional view of the captured z-stacks within those planes. Scale bar is 50 μm. (b) C. elegans tissue-specific GFP-expression strains in the epidermis (top), body wall muscle (middle), and neurons (bottom), were infected with N. displodere and imaged at 3 dpi by FISH and DAPI. The neuron infected was in the ventral nerve cord (bottom). (c) Tissue-specific GFP strains were infected and imaged at 5 dpi with FISH and DY96 to stain clusters of spores (Sp). GFP-positive tissues that are difficult to see due to heavy infection are outlined with dashed white lines. The neuron infected was in the pre-anal ganglia (bottom). Scale bars are 20 μm. (d) The mid-body of the C. elegans intestinal GFP-expression strain infected with N. displodere at 1 dpi (top) and 3 dpi (bottom). Infection events are labeled as either inside (in) or outside (out) of the GFP-labeled intestine. Scale bar is 10 μm. (e) The tissue distribution of proliferating N. displodere infection was analyzed at 3 dpi, and was calculated individually in each C. elegans tissue-expression strain as the percent of FISH-stained meront clusters occurring in the GFP-positive tissues compared to the total number of events throughout the animals. Data are represented as the mean with SD of four replicates across two experiments, with a total of 50 animals counted for each replicate. (f) A comparison of the percent of animals infected in the specified GFP-positive tissue at three time points at which the three main stages of N. displodere infection occur, with sporoplasms analyzed at 1 dpi, meronts at 3 dpi, and new spores at 6 dpi. Each time point was calculated individually in each C. elegans tissue-expression strain as the percent of 50 animals that show a given symptom in the GFP-positive tissues. Data are represented as the mean with SD of four replicates across two experiments (ns = not significant, comparing intestine to muscle (p = 0.55) or intestine to epidermis (p = 0.11) at 1 dpi; *p = 0.03 comparing intestine to muscle and comparing intestine to epidermis at 6 dpi, two-tailed Mann-Whitney test).
Fig 4.
N. displodere induces an intestinal response, and host feeding is required for infection.
(a) Normalized GFP induction after N. displodere or N. parisii infection of an intestinal infection reporter strain (ERT54 F26F2.1p::GFP, left), and an epidermal infection/cuticle damage reporter (AU189 nlp-29p::GFP, col-12p::dsRed, right), as measured by a COPAS Biosort. Experimental replicates were normalized by animal body size for ERT54 or by red fluorescence (col12p::dsred) for AU189. For ERT54, data are represented as mean values with SD from n = 882 animals from six replicates across two independent experiments (****p<0.0001, two-tailed Mann-Whitney test). For AU189, due to a batch effect, only data are shown from three replicates in one experiment, with mean values shown with SD from n = 900 animals (****p<0.0001, ns = not significant, two-tailed Mann-Whitney test). Data from the other AU189 replicates are shown in the supplement (S4 Fig). (b) Comparison of N. displodere infection of daf-2(ts) animals at the L3 stage (maintained at 15°C) or daf-2(ts) animals induced to form dauer larvae (maintained at 25°C). As controls, N2 animals were maintained at 15°C and infected with N. displodere as L3 animals at 25°C, and N. parisii spores were used to infect daf-2(ts) dauer larvae. (c) Comparison of the number of invasion events (counted as sporoplasms) occurring in N2 and eat-2 animals at 1 dpi. Events were counted as either intestinal (co-localizing with intestinal gut) or non-intestinal. Data are represented as mean values with SD from n = 75 animals from three independent experiments (****p<0.0001, two-tailed Mann-Whitney test). (d) Comparison of the number of invasion events (counted as sporoplasms) occurring in dyn-1(ts) and N2 animals at 30°C for 30 minutes for N. displodere (left) and N. parisii infection (right). Infection events were distinguished as either intestinal or non-intestinal as above. dyn-1(ts) animals are paralyzed and cease to feed at the non-permissive temperature (30°C). Data are represented as mean values with SD from n = 80 dyn-1(ts) animals and n = 50 N2 animals across two independent experiments (****p<0.0001, two-tailed Mann-Whitney test).
Fig 5.
N. displodere likely accesses non-intestinal tissues from the intestinal lumen.
(a) C. elegans intestinal GFP expression strain ERT413 at 1 dpi stained with N. displodere rRNA FISH. Sporoplasms are seen inside and outside of the GFP-labeled intestine, in close proximity to the intestine, but never anterior to the posterior bulb (PB). Scale bar is 10 μm. (b) Exterior polar tubes associated with a spore were measured for N. displodere and N. parisii small spores. Each data point represents a measured polar tube, with the line and error bars showing the mean and SD of n = 40 for N. displodere and n = 41 for N. parisii. Note polar tubes of N. displodere and N. parisii were measured with separate techniques on separate occasions. The image (right) shows an N. displodere spore stained by Calcofluor white (CFW) with the associated polar tube stained by Concanavalin A-rhodamine (ConA). Scale bar is 10 μm. (c) The widths of GFP-labeled intestine from L3 larvae and young adults were measured in the anterior and posterior regions of the animal and halved to estimate the distance from the lumen to the basal lateral side of the intestine. The mean and SD from n = 14 L3 animals and n = 8 young adults are indicated (***p = 0.0002, two-tailed Mann-Whitney test). (d) The tissue distribution of invasion events of N. displodere infection (sporoplasms) was analyzed after 30 minutes of infection in L3 larvae versus young adults, and was calculated in the tissue-specific strains expressing GFP in the intestine (left) and muscle (right). Invasion events were calculated as the percent of FISH-stained sporoplams occurring outside of the GFP-expressing intestine (left) or inside the GFP-expressing muscle (right) compared to the total number of events throughout the animals. Data are represented as mean with SD of four replicates across two experiments, with a total of 25 animals counted for each replicate (*p = 0.0286, two-tailed Mann-Whitney test).
Fig 6.
N. displodere spores exit through a bursting route.
(a) Time course comparing the total number of internal spores compared to shed spores in N. displodere-infected (left) and N. parisii-infected (right) animals at 15°C. Note that only intact (non-burst) animals were picked for this assay. Internal spores indicate the average number of internal spores per animal, while external spores indicate the average number of spores shed by twenty animals into the media in four hours. Data points indicate the mean with SD of n = 6 replicates of 20 animals across 3 experiments for internal N. displodere spores and n = 4 replicates of 20 animals across 2 experiments for N. displodere shed spores and all N. parisii data. (b) Time course depicting the percent of animals with a burst vulva phenotype of uninfected, N. displodere-infected, and N. parisii-infected animals at 15°C. Data points depict the mean and SD from n = 4 independent experiments for uninfected and N. displodere and n = 3 experiments for N. parisii where each experiment consisted of triplicate samples containing at least 150 animals per replicate. (c) Image from a plate of wild-type C. elegans infected with N. displodere at 10 dpi. Indicated are adult animals with a burst vulva (BV) and internal organs spilling out. Image taken from a Nikon SMZ800 dissecting scope with an iPhone 5S. (d) Analysis of spores shed by late stage N. displodere-infected animals split into two groups, intact animals versus animals with a burst phenotype. Each data point indicates the number of spores shed by twenty animals for four hours of a single replicate, with the line and error bars showing the mean and SD of n = 4 replicates across two independent experiments (*p = 0.0211, two-tailed Mann-Whitney test; ns = not significant, p = 0.298).
Fig 7.
Analysis and comparison of N. displodere, N. parisii, and N. sp. 1 genomes.
(a) Phylogenomic tree of N. displodere and 18 other microsporidia genomes, with Rozella allomycis as an outgroup. Bootstrap support is indicated next to each node. Scale bar indicates changes per site. The tree was created with FigTree 1.4.2 (http://tree.bio.ed.ac.uk/software/figtree/). (b) Histogram of intergenic region lengths of the three Nematocida species. (c) Comparison of protein content among the three Nematocida species. Proteins were classified into 7 categories: proteins shared with all Nematocida and at least 1 other non-microsporidian eukaryotic species (eukaryotic), proteins shared between all Nematocida and at least 1 other microsporidian species (microsporidia), proteins shared only between all the three Nematocida species (Nematocida), proteins shared by N. displodere and N. parisii, proteins shared by N. displodere and N. sp. 1, proteins shared by N. parisii and N. sp. 1, and proteins not in any other species (unique). (d) Protein schematic of a generalized member of each of the large gene families in the Nematocida species, which contain signal peptides (SP). The average size of the gene family and the number of proteins in each species are indicated at the right.