Antagonistic fungal enterotoxins intersect at multiple levels with host innate immune defences

Abstract

Animals and plants need to defend themselves from pathogen attack. Their defences drive innovation in virulence mechanisms, leading to never-ending cycles of co-evolution in both hosts and pathogens. A full understanding of host immunity therefore requires examination of pathogen virulence strategies. Here, we take advantage of the well-studied innate immune system of Caenorhabditis elegans to dissect the action of two virulence factors from its natural fungal pathogen Drechmeria coniospora. We show that these two enterotoxins have strikingly different effects when expressed individually in the nematode epidermis. One is able to interfere with diverse aspects of host cell biology, altering vesicle trafficking and preventing the key STAT-like transcription factor STA-2 from activating defensive antimicrobial peptide gene expression. The second increases STA-2 levels in the nucleus, modifies the nucleolus, and, potentially as a consequence of a host surveillance mechanism, causes increased defence gene expression. Our results highlight the remarkably complex and potentially antagonistic mechanisms that come into play in the interaction between co-evolved hosts and pathogens.

Fig 1. Three enterotoxins have different expression patterns in the epidermis.

(A) Schematic overview of the plasmid insert used to express virulence factors. The candidate gene of interest (GOI), corresponding to a virulence factor without its signal peptide (ΔSigP) and stop codon was cloned between the col-19 promoter and unc-54 3’ UTR, and expressed as a fusion protein with FLAG, tobacco etch virus (TEV) protease cleavage site, degron and mKate2. (B) Schematic overview of three selected enterotoxins from D. coniospora. The signal peptide (SigP) is represented in red, the heat-labile enterotoxin alpha chain domain (PFAM: PF01375) in green, a D. coniospora specific 90-residue C-terminal domain in DcEntA (hatched blue), the DcEntB nucleolar targeting sequence (NoLS) predicted by NoD [33] in yellow, the regions similar (p = 9.44e^-3) to part of an adenylate kinase (ADK) domain (PRK13808), in brown, and DNA double-strand break repair ATPase Rad50 (PRK03918; 4.4e^-5) in dark blue for DcEntC. The remainder of the proteins’ sequences is in light blue. Representative confocal fluorescence images of young adult worm expressing DcEntA::mKate2 (IG1926; C), DcEntB::mKate2 (IG1925; D), and DcEntC::mKate2 (IG1880; E). (C) Left panel: DcEntA (arrowheads) adjacent to seam cells (asterisks). Middle: DcEntA (red) accumulation around nuclei (blue) is highlighted with arrows in the right panel showing a magnified view of the boxed area. Scale bars, left and middle, 20 μm, right, 5 μm. (D) Left panel: young adult IG1925 worm expressing DcEntB::mKate2. Scale bar, 20 μm. Right: Image of young adult IG1984 worm expressing DcEntB::mKate2, FIB-1::GFP, and BFP-NLS (a: red channel; b: green channel; c: overlay red, green and blue channels). Scale bar, 5 μm. (E) DcEntC::mKate2 appears as a punctate cytoplasmic pattern. Scale bar, 20 μm.

Fig 2. Expression of DcEntA or DcEntB makes worms sick and die precociously.

(A) Representative images of control and enterotoxin-expressing worms on the third day of adulthood. From the top: control, DcEntA-, DcEntB- and DcEntC-expressing worms (JDW141, IG1926, IG1925 and IG1880, respectively); scale bar, 100 μm. (B) Lifespan counted from the L4 stage at 25°C of worms of these 4 strains. For each strain, n = 50. **** p < 0.0001, one-sided log rank test. (C) Expression of DcEntA increases cuticle fragility in 2- and 3-day old adult worms. Tukey boxplots (n > 20, for each condition); unpaired t test, ** p < 0.01; **** p < 0.0001. (D) Survival of worms carrying frIs7 [nlp-29p::GFP and col-12p::dsRed] and a hygR transgene (control; IG1864) or also expressing DcEntA (IG1942) or DcEntB (IG1941) after infection as young adults with D. coniospora at 25°C (n = 91, 89 and 89 respectively). **** p < 0.0001, one-sided log rank test. Representative of 3 independent biological replicates.

Fig 3. DcEntA blocks AMP gene expression after infection.

(A) Representative images of DcEntC;frIs7 (IG1883; upper panels) and DcEntA;frIs7 (IG1942; lower panels) worms either not infected (left), or 20 h after infection with D. coniospora (right) as young adults. frIs7 includes nlp-29p::GFP and col-12p::dsRed transgenes; red and green florescence is visualized simultaneously. Scale bar, 200 μm. (B) Quantification of relative green fluorescence in worms carrying frIs7 and hygR transgenes (control; IG1864) or also expressing DcEntA (IG1942) or DcEntC (IG1883) either not infected (NI) or 20 h after infection (I) as young adults with D. coniospora at 25°C. **** p < 0.0001; ns, not significant, one way ANOVA test, n > 80. (C) Quantitative RT-PCR analysis of the expression of nlp and cnc genes in worms carrying frIs7 and hygR transgenes (control; IG1864) or also expressing DcEntA (IG1942) after 18h of infection by D. coniospora. Results from 2 independent experiments are shown relative to the expression levels in age-matched uninfected worms. (D-H) DcEntA acts parallel to or downstream of gpa-12 and pmk-1 to block AMP gene expression. (D) D. coniospora infection activates a signal transduction pathway that, via the Gα protein GPA-12, the p38 MAPK PMK-1 and the STAT-transcription factor-like protein STA-2, positively regulates the expression of AMP genes of the nlp family. The expression of cnc family AMP genes is also induced, but via a distinct pmk-1-independent pathway [40] that converges on sta-2 [48]. In the absence of infection, expression of a constitutively active Gα protein (GPA-12*) or uncleavable p38 MAPK (PMK-1*) leads to higher expression of AMP genes of both the nlp and cnc families. (E, G) Representative images of young adult worms carrying frIs7, expressing GPA-12* (E) or PMK-1* (G) and expressing (right, IG1948 and IG1963) or not (left, IG1389 and BPW24) DcEntA. Red and green florescence is visualized simultaneously. Scale bar, 200 μm. (F, H) Quantitative RT-PCR analysis of the expression of nlp and cnc genes in worms expressing a constitutively active Gα protein (GPA-12* (IG1389), F) or uncleavable p38 MAPK (PMK-1* (BPW24), H) and worms also expressing DcEntA (DcEntA;GPA-12* (IG1948), F; DcEntA;PMK-1* (IG1963), H). Results from 2 independent experiments are shown relative to the expression levels in age-matched IG1864 worms.

Fig 4. DcEntA alters the localisation of key immune regulatory proteins.

(A, B) DcEntA disrupts the vesicular pattern of SNF-12 in the epidermis. Confocal images of adult worms expressing SNF-12::GFP together with DcEntA::mKate2 (IG1998, right-hand panels, in order, green, red, and green and red channels together) and their siblings without the DcEntA transgene (left panel, green channel only). (A) The normal vesicular expression of SNF-12 in the lateral epidermis (lat) is disrupted in young adult worms expressing DcEntA and an accumulation of green fluorescence (arrow) is observed at the junction with the seam cell (sc). (B) When DcEntA expression is low in very young adults (upper panels), some of the SNF-12 vesicular pattern is retained. In older worms, when DcEntA is more highly expressed (lower panels), the SNF-12 pattern is diffuse in the lateral epidermis (lat) and underneath the muscle (ventral) and some aggregates can be observed to colocalise with DcEntA (arrow). Scale bar 10 μm, (*, vulva), n > 10. (C-E) DcEntA decreases STA-2 nuclear accumulation. Confocal images of young adult worms expressing STA-2::GFP with DcEntA::mKate2 (IG1971, D; left panel green channel only, right panel green and red channels shown together) and their siblings without the DcEntA transgene (C). STA-2::GFP levels are reduced in the epidermal nucleus (white box, enlarged insert; highlighted by white oval) in the presence of DcEntA. Scale bar, 20 μm. (E) Intensity of green fluorescence in the nuclei of DcEntA;STA-2::GFP worms (DcEntA, IG1971, brown) and their siblings without the DcEntA transgene (Control, blue), plotted against nuclear size, measured using Fiji. Each dot represents a nucleus; n > 20. The difference between the 2 populations is significantly different (**** p < 0.0001; unpaired t-test).

Fig 5. DcEntA induces expression of ifas-1, a target of negative regulation by sta-2, and inhibits translation.

(A) Quantitative RT-PCR analysis of the expression of ifas-1 and nlp-34 in control (hygR;frIs7 IG1864) worms and worms expressing GPA-12* (IG1389) following RNAi against sta-1 or sta-2. Data from three independent experiments are shown. (B) Quantitative RT-PCR analysis of the expression of ifas-1, nlp-34 and cnc-2 following RNAi against sta-1 or sta-2 in worms infected for 18h (Infec) and non-infected (NI) controls in the epidermis-specific RNAi strain IG1502. Data from two independent experiments are shown. (C, D) Quantitative RT-PCR analysis of the expression of ifas-1 (C) and irg-1 (D) genes in young adult worms carrying hygR and frIs7 with (IG1942) or without DcEntA (control; IG1864). (E, F) Quantitative RT-PCR analysis of the expression of ifas-1 (E) and irg-1 (F) in control worms (carrying frIs7 [nlp-29p::GFP; col-12p::dsRed]; IG274) or following exposure to cyclohexamide (CHX) for 6 h. (G) Representative fluorescence images of adult worms carrying an atf-4p(uORF)::GFP (Green) reporter also expressing DcEntA::mKate2 (Red, IG2044; bottom panels) and their siblings without the DcEntA transgene (upper panels). Scale bar, 200 μm. (H) Relative green fluorescence plotted against red fluorescence in the mixed progeny of IG2044 worms. Each dot represents a single worm (left panel). As the DcEntA transgene is not integrated, the population contains worms expressing or not DcEntA. They can be distinguished on the basis of the red fluorescence associated with DcEntA. Right panel: the proportion of the worms with a high (> 150 a.u.) green fluorescence among those with a low (<50, population I, n = 892) or high (>50, population II, n = 102) level of red fluorescence. In population I, that corresponds to worms that do not express DcEntA, only 0.78% have Green >150, while in population II, 44.12% do. (A-F) paired one-sided t test, * p < 0.05; ** p < 0.01; *** p < 0.001. (A) and (B) the fold-change in expression level between the 2 indicated conditions is significantly different.

Fig 6. DcEntA interacts with many candidate host proteins and disrupts endocytosis and the host cytoskeleton.

(A) The relative abundance of proteins co-precipitated with DcEntA::FLAG::Degron::mKate2 was assessed by mass spectrometry. Volcano plot showing specific interaction partners (in red) of DcEntA::FLAG::Degron::mKate2 (DcEntA in blue). The mean values for fold change from 3 independent experiments are shown. The SAM (significance analysis of microarrays) algorithm was used to evaluate the enrichment of the detected proteins. Proteins that met the combined enrichment threshold (hyperbolic curves, t₀ = 2) are coloured in red. The 2 C. elegans ARF proteins are shown in brown, known members of the nuclear pore complex (NPP) are in green, eukaryotic initiation factor proteins (eIF) in purple and proteins corresponding to Nipi (No Induction of Peptide after Drechmeria Infection) genes in orange. (B) Confocal images of young adult worms expressing RAB-5::GFP together with DcEntA::mKate2 (IG2022, panels, in order, green, red, and green and red channels together) in the epidermis. Selected instances of DcEntA’s colocalisation with RAB-5 are highlighted with arrows. Scale bar 10 μm in the upper panel, 5 μm in the lower panel. (C) Confocal images of young adult worms expressing RAB-5::GFP together with DcEntA::mKate2 (IG2022, right-hand panels, in order, green, red, and green and red channels together) in the epidermis, and their siblings without the DcEntA transgene (left panel). When DcEntA expression is higher, some RAB-5 signal becomes diffuse in the cytoplasm of the lateral epidermis (lat), colocalising with DcEntA (arrow). (D) Confocal images of young adult worms expressing ERM-1::mNG together with DcEntA::mKate2 (IG2051, right-hand panels, in order, green, red, and both channels together) in the epidermis, and their siblings without the DcEntA transgene (left panels). When DcEntA expression is low (upper panels), some ERM-1 is still observed in a fibre pattern, as well as with a localisation at the junction with the seam cell, co-localised with DcEntA (arrow). When DcEntA expression is higher, (lower panels), the ERM-1 pattern becomes diffuse at the seam cell (sc) boundary and some aggregates can be observed in the cytoplasm, co-localised with DcEntA. Scale bar 10 μm, n> 10. (E) DcEntA disrupts the actin cytoskeleton. Confocal images of young adult worms expressing Lifeact::GFP together with DcEntA::mKate2 (IG2024, right-hand panels, in order, green, red, and green and red channels together) in the epidermis, and their siblings without the DcEntA transgene (left panel). Scale bar 10 μm.

Fig 7. DcEntB makes nucleoli irregular and larger.

(A) Representative Nomarski image of a young adult worm expressing DcEntB::mKate2. Scale bar, 10 μm. Enlarged views of the boxed regions show the large and irregular nucleoli (white arrows). (B) Representative confocal image of a young adult worm expressing DcEntB::mKate2 and STA-2::GFP (IG1977) in hyp7. White arrow points to an irregular nucleolus (red). Scale bar, 10 μm. (C, D) Quantification of nuclear size (C) and nucleolus/nucleus ratio (D) in the epidermis of young adult IG1977 worms (DcEntB) and their siblings without the DcEntB transgene (Control). n > 20 in each strain; bars represent the mean; * p < 0.05, unpaired t-test **** p < 0.0001, unpaired t-test. (E) Quantification of nucleolus area and Feret’s diameter in hyp7 of young adult worms expressing FIB-1::GFP together with (DcEntB; IG1984) or without (Control; IG1596) DcEntB. **** p < 0.0001. Statistical significance was determined using a nonparametric Mann Whitney test. (F) Confocal images of hyp7 nuclei in young adult worms expressing FIB-1::GFP together with (DcEntB; IG1984; lower panels) or without (Control; IG1596; upper panels) DcEntB. All worms also express BFP-NLS; panels from left to right: green, blue, red, and the 3 channels together, scale bar, 5 μm.

Fig 8. DcEntB induces AMP gene expression in a sta-2-dependent manner.

(A, B) Representative images of adult worms, 3 days after the L4 stage, carrying frIs7 and expressing DcEntC (IG1883) or DcEntB (IG1941) on E. coli OP50 (A) or following RNAi against sta-1 or sta-2 (B). frIs7 includes nlp-29p::GFP and col-12p::dsRed transgenes; red and green florescence is visualized simultaneously. Scale bar, 200 μm. A difference in GFP levels for worms with frIs7 between OP50 and RNAi (HT115) bacteria has been observed regardless of the genetic background. (C) Quantitative RT-PCR analysis of the expression of nlp, cnc and ifas-1 genes in worms expressing DcEntB (IG1941) following RNAi against sta-1 or sta-2. Results from 2 independent experiments are shown relative to the expression levels in age-matched control (hygR;frIs7 IG1864) worms. (D) Intensity of green fluorescence in the nuclei of DcEntB;STA-2::GFP worms (DcEntB, IG1977, green), and their siblings without the DcEntB transgene (Control, blue), plotted against the nuclear-nucleolar size, measured using ImageJ. Each dot represents a nucleus; n>20, **** p < 0.0001, unpaired t-test. (E) Survival of control (IG1864) worms and worms expressing DcEntA (IG1942) or DcEntB (IG1941) after infection as young adults with a concentration of D. coniospora spores 10 times higher than usual at 25°C (n = 92, 91 and 87 respectively). **** p < 0.0001, one-sided log rank test. The curves here are representative of 2 independent biological replicates for which the ratio of median survival (TD50) between worms expressing DcEntA or DcEntB to control worms are shown in (F).

Fig 9. Relationship between nucleolar morphology, translation and AMP gene expression.

(A, B) Quantification of nucleolus area and Feret’s diameter in young adult worms expressing FIB-1::GFP together with (DcEntB; IG1984) or without (Control; IG1596) DcEntB following RNAi against sta-1 or sta-2 (A) or in IG1596 worms following 24 h infection (B). (C) Representative confocal images of young adult IG1596 worms (that express BFP-NLS in addition to FIB-1::GFP) after exposure for 6 h to cycloheximide (CHX; lower panels) compared to control (upper panels); left and right panels: green, and green and blue channels together, respectively), scale bar, 5 μm. (D) Quantification of nucleolus area and Feret’s diameter in young adult worms expressing FIB-1::GFP (IG1596) after exposure for 6h to CHX, compared to control untreated worms. (E) Quantification of relative green fluorescence of young adult worms carrying frIs7 ([nlp-29p::GFP; col-12p::dsRed]; IG274) after exposure for 6 h to CHX, compared to control untreated worms. The results from 2 independent experiments are shown. (F) Quantitative RT-PCR analysis of the expression of nlp and cnc genes in worms carrying frIs7 (IG274) following exposure as young adults to CHX for 6 h. Results from 3 independent experiments are shown as averages with standard deviation, relative to the expression levels in age-matched control worms. Statistical significance was determined using a nonparametric Mann Whitney test; * p < 0.05, **** p < 0.0001, ns, not significant. (G) Representative images (left panels: white light, right panels: green fluorescence) of one day old adult worms carrying an atf-4p(uORF)::GFP reporter also expressing DcEntB (IG2045; bottom panels), and their siblings without the DcEntB transgene (upper panels). Scale bar, 200 μm.

Fig 10. Models of the innate immune response to D. coniospora infection and the action of DcEntA and of DcEntB.

(A) Infection by D. coniospora activates a pathway involving SNF-12 and STA-2. The translocation of STA-2 into the nucleus leads to an increase in the expression of nlp and cnc genes. An uncharacterized parallel pathway, potentially linked to surveillance of translation, activates ifas-1, which is negatively regulated by STA-2. (B) DcEntA, via its interaction with host ARF GTPases and other protein partners (brown circle), through ADP ribosylation, potentially interferes with the normal innate immune response at multiple levels. It alters SNF-12 localization, blocks STA-2 nuclear translocation and AMP gene expression. It inhibits translation, leading to an increase in ifas-1 expression, accentuated by the loss of the repressive function of STA-2. DcEntA interferes with endocytosis, disrupts the cytoskeleton and causes increased cuticle fragility, potentially further increasing susceptibility to infection. (C) When DcEntB is expressed in hyp7, it localises to the nucleolus. Expression of DcEntB drives STA-2 into the nucleus, leading to an increase in nlp and cnc gene expression and suppressing ifas-1 expression. DcEntB alters the shape and size of the epidermal nucleoli. This provokes a surveillance mechanism leading to the expression of nlp but not cnc genes.