The role of mycelium production and a MAPK-mediated immune response in the C. elegans-Fusarium model system

Abstract

Fusariosis is an emerging infectious complication of immune deficiency, but models to study this infection are lacking. The use of the soil nematode Caenorhabditis elegans as a model host to study the pathogenesis of Fusarium spp. was investigated. We observed that Fusarium conidia consumed by C. elegans can cause a lethal infection and result in more than 90% killing of the host within 120 hours, and the nematode had a significantly longer survival when challenged with Fusarium proliferatum compared to other species. Interestingly, mycelium production appears to be a major contributor in nematode killing in this model system, and C. elegans mutant strains with the immune response genes, tir-1 (encoding a protein containing a TIR domain that functions upstream of PMK-1) and pmk-1 (the homolog of the mammalian p38 MAPK) lived significantly shorter when challenged with Fusarium compared to the wild type strain. Furthermore, we used the C. elegans model to assess the efficacy and toxicity of various compounds against Fusarium. We demonstrated that amphotericin B, voriconazole, mancozeb, and phenyl mercury acetate significantly prolonged the survival of Fusarium-infected C. elegans, although mancozeb was toxic at higher concentrations. In conclusion, we describe a new model system for the study of Fusarium pathogenesis and evolutionarily preserved host responses to this important fungal pathogen.

Introduction

Fusarium spp. are filamentous fungi widely distributed in nature within environmental niches that include the water and soil, and are capable of causing infections in both plants and humans. In humans, Fusarium spp. are involved in superficial, locally invasive, and disseminated infections [1,2]. The most common etiologic agent among Fusarium spp. is Fusarium solani followed by F. oxysporum, F. moniliforme and F. verticillioides [2]. Immunocompromised patients, such as those with transplants, hematological malignancies, and those using immunosuppressant agents, are more prone to disseminated infections [2–4]. The prognosis of systemic fusariosis is poor and the response to treatment is not favorable [5–7]. Due to the lack of formal clinical trials regarding fusariosis treatment, the recommendations for treatment are based mainly on expert opinion and case series [8–12].

A host-pathogen interaction model may elucidate new information about Fusarium pathogenesis and the corresponding host immune response. Caenorhabditis elegans, a free-living soil nematode, is a widely used invertebrate model in the in vivo study of host-pathogen interactions and antifungal agent discovery [13–17]. In the mycology field, C. elegans has proven to be a useful tool to study the pathogenesis of Candida spp. and Cryptococcus neoformans infections [18–21]. C. elegans is an amenable model because the nematode has a short reproductive life cycle, produces genetically identical progeny, is inexpensive, is ethically acceptable, its genome has been annotated, and it has been reported that the virulence factors of multiple pathogens in C. elegans correlate to those in mammals [17]. C. elegans has an innate immune response mechanism that is activated against different microbes [22–24], involving several pathways such as the DBL-1 and the p38 mitogen-activated protein kinase (MAPK) pathways. Therefore, it can serve as a basis to explore the innate immune response in higher organisms including mammals [22,25–27]. In the latter, the Toll like receptors (TLRs) are important components of the innate immune response against pathogens that contain an intracellular Toll/IL1 resistance (TIR) domain [28–30]. Sensing microbial pathogen-associated molecular patterns through TLRs leads to recruitment of adaptor proteins, such as MyD88, to form a complex that activates signaling pathways, including the p38 MAPK cascade which plays an important role in pathogen resistance [31,32]. The TIR domain is evolutionarily conserved in C. elegans and mammals and in C. elegans, the gene tir-1 (encoding the TIR domain) functions upstream of pmk-1 (the homologue of the mammalian p38 MAPK) [33,34].

This work illustrates how Fusarium germination and mycelium production can kill the host (C. elegans) and demonstrates the contribution of two evolutionarily conserved entities, the TIR domain and a MAPK, in the immune response of C. elegans against Fusarium. We also demonstrate the use of this model for the study antifungal compounds against Fusarium and in a proof of principle study we report the efficacy of two compounds, mancozeb and phenyl mercury acetate, against Fusarium utilizing this system. Therefore, C. elegans is a potential substitute host to study Fusarium infection, Fusarium-host interactions, and the efficacy of antifungal agents.

Materials and methods

Strains and media

Isolates of F. solani (FsCI-1), F. oxysporum (FoCI-1), and F. proliferatum (FpCI-1) were obtained from the mycology laboratory at Massachusetts General Hospital (sources are from cultures of blood, tissue, and bronchial wash, respectively). The strains were maintained on Potato Dextrose Agar (PDA) plates containing 45 μg/ml kanamycin, 100 μg/ml ampicillin, and 100 μg/ml streptomycin at room temperature. Microconidia were the only conidial type used in this work. The C. elegans strains were the wild and mutant strains N2 (wild type), cnc-2 (ok3226), dbl-1 (nk3), tir-1 (qd2), fer15 (b26); fem1 (hc17), and pmk-1 (km25). All the strains were obtained from the Caenorhabditis Genetics Center (CGC) which is funded by the NIH National Center for Research Resources (NCRR) except the tir-1 (qd2) strain [35]. Synchronized L4 stage C. elegans were used in all experiments.

Identification of Fusarium spp. isolates

The species of the Fusarium isolates were identified by sequencing of the translation elongation factor (TEF) 1 α coding region. PCR and the primers EF1 and EF2 were used for amplification and then the product cloned into pGEM-T Easy (Promega, Madison, WI, USA) and bi-directionally sequenced using EF1 and EF2 [36] at the MGH DNA sequencing core facilities. Isolates were further subjected to multilocus sequence typing as described previously [37–39] and MLST group was identified using the blastn analysis function at NCBI or the FUSARIUM-ID database [39].

Pre-infection and nematode killing assay

V8 agar plates (100 ml V8 juice (Campbell Soup Company), 400 ml distilled H₂O, 12 g agar) containing 45 μg/ml kanamycin, 100 μg/ml ampicillin, and 100 μg/ml streptomycin were inoculated with Fusarium and kept at room temperature for 6 days, after which the plates were washed with phosphate buffered saline (PBS), the conidia suspension was filtered through two layers of cheese cloth and centrifuged. The pellet was washed twice with PBS, and diluted in 20% Brain-Heart-Infusion (BHI) broth: i.e., 80% M9 buffer (v/v) containing 45 μg/ml kanamycin, 100 μg/ml ampicillin, 100 μg/ml streptomycin, and 5 μg/ml cholesterol.

C. elegans nematodes were incubated with 2 ml of 20% BHI broth: i.e., 80% M9 buffer (v/v) containing Fusarium conidia, 45 μg/ml kanamycin, 100 μg/ml ampicillin, 100 μg/ml streptomycin and 5 μg/ml cholesterol and maintained at 25°C for 4 h. Nematodes were washed with M9 and transferred to fresh medium (20% BHI broth: 80% M9 buffer (v/v) containing 45 μg/ml kanamycin, 100 μg/ml ampicillin, 100 μg/ml streptomycin, and 5 μg/ml cholesterol) and incubated at 25°C for survival observation. Nematodes were considered dead when they did not respond to touch. All C. elegans experiments were performed in duplicate and representative experiments are presented. For heat killed conidia preparations, Fusarium conidia were incubated in a water bath at 85°C for 60 min. Part of the cell suspension was plated on PDA plates to evaluate the viability of Fusarium conidia after heat exposure and conidia were inspected under the microscope for structural integrity.

Fungal burden assessment

Infected nematodes at a specific time point were washed with M9 and homogenized in 1 ml of PBS. Fifty μl of the homogenate was inoculated on PDA plates supplemented with 45 μg/ml kanamycin, 100 μg/ml ampicillin, and 100 μg/ml streptomycin, and incubated for 48 h at 30°C before determination of the colony forming units (CFU).

Microscopy

Nematodes were pre-infected with Fusarium for 4 h (as indicated above) and then washed multiple times with M9 buffer and transferred to fresh medium (20% BHI broth: 80% M9 buffer (v/v) containing 45 μg/ml kanamycin, 100 μg/ml ampicillin, 100 μg/ml streptomycin, and 5 μg/ml cholesterol) and incubated at 25°C. After 72 h, the worms were placed on pads of 2% agarose with 5 μl of 100 μM levamisol hydrochloride in M9 buffer. A confocal laser microscope (TCS NT, Leica Microsystems) was used for observation.

Study of antifungal agents

To study the efficacy of antifungal agents against Fusarium in the C. elegans infection model, amphotericin B, voriconazole, fluconzole, mancozeb, and phenyl mercury acetate (PMA) were diluted in dimethyl sulphoxide (DMSO) and added to the liquid assay to the target concentration. All chemicals were obtained from Sigma (St Louis, MO, USA).

Statistical analysis

Survival curves were plotted using Excel software (Microsoft), and differences in survival were calculated using the log-rank test (STATA 6), and statistical analysis of the fungal burden was calculated using t-test function in the Excel software (Microsoft). A P value of <0.05 was considered statistically significant.

Results

Killing C. elegans by Fusarium

To test whether Fusarium spp. can infect and kill C. elegans, nematodes were incubated with Fusarium conidia in liquid medium and transferred to a pathogen-free liquid medium and monitored for survival. This procedure resulted in killing of more than 90% of the nematodes within 120 h. Interestingly, C. elegans when challenged with F. proliferatum showed significantly increased survival when compared to F. solani and F. oxysporum (Fig. 1). To confirm that conidia had an impact on nematode death, nematodes were incubated with heat killed conidia in parallel with non-heat killed conidia. Incubating the nematodes with heat killed conidia caused no death of the worms (data not shown). During the infection process, the number of fungal cells recovered from each pre-infected nematode began to decrease after 4 h post-infection (72 cells/nematode) and remained constant after 24 h post-infection (13 cells/nematode) (Fig. 2). In contrast to Fusarium, challenging C. elegans with the filamentous fungus Neurospora crassa led to no killing 90 h after infection (unpublished observation by the same authors) [40].

Fig. 1.

Killing of Caenorhabditis elegans by Fusarium spp. C. elegans were pre-incubated with Fusarium conidia for 4 hours and transferred to pathogen-free medium for survival observation. Three clinical isolates (F. solani, F. oxysporum and F. proliferatum) are shown. 120 hours post infection > 90% of the nematodes were dead. (P < 0.001 between F. solani and F. proliferatum) (P < 0.01 between F. oxysporum and F. proliferatum)(P = 0.30 between F. solani and F. oxysporum). Eighty nematodes were assayed per condition.

Fig. 2.

Assessment of fungal burden within the nematode. The fungal burden decreased after 4 hours post-infection and stabilizes 24 hours post-infection. Caenorhabditis C. elegans were pre-incubated with conidia of Fusarium F. oxysporum (clinical isolate) for 4 hours and transferred to pathogen-free medium. Fifteen worms per condition were used. Each bar represents the mean of three experiments. The error bar represents the standard deviation. *P < 0.05.

It should be noted that the method in this work is modified and slightly different from the previously reported Candida albicans-C. elegans pathogenesis assay [15]. More specifically, in this work, L4 C. elegans (fer15 (b26); fem1 (hc17)) nematodes were used (compared to glp-4(bn2);sek-1 (km4) in the C. albicans-C. elegans assay) and the pre-infection took place in liquid medium (instead of solid media as reported in the original C. albicans-C. elegans assay; see also Material and methods). Of note is that a 4-h incubation of C. elegans with Fusarium conidia was sufficient to establish infection before transferring the worms to a pathogen-free medium.

Mechanism of killing

During assessment of the killing assay and scoring of the nematodes, some of the dead nematodes exhibited mycelium protruding from inside the nematode and penetrating the cuticle (Fig. 3). Alternatively, an additional population of nematodes died without visible mycelium penetration of the cuticle as viewed under the dissecting microscope. Interestingly, using higher magnification and confocal microscopy to visualize nematodes infected with F. oxysporum, we identified mycelia even within the bodies of the nematodes in which the fungus did not penetrate the cuticle (Fig. 3B). Thus, although the mycelium did not penetrate the cuticle, they were still present internally, and are probably responsible for nematode death by disrupting the internal structure and organs of the worm (Fig. 3B).

Fig. 3.

Fusarium mycelium disrupts the nematode cuticle and internal structures. Caenorhabditis C. elegans were pre-incubated with conidia of F. oxysporum (clinical isolate) for 4 hours and transferred to pathogen-free medium. (A) Nematodes not infected with Fusarium. (B) Mycelium disrupting the internal nematode structures but not protruding from the nematode cuticle (arrow head). (C) Mycelium protruding from the nematode cuticle (arrow). (D) Higher magnification of a section of the nematode shown in panel C. Images were taken 72 hours post infection.

C. elegans immune response toward Fusarium

To explore the immune response of C. elegans against Fusarium, we challenged different C. elegans mutant strains (tir-1, cnc-2, dbl-1, and pmk-1) with F. oxysporum and observed the nematode survival. The cnc-2 gene encodes an antimicrobial peptide from the Caenacin (cnc) family that plays a role in C. elegans immunity when challenged with another fungal pathogen, Drechmeria coniospora [25]. The dbl-1 gene encodes a transforming growth factor- β-like protein (TGF- β) that is involved in the immune response of C. elegans against different pathogens [25,26,41,42]. The expression of cnc genes including cnc-2 are regulated through the DBL-1 protein.

After challenging C. elegans with F. oxysporum, the differences in the survival of both dbl-1 and cnc-2 mutant strains were not significant compared to the N2 wild type strain. However, the survival of pmk-1 and tir-1 mutant strains was significantly shorter than the parental strain (N2 wild type) when challenged with F. oxysporum (Fig. 4). These results suggest the importance of both the TIR domain containing protein and the MAPK pathway in the immune response of the nematodes in response to Fusarium.

Fig. 4.

Caenorhabditis C. elegans immune response mutants are shorter lived when challenged with Fusarium oxysporum. Nematodes were pre-incubated with conidia for 4 hours and transferred to pathogen-free medium for survival observation. tir-1 and pmk-1 mutants lived significantly shorter compared to the N2 wild type nematodes (P = 0.015 between N2 and pmk-1; P = 0.0132 between N2 and tir-1). Fifty nematodes were assayed per condition.

Utilization of C. elegans as a tool to study antifungal compounds against Fusarium

To assess whether C. elegans could be used to evaluate antifungal efficacy against Fusarium spp., nematodes were pre-infected in liquid medium with F. oxysporum conidia, and transferred to a pathogen-free medium containing antifungal agents. Two clinically relevant antifungal agents were tested, i.e., amphotericin B (minimum inhibitory concentration MIC = 2 μg/ml), and voriconazole, both of which are used in Fusarium infection therapy, and fluconazole (MIC > 64 μg/ml), which has limited efficacy against Fusarium in the clinical setting. Two other compounds, mancozeb and phenyl mercury acetate (PMA), were evaluated for their efficacy against Fusarium and in vivo toxicity. Mancozeb is an antifungal agent used in agriculture, and PMA is a fungicide. Amphotericin B (Fig. 5A) or voriconazole (Fig. 5B) treated nematodes lived significantly longer than the control group (P < 0.001), while the survival of fluconazole treated nematodes was not statistically different from DMSO treated nematodes (Fig. 5A).

Fig. 5.

Assessment of antifungal efficacy of clinically relevant compounds in the Fusarium-Caenorhabditis C. elegans assay. Nematodes were pre-incubated with F. oxysporum conidia for 4 hours and transferred to liquid medium containing (A) amphotericin B (8 μ/ml) or fluconazole (32 μg/ml) (P < 0.0001 between amphotericin B and solvent alone, DMSO) or (B) voriconazole (2 μg/ml) (P < 0.001 between voriconazole and solvent alone, DMSO). Eighty nematodes were assayed per condition.

To broaden our knowledge about the in vivo efficacy of mancozeb and PMA against Fusarium and the toxicity of these compounds, nematodes were treated with increasing concentrations of mancozeb ranging from 8–128 μg/ml and with increasing concentrations of PMA ranging from 0.183–0.732 μg/ml. When the concentration of mancozeb was increased, the number of dead nematodes increased (Fig. 6). After 24 h, all the nematodes treated with 128 μg/ml of mancozeb had died. This result suggests that mancozeb could be toxic to the nematodes at higher concentrations and toxicity outweighs the antifungal effect of this compound. Treating the nematodes with increasing concentration of PMA resulted in increased survival and no difference in survival was observed after treating the nematodes with a concentration of more than 0.549 μg/ml of PMA (Fig. 7). This finding suggests that C. elegans can serve as substitute host to study the antifungal efficacy and the toxicity of potential antifungal compounds.

Fig. 6.

The efficacy and toxicity of mancozeb assessed in the Fusarium-Caenorhabditis C. elegans pathosystem. C. elegans were pre-incubated with F. oxysporum conidia for 4 hours and transferred to liquid medium containing mancozeb at the indicated concentrations. The toxicity is positively correlated to the concentration of the mancozeb in the medium. (P = 0.0011 between 8 μg/ml treatment and 32 μg/ml treatment; P < 0.001 between 8 μg/ml treatment and 128 μg/ml treatment; P = 0.0027 between 16 μg/ml treatment and 32 μg/ml treatment; P < 0.0001 between 16 μg/ml treatment and 128 μg/ml treatment; P < 0.0001 between 32 μg/ml treatment and 128 μg/ml treatment). Fifty worms per concentration were used.

Fig. 7.

Assessment of the efficacy and the toxicity of PMA using the Fusarium-Caenorhabditis elegans assay. C. elegans were pre-incubated with F oxysporum conidia for 4 hours and transferred to liquid medium containing the indicated concentration of PMA. The efficacy is positively correlated to the concentration of PMA in the medium. (P = 0.0118 between 0.183 μg/ml treatment and 0.549 μg/ml treatment). Forty worms per concentration were used.

Discussion

In this work, C. elegans was investigated as a potential substitute model host to study the pathogenic fungi of the genus Fusarium. Fusarium conidia ingested by C. elegans led to a lethal infection and resulted in more than 90% killing of the nematode within 120 h. C. elegans was able to live significantly longer after infection with F. proliferatum than F. solani or F. oxysporum. Interestingly, Fusarium germination and mycelium production appears to be a key factor responsible for C. elegans killing.

The existence of both fungi and nematodes in the same environmental niche (soil, plant tissue, and rhizosphere) provides opportunities for these organisms to interact. The key determinants of the pathogenesis of fungi could have partly evolved as a result of this existence and interaction. Moreover, the feeding habit of nematodes on fungi has been reported [43]. Interestingly, fungi can live commensally within the host and provide protection against nematodes. For example, in bananas, a naturally occurring endophytic F. oxysporum strain has an antagonist effect against the burrowing nematode Radopholus similis, conferred through production of substances by the fungus [44]. Additionally, Fusarium has the ability to produce a variety of substances and mycotoxins that have a wide range of effects on the host [45,46]. It has been shown that the culture filtrate of F. oxysporum has efficacy against the root-knot nematode Meloidogyne incognita [47]. Finally, beauvericin, a compound isolated from F. bulbicola, has been shown to have a nematicidal activity against C. elegans [44,47–49]. However, specific environmental stimuli may be required for mycotoxin production [50–52]. These examples augment the idea that the virulence factors of fungi could have evolved as a result of trial to protect themselves from nematodes or other organisms.

C. elegans has played a significant role as an infection model to study the pathogenesis of multiple microbes [17,18,20,21,53], including their virulence factors. For example, it has been confirmed that hyphal formation, a virulence factor of C. albicans, is also a key determinant of its pathogenicity of C. elegans [21]. Moreover, mycelium production by Fusarium conidia is an important part in its pathogenesis of mammals [2,54]. In the current work, internal germination of the conidia and mycelium production and penetration of the body appears to be the major mechanism and the key factor involved in nematode killing. Therefore, the infection process of Fusarium in C. elegans has similar hallmarks of the infection process seen in higher organisms.

C. elegans has also contributed in the development of the immunology field and in particular, the host immune response against fungi [55]. In our study, tir-1 and pmk-1 mutant strains lived shorter than the N2 strain after challenge with Fusarium. This result is in agreement with those previously reported that showed down regulation of tir-1 using RNA interference led to increased susceptibility of C. elegans to the fungus D. coniospora [41]. Moreover, these results highlight the importance of the TIR domain and MAPK pathway in the immunity of C. elegans against fungal pathogens. However our data cannot exclude other immune pathways involved in the response to Fusarium. The two entities are conserved in C. elegans and mammals and they have significant roles in the immune system, supporting the importance of C. elegans in exploration of the immunology field.

The most widely used and accepted antifungal agents to treat patients with fusariosis are amphotericin B and voriconazole, but these are not sufficient for the treatment in every patient [7]. Therapy failure and/or high inhibitory concentrations for both agents have been reported [7]. Using C. elegans as a host to identify new compounds with antifungal properties is a novel and promising approach [56]. In the antimicrobial discovery, researchers have been attracted to C. elegans as a whole animal tool because of several reasons including measurement of the toxicity of the compounds, it is inexpensive, ethically acceptable and C. elegans can detect compounds that have immunomodulatory activity [15,16,57]. However, the bioavailability of the compounds cannot be determined in this system, a fact that poses some limitation. In this study, the possibility of using C. elegans as a host for the in vivo assessment of antifungal agents against Fusarium has been also confirmed. The efficacy of an antifungal agent used in agriculture, mancozeb, against Fusarium was assessed, and determined a balance between the toxicity and efficacy of this compound. Additionally, PMA had efficacy against Fusarium and PMA had been previously identified as harboring antifungal properties [56,58].

It is worth mentioning that Fusarium pathogenesis has been studied using different alternative hosts such as the fruit fly Drosophila melanogaster and the larvae of the greater wax moth Galleria mellonella [59,60]. In the D. melanogaster model, Fusarium conidia are injected directly into the body of the fly, but in the C. elegans model the conidia were ingested. In addition, the antifungal agents were mixed with food in D. melanogaster model, and they were mixed with the liquid medium in the C. elegans model system. However, Fusarium can cause a lethal infection in wild type C. elegans, but in D. melanogaster model, immunosuppressed flies were necessary to establish an infection. Also G. mellonella has offered a unique system that facilitated the study of the Fusarium pathogenesis at a temperature relevant to the human temperature [59].

In conclusion, C. elegans can successfully be utilized as a model host to study Fusarium pathogenesis, Fusarium-host interaction, and to identify antifungal agents.