Caenorhabditis elegans as an Infection Model for Pathogenic Mold and Dimorphic Fungi: Applications and Challenges

The threat burden from pathogenic fungi is universal and increasing with alarming high mortality and morbidity rates from invasive fungal infections. Understanding the virulence factors of these fungi, screening effective antifungal agents and exploring appropriate treatment approaches in in vivo modeling organisms are vital research projects for controlling mycoses. Caenorhabditis elegans has been proven to be a valuable tool in studies of most clinically relevant dimorphic fungi, helping to identify a number of virulence factors and immune-regulators and screen effective antifungal agents without cytotoxic effects. However, little has been achieved and reported with regard to pathogenic filamentous fungi (molds) in the nematode model. In this review, we have summarized the enormous breakthrough of applying a C. elegans infection model for dimorphic fungi studies and the very few reports for filamentous fungi. We have also identified and discussed the challenges in C. elegans-mold modeling applications as well as the possible approaches to conquer these challenges from our practical knowledge in C. elegans-Aspergillus fumigatus model.


INTRODUCTION
Pathogenic fungi pose an enormous global threat to humanity, leading to millions of deaths and substantial financial losses annually (Fisher et al., 2012;Rhodes, 2019). Morbidity and mortality rates from opportunistic fungal pathogens, such as Candida albicans, Aspergillus fumigatus, and Cryptococcus neoformans, have been increasing for some years, especially in immunocompromised patients (Pal, 2017;Linder et al., 2019;de Sousa-Neto et al., 2020). Addressing the pathogenesis of these fungal pathogens and finding controllable strategies are crucial and urgent. To tackle this threat, model organisms are required to conduct research focusing on the identification of virulence factors, screening of effective antifungal agents, and exploring appropriate treatment approaches.
C. elegans is a microscopic multicellular nematode that lives freely in soil (Muhammed et al., 2012;Kim et al., 2017). Advantages, such as short life cycle, physiological simplicity, transparent body, complete sequenced genome, mature genetic manipulation system, and no requirement for ethical license, have greatly encouraged the wide adoption of this nematode as a model organism in scientific research with assorted applications across several research fields (Okoli et al., 2009;Ballestriero et al., 2010;Huang et al., 2014;Jiang and Wang, 2018). Some of these applications have been established for decades now whereas others are still in their nascent stages undergoing several studies. Nematode infection by the natural nematophagous obligate filamentous fungus Drechmeria coniospora is a common incidence in nature. C. elegans is usually applied for studying the innate immunity of nematodes to this fungus (Engelmann et al., 2011;Couillault et al., 2012;Zugasti et al., 2016). This nematode model has also been explored as an in vivo model for studying infections of human pathogenic filamentous fungi (Okoli and Bignell, 2015;Ahamefule et al., 2020a).
Application of the nematode model for dimorphic pathogenic fungi studies has resulted in numerous publications whereas only a few publications thus far have been recorded for human filamentous pathogenic fungi studies, such as A. fumigatus (Okoli and Bignell, 2015;Ahamefule et al., 2020a;Eldesouky et al., 2020a). Here, we have extensively portrayed C. elegansdimorphic fungi (in particular Candida spp.) infection models for determining virulence factors (reported within the last decade) and evaluated the effectiveness of anticandidal agents, including drugs, bioactive compounds, and live biotherapeutic products (reported within the last 5 years). The practical challenges constraining the applications of the C. elegans model for filamentous fungi are elaborated, and possible solutions are raised for future improvement.
Virulence factors of C. albicans such as genes involved in hyphal filamentation and biofilm formation (Romanowski et al., 2012;Sun et al., 2015;Holt et al., 2017), intestinal adhesion and colonization (Rane et al., 2014a;Muthamil et al., 2018;Priya and Pandian, 2020), important virulence enzymes (Ortega-Riveros et al., 2017;Song et al., 2019), transcription factors (Jain et al., 2013;Hans et al., 2019a), and environmental and nutrient factors (Hammond et al., 2013;Lopes et al., 2018;Hans et al., 2019b;Wong et al., 2019) have been identified in a C. elegans model to strengthen our understanding of the in vivo pathogenesis of this important fungal pathogen ( Table 1). The virulence traits of some other non-albicans species (both dimorphic and nondimorphic) have also been investigated with this nematode model (Table 1). Similarly, virulence factors such as pigmentation and hyphal filamentation have been demonstrated to be critical pathogenic features of T. marneffei in a C. elegans infection model (Huang et al., 2014;Sangkanu et al., 2021). C. elegans glp-4; sek-1 worms have mostly been used in these studies (aside from the wildtype strain, N2) because of their inability to produce progeny at 25°C due to the glp-4 mutation and their susceptibility to pathogens due to sek-1 mutation, thus making the worms immunocompromised for infection by opportunistic human fungi (Huang et al., 2014;Okoli and Bignell, 2015;Ahamefule et al., 2020a) Moreover, the adoption of a C. elegans model for searching and screening of effective bioactive compounds against several species of Candida has also received much attention. Effective bioactive compounds from marine habitats (Subramenium et al., 2017;Ganesh Kumar et al., 2019), plant parts (Shu et al., 2016;Pedroso et al., 2019), and other sources ( Table 2) have been discovered because of their in vivo efficacies against several Candida species and were simultaneously evaluated for their cytotoxicity in a C. elegans model. Compounds such as alizarin, chrysazin, sesquiterpene, and purpurin were discovered to be quite effective in in vivo assays with effective doses ranging from 1 to 10 µg/ml ( Table 2), indicating potential future prospects for antifungal drug research and discovery. Other compounds such as thymol (Shu et al., 2016), coumarin (Xu et al., 2019), and theophylline (Singh et al., 2020), were only effective at high concentrations of 64, 2, and 1.6 mg/ml, respectively ( Table 2). Most of these compounds were certified as nontoxic at such effective concentrations as they were able to rescue infected nematodes and significantly elongated their lifespan ( Table 2). The drug resistance threat of Candida species, similar to most other pathogens, is constantly increasing, leading to increased incidences of mortality and morbidity (Sanguinetti et al., 2015;Popp et al., 2017;Popp et al., 2019;Prasad et al., 2019). C. elegans has also proven to be an effective in vivo model for studying the infection of several azole-resistant C. albicans Sun et al., 2018) and C. auris (Eldesouky et al., 2018a;Eldesouky et al., 2018b) strains. Studies have demonstrated the in vivo efficacy of some bioactive compounds applied singly or in combination with initially resistant antifungal drugs in the treatment of infected nematodes ( Table 3).
The search for alternative treatment drugs with new inhibition mechanisms against pathogenic fungi such as C. albicans is a pressing need. Obtaining effective compounds that may not necessarily have a direct effect on Candida planktonic cells but affect critical virulence factors has recently been made possible by evaluating the efficacy of the compounds in a C. elegans infection model (Graham et al., 2017;Subramenium et al., 2017;Manoharan et al., 2018) (Table 4).
These compounds usually rescue infected nematodes through other pathways such as direct effects on cardinal virulence factors and/or by stimulating/enhancing the immune responses of the host against pathogens (Okoli et al., 2009;Peterson and Pukkila-Worley, 2018;Ahamefule et al., 2020b). Such compounds may only be screened and identified through in vivo assays since they usually show little or no antimicrobial activities in in vitro assays. The adoption of simple in vivo models such as C. elegans significantly supports the screening and identification of more such compounds, which may expand the narrative of the usual antifungal therapies that primarily address direct effects on causative pathogens.
The application of live biotherapeutic products (LBPs) consisting mainly of probiotics is another alternative approach for the treatment of nematode candidiasis. Such alternative Priya and Pandian (2020) C. albicans, C. glabrata, and C. tropicalis     C. albicans N2 L4/young adult worms Theophylline (THP) 1,600 THP gave over 50% more survival rate of infected worms than the untreated infected control after 6 days postinfection. THP was able to drastically lower the persistence of C. albicans in nematode gut. Additionally, THP did not show any toxicity at 1.6 mg/ml compared with untreated control for 6 days of treatment   Piperine treatment helped worms to combat infection in a dose-dependent manner leading to a significant (p < 0.05) reduction in C. albicans load. Piperine did not result in cytotoxity at sub-BIC, BIC, and 2 × BIC in worms. Piperine in vivo efficacy was mainly through hindering C. albicans colonization in nematode intestine by downregulating some important hyphae-specific genes but not affecting the growth and metabolism of the pathogen Priya and Pandian (2020) C. albicans, C. glabrata, and C. tropicalis Camphor and fenchyl alcohol 0.01% Treatment of infected worms with camphor and fenchyl alcohol significantly (p < 0.05) increased the survival rates of infected worms to >70% and >50%, respectively, compared with 5% untreated control. These compounds had no effect on worm survival and viability at concentrations of 0.05% and 0.1% in 4 days, but they became significantly toxic (p < 0.05) at 0.5%. Camphor and fenchyl alcohol at BIC (approximately 50 times the MIC) had effect on C. albicans biofilm and hyphal filamentation but not on the planktonic cells Manoharan et al. (2017b) C. albicans L4 worms 5-Hydroxymethyl-2furaldehyde (5HM2F) MBIC (400) Increased survival time of infected worms when treated with 5HM2F (120 h) compared with 96 h of control group. 5HM2F displayed no cytotoxic effect on worms by120 h. 5HM2F below 500 µg/ml does not have antifungal effect on C. albicans except on some virulence factors such as biofilm formation, morphological transition, and production of secreted hydrolases Subramenium et al. (2017) a BICs for C. albicans, C. glabrata, and C. tropicalis in combination with QA/UDA were 100/5, 100/10, and 200/20 µg/ml, respectively. BIC, biofilm inhibition concentration; MBIC, minimum biofilm inhibitory concentration.
therapy is an interesting and promising option since pathogenic fungi are currently developing resistance to the few clinically available antifungal drugs (Sanguinetti et al., 2015;Prasad et al., 2019). Several species of Lactobacillus such as L. rhamnosus (Poupet et al., 2019a;Poupet et al., 2019b) and L. paracasei (de Barros et al., 2018) as well as probiotic yeasts-Saccharomyces cerevisiae and Issatchenkia occidentalis (Kunyeit et al., 2019)have demonstrated efficient rescue of worms infected with a number of Candida species. These therapeutic microorganisms drastically reduced the burden of the pathogens in the C. elegans intestine approximately 2 to 4 h postinfection treatment ( Table 5).
The efficacy of these LBPs in reducing and/or eliminating fungal burden implies the future potential of LBPs in addressing the fungal menace. The demonstrated significant increase (p < 2 × 10 −16 ) in worm mean lifespan (Poupet et al., 2019a;Poupet et al., 2019b) is so high that it has not been reported in any potent bioactive compounds or even established antifungal drugs. The fact that most of these LBPs are already established probiotics is yet another important parameter that would advance future research beyond nematode models.
Given that decades of searching for new antifungal agents have not truly resulted in new antifungal drugs, drug repurposing is a less expensive and welcome research prospect. The C. elegans infection model for evaluating the efficacy of repurposed drugs on candidiasis has attracted attention (Eldesouky et al., 2020b;Singh et al., 2020) (Table 6) due to the advantages of saving extensive time, cumbersome labor, and enormous cost of searching and obtaining new antifungal drugs.

C. elegans and Pathogenic Molds
The deadly opportunistic mold pathogen, A. fumigatus, ranks as the number 1 aetiological agent for aspergilloses in immunocompromised patients (Snelders et al., 2009;Fang and Latge, 2018;Geißel et al., 2018) with an almost 100% mortality rate in some groups of patients (Darling and Milder, 2018;Geißel et al., 2018;Linder et al., 2019). This pathogen had not been well studied in C. elegans until recently. Okoli and Bignell (2015) were the first to demonstrate the possibility of adopting C. elegans for A. fumigatus infection. They set up the nematode model to study the pathogenicity of the clinical strain A. fumigatus Af293 for 72 h postinfection after an initial preinfection of 12 h. We recently reported a breakthrough in overcoming some of the challenges usually encountered in the C. elegans-mold infection system, one of which is removing spores that were not ingested by worms through a hand-made filter with a membrane-attached-on-tube. We were able to develop a stable and consistent C. elegans model for evaluating the virulence of A. fumigatus mutant strains that had previously been studied in other established models, including mice and insects. We also successfully demonstrated the possibility of in vivo testing of antifungal agents on nematode aspergillosis using the established model (Ahamefule et al., 2020a).
The established C. elegans-A. fumigatus model clearly demonstrated the progression of aspergillosis infection in nematodes using the A. fumigatus fluorescence strain, Af293-dsRed, showing that hyphal filamentation could actually emanate from any part of the infected worms against the previously reported concept of mainly the tail region (Okoli and Bignell, 2015;Ahamefule et al., 2020a). Our worm model was able to identify important virulence factors of A. fumigatus such as a-(1,3)-glucan synthase, melanin pigmentation, iron transporter, Zn2Cys6-type transcription factor, and mitochondrial thiamine pyrophosphate transporter, as mutant strains without these components (triple agsD, pksPD, DmrsA, DleuB, and DtptA, respectively), all of which gave significantly attenuated virulence compared with the A. fumigatus parent strain KU80D. These reduced virulence patterns obtained by our C. elegans model were similar to previously reported attenuated virulence patterns of these A. fumigatus mutants in both vertebrate and insect models. The nematode model was also demonstrated to be an easy in vivo system to evaluate antifungal drug efficacy thus presenting the model as a desired platform for screening antifungal agents against A. fumigatus in the future (Ahamefule et al., 2020a).

CHALLENGES OF C. elegans APPLICATIONS IN MODELING PATHOGENIC MOLD
One of the biggest challenges usually encountered in the applications of the C. elegans model for filamentous fungal infection is the difficulty in infecting the worms through conidia. Worms usually avoid eating conidia unless they starve with no other option (Okoli and Bignell, 2015). This avoidance is unlike the case of dimorphic fungal and bacterial pathogens, where infection is never much of a problem as worms easily feed on the cells of these pathogens when they replace or are mixed up with nematode choice food (E. coli OP50 or HB101) (Breger et al., 2007;Johnson et al., 2009;Kirienko et al., 2013;Okoli and Bignell, 2015).
Giving the worms more time to starve and more access to the conidia (placed at four cardinal points) for ingestion is very important for establishing mold preinfection assays. Okoli and Bignell (2015) adopted a 12-h preinfection technique, while we modified to 16 h (Ahamefule et al., 2020a). The fact is that worms must be given such ample time to "force" them to ingest the mold conidia in a preinfection system since coinfection approach (which is usually adopted for most dimorphic fungi modeling) cannot work well for mold pathogens (Okoli et al., 2009;Okoli and Bignell, 2015;Ahamefule et al., 2020a). As conidia germinate very fast even before the worms have ingested enough spores in killing assay medium, a relatively less nutritious medium was adopted for pre-infection assay to avoid the quick growth and flooding of hyphal filaments in the rich killing assay medium (brain heart infusion medium); otherwise later experimental procedures will be severely limited (Okoli and Bignell, 2015;Ahamefule et al., 2020a).
Another challenging aspect in setting up the C. elegans-mold model is the separation of noningested conidia from worms after pre-infection stage. Failure at this stage leads to the germination of unseparated spores in killing or antifungal screening media thus obstructing experimental progress. Although our designed membrane-attached-on-tube filter (with a 35-µm pore diameter) was able to remove a great deal of noningested conidia, the separation was not 100% efficient. Modifying the membrane pore size to an appropriate diameter should help improve the filtration efficiency by allowing faster and better removal of conidia while keeping the preinfected L4/ young adult worms (Figure 1). Even though the separation efficiency of noningested spores becomes 100% or close to it, hyphae growth in killing medium would still not be completely eliminated, particularly if the experiment is scheduled to go beyond 72 h postinfection. This is because we have discovered that some conidia could be egested out of the nematode intestine into the killing medium and still retain their viability of germinating to hyphae, which is a big challenge to tackle and severely affect the experiment.
Hyphal filamentation usually occurs in infected worms. Unlike most studied dimorphic fungi whose external hyphae protrude when worms were already dead (and could therefore be easily transferred), numerous worms infected with filamentous fungi such as A. fumigatus (Ahamefule et al., 2020a), A. flavus, and some strains of Penicillium (that we have studied in our laboratory), were discovered to still be alive with protruded hyphae. This makes these worms stuck to the killing assay plates and therefore difficult to remove (Ahamefule et al., 2020a). Such filamentation usually becomes profuse, growing and spreading very fast and may eventually obstruct visibility and affect the experimental results. Regulating the number of immunocompromised worms in killing assays, especially for highly virulent pathogenic molds, is an option to ameliorate this menace (Figure 1).

Caenorhabditis elegans as an Infection Model for Fungi
Frontiers in Cellular and Infection Microbiology | www.frontiersin.org October 2021 | Volume 11 | Article 751947 meaningful not only for dimorphic fungi but also for mold pathogens. Our review of the breakthrough applications of C. elegans for dimorphic fungi studies and progress/ modifications of the C. elegans-mold infection model will provide a reference for studying fungal infections and developing antifungal agents.

AUTHOR CONTRIBUTIONS
CA and BE wrote the initial manuscript. JO, AM, AI, BW, CJ, and WF revised the manuscript. WF supervised the manuscript.

A B C
FIGURE 1 | Modifications of the preinfection to killing assays for the C. elegans-mold infection model. (A) The previously described procedure (Okoli and Bignell, 2015).