The global impact of opportunistic fungal infections has gone underrecognized for a long time (1). However, with the increase in chronic and immunosuppressive health conditions including HIV/AIDS, cancer, cystic fibrosis and diabetes, antimicrobial therapies and invasive procedures that leave individuals vulnerable to opportunistic infections, the impact of these infections are becoming more apparent (1, 2). Fungi cause disease through direct infection of the host or through their secondary metabolites, mycotoxins, pigments that can contaminate the environment, food products and air (3). The disease burden ranges from superficial to invasive fungal infections and is estimated to be in the 100s of millions of patients per year, resulting in >1.5 million deaths/year (2, 4). These infections are caused by long recognised pathogens such as Aspergillus fumigatus and Candida albicans (5–7), neglected tropical diseases like eumycetoma (8, 9), and newly emerged pathogens, such as Candida auris (10, 11).

With the development of advanced molecular and cellular biology technologies, fungal pathogenicity and virulence factors are being studied in greater detail (12–14). However, the fungal threat continues to grow while the development of novel effective antifungal therapies remains inadequate (15, 16). As a result, in 2022 the WHO published the WHO fungal priority pathogens list, classifying 18 medically relevant fungal species as “Critical”, “High” or “Medium” priority, according to the perceived public health burden (2, 17).

The use of model organisms is one of the technologies that has been developing over time and has become indispensable to investigating the nuances of host-pathogen interactions (18, 19). A cursory search of PubMed using the keywords “Drosophila” AND “fungi” yielded 8,617 results (1948 – 2023), with over a third (36.2%) of the publications having been released in the last decade alone. Seminal proof of concept studies in the 1990s and early 2000s, utilising wild-type and mutant Drosophila strains and fungi, provided a comprehensive framework for employing Drosophila in fungal research (20–25). Over the last decade, more extensive Drosophila-fungi work has taken place, leading to a better understanding of virulence, pathogenicity, and host immune responses (26–30).

This review sets out to provide a brief update on tools currently being applied to host-fungal interaction studies in Drosophila and highlight examples of research in the last 5 years with a focus on the WHO’s fungal priority pathogens list (17).

Drosophila is currently being utilised to investigate how medically relevant fungi interact with host immunity, and how they transition from colonization to infection (26–30). Wild-type and genetic mutant strains (e.g., Toll-deficient) are commercially available for distribution across the world from stock centres, such as the Bloomington Drosophila Stock Centre and Kyoto Stock Centre (54, 55). Table 1 summarises Drosophila strains used in fungal research studies included in the current review. The Drosophila Genomics Resource Centre and ATCC are some of the suppliers who distribute Drosophila cell lines, like Schneider’s Drosophila Line 2 (S2) cell line, GFP-tagged cells, and cells from various organs for ex vivo studies (67, 68). These fly strains and cell lines are relatively inexpensive to purchase and maintain, increasing accessibility of the model (68). The model systems are infected or exposed to fungi, fungal secondary metabolites, and antifungal compounds to investigate these interactions (26, 55) via feeding, rolling over, or co-culture and in a standardised manner via needle pricking or microinjection, allowing for rapid inoculation of experimental groups (57, 58). Infection progression can be measured through survival, microbial load, mRNA quantification, melanisation and microscopy assays (55, 64). The efficacy and toxicity of antimicrobial compound screens can be measured in similar ways to determine their efficacy and toxicity (69, 70). Examples of microscopy techniques include confocal microscopy for visualising phagocytosis in fungi-stimulated plasmatocytes (71), electron microscopy for imaging effects of treatment on host cell morphology (35) and fungal burden (72). Immunofluorescence staining and bioluminescence allow visualisation of individual cell types in tissue, larvae or adult flies, and can be done through RNA in situ hybridization (30, 73, 74), intravital 2-d photon microscopy and reporter systems (GFP, lacZ) (62).

The development of molecular techniques including DNA and RNA sequencing, RNAi gene silencing and CRISPR/Cas9 has advanced the field in leaps and bounds. Molecular techniques have made it possible to sequence the Drosophila genome (75); sequence coding and non-coding RNA and determine functionality through RNAi-based screening assays and gene silencing or overexpression (66, 76, 77). They facilitate quantification of messenger RNA (mRNA) or transfer RNA (tRNA) through quantitative PCR, reverse transcriptase PCR and modified-induced misincorporation tRNA sequencing (mim-tRNASeq) (78, 79). The gene editing tool CRISPR/Cas9 utilises guide RNA which matches with target gene (CRISPR) and CRISPR-associated protein 9 (Cas9), an endonuclease which helps to break the dsDNA and facilitate editing of the target gene (80) and it can be used for loss-of-function studies (12, 45, 81). Bioinformatics tools have been developed and adapted for genomic studies across microbial, Drosophila and human genomes and these allow for rapid screening of genomes and vast publicly available pathogen and fly data for potential targets for further study (79, 82). These software tools coupled with publicly accessible databases such as Drosophila Evolution over Space and Time (DEST), FlyRNAi (Drosophila RNAi Screening Center and Transgenic RNAi Project (DRSC/TRiP)) and FlyBase form a powerful computational component of the Drosophila tool kit (67, 82, 83).

With a focus on the WHO priority pathogens, we will highlight some examples of how Drosophila has been used to address key questions around host-fungal pathogen interactions, immunology and drug interactions with a focus on developments in the last five years.

The WHO classified Cryptococcus neoformans, A. fumigatus, C. albicans, and the recently emerged C. auris as “Critical” pathogens. A. fumigatus is a filamentous, airborne pathogen that causes invasive aspergillosis, in vulnerable populations, like cystic fibrosis patients (32, 84). Drosophila has been used to study A. fumigatus pathogenesis since as far back as 2005 (85, 86). In 2010, Chamilos and colleagues showed that pathogenicity of A. fumigatus strains in a Toll-deficient fly model was comparable to that in a mice model (25). Since then, the fly model has been used to study the virulence of A. fumigatus mating types, effects of fungal volatile organic compounds on larval development and comparative pathogenicity of Aspergillus strains collected from diverse sources (environmental, clinical, airborne) (29, 57, 87).

Fungi produce volatile organic compounds (VOCs) which are easily vapourised, carbon-based compounds made up of “alcohols, aldehydes, acids, ethers, esters, ketones, terpenes, thiols and their derivatives” (56, 57, 88). A study of the effects of A. fumigatus VOCs in Drosophila was carried out over a 15-day period, by co-culturing the fungi and fly model. Quantitative measurements of VOCs production showed that greater volumes of VOCs were secreted when A. fumigatus was cultivated at 37°C, than at the fly’s preferred incubation temperature of 25°C (57, 89). In addition, exposure to VOCs resulted in varying levels of toxicity, ranging from mild to severe, including reduced speed and success rate of metamorphosis or death of 3rd instar larvae (57, 90). Gas-chromatography mass spectrometry analysis of A. fumigatus VOCs detected isopentyl alcohol 1-octen-3-ol at the highest volume (57). A Drosophila infection model was subsequently used to show that 1-octen-3-ol caused greater sensitivity in male than female flies, resulting in reduced dehydrogenase activity and nitric oxide production, and increased ROS production (35). The connection to sex may explain the similar sensitivity distribution witnessed in humans postexposure to mould (35, 91). The Drosophila models used to investigate the immune response to mycotoxins and studies have shown that the Toll pathway and secreted Bomanins, specifically neuronal BomS6, mitigate the symptoms of Aspergillus mycotoxin exposure, namely restrictocin and verruculogen (55). This could contribute to our understanding of how mammalian immunity interacts with mycotoxins.

The study by Almaliki (57) investigated the effect of VOCs produced by a single C. neoformans strain and found that the VOCs of this severe pathogen caused more severe morphological effects and higher death rates than all the A. fumigatus strains that were tested (57). C. neoformans is a pathogenic yeast able to establish invasive infections in immunocompromised patients. It has been frequently associated with HIV/AIDS and accounts for as much as 15% of HIV-related deaths (17). A Drosophila S2 protein expression system has been used to produce and purify a recombinant cryptococcal protease, May1. This protease was used for further investigation as a target to identify compounds that could simultaneously inhibit the fungal protease and HIV-1 protease, which would provide dual protection and lower toxicity for HIV/AIDS patients (92). In the study by Almaliki and colleagues (57) regarding the toxicity of VOCs in Drosophila, C. neoformans VOCs cause significant delays in metamorphosis with eclosion rates of 44% compared to 80% for controls.

Contemporaneously with the growing number of studies in Aspergillus, Drosophila has been used to study C. albicans, one of the most common causes of candidiasis and blood stream infections (7, 93). In 2004, Alarco and colleagues published a Toll-deficient Drosophila model through which they demonstrated concordant C. albicans pathogenicity findings with mouse models, giving validity to the use of fly models (94). This was further corroborated by similar study in Candida glabrata mutant libraries (65, 95). Drosophila studies have been used to investigate host adaptation by C. albicans. Liu et al. (59) demonstrated the necessity of phosphate transporter, Pho48, in establishing candidiasis in the wild type Oregon^R fly via infection with wild type C. albicans and Pho48 null mutants (59). Null mutants were 3.5 times less likely to cause fly death than wild type strains 5 days post-infection (59). Glittenberg and colleagues (66) via a targeted genetic screening of 5698 RNAi lines described the protective impact of fucosylation in immune defence against C. albicans. A recent study in a Bom ^Δ55C fly model, (lacking the ability to produce the full range of Bomanin peptides) highlighted the ability of Candida sp. (including C. albicans and C. auris) to break down proline for energy, which may promote virulence. Moreover C. albicans mutants lacking the Proline UTilization genes put1, put3 or put1/put2 genes) showed reduced virulence compared to control fungal strain in the same fly infection model (62).

Drosophila infection models have been used to investigate the efficacy and toxicity of potential antifungal compounds. These include a Toll heterozygous Drosophila, Tl[r3]/+, used to test the naturally occurring compound, acid ellagic acid, against C. albicans where researchers showed statistically significant survival rates, and no toxicity at the proposed effective doses (64). Raj et al. (96) demonstrated a >70% survival rate of wild-type Drosophila infected with C. albicans when treated with Syzygium samarangense leaf extracted in methanol and dissolved in dimethyl sulfoxide. While the dosage applied to the Drosophila infection model was not specified, 50 mg of the Syzygium samarangense leaf extract was effective at clearing colonisation in an ex vivo porcine tongue and skin model, suggesting it could have utility as part of a topical treatment (96). Drosophila infection models can also be applied to antifungal studies for known compounds with the goal of reintroducing or repurposing old therapies. Clioquinol was administered orally to treat parasitic infections in the mid-1900s, however its use was discouraged due to perceived side effects (63, 97). Researchers investigated the antifungal efficacy and toxicity of Clioquinol in a Toll-deficient Drosophila model infected with C. albicans (63).

Drosophila has been utilised to investigate the novel pathogen C. auris. Wurster et al. (11) used a Toll-deficient mutant, Tlr632/TlI-RXA (which shows reduced AMP production and reduced phagocytic ability) to investigate the pathogenicity of C. auris clades identified at the time (Clade I-IV), and to determine the efficacy of azole to treatment (11). Their findings suggested that there was variability among the strains’ pathogenicity, though all strains were more pathogenic than C. albicans (11, 98).

Species of non-C. albicans (NCA) have been investigated using Drosophila models. While NCAs have typically accounted for a smaller fraction of candidiasis infections, their prevalence and resistance to azoles and echinocandins is on the rise (17, 53, 99). NCAs in the high priority group include Nakaseomyces glabrata (C. glabrata), Candida tropicalis and Candida parapsilosis.

In 2018, researchers harnessed CRISPR/Cas9 for the targeted deletion of individual Drosophila Bomanin genes to determine their immunoprotective role against C. glabrata (43). Using in vivo and ex vivo infection models, they demonstrated that Bomanin genes do not act in tandem and the short-form Bom peptide was immunoprotective against C. glabrata on its own (43). They showed that flies lacking 10 out of the 12 Bomanin genes (42) were as susceptible to infection as Toll-deficient flies, highlighting the importance of Boms in host immunity (43). Studies in Drosophila cell lines have been used to identify mechanisms by which C. glabrata evades innate immunity strategies, like AMPs and ROS, and potential drug targets. A study by Kounatidis and colleagues showed that C. glabrata ADA2 gene is essential for the pathogen to resist oxidative stress as the ADA2 knockout yeast could only grow in flies with suppressed ROS, while overexpressing ADA2 promoted C. glabrata growth and resulted in lower host survival rates (65). The role of the potassium transporter C. glabrata TRK1 was elucidated through infection of MyD88 and Bom ^Δ55C Drosophila strains with wild type and C. glabrata trk1 knockout (60). Loss of TRK1 gene resulted in cell wall modifications and reduced virulence within the host environment, in a potassium concentration dependent manner (60).

C. parapsilosis is associated with neonatal infections in addition to candidemia and candidiasis in immunosuppressed patients (17, 53). The Toll pathway has been shown to be crucial for Drosophila survival when infected by C. parapsilosis (which was not the case for Persephone protease), by comparing the susceptibility of wild type and mutant MyD88−/− flies to C. parapsilosis (53).

In addition, the high group includes Mucorales, Fusarium sp. Histoplasma sp. and eumycetoma causative agents (17). Mucorales are a large group of ubiquitous, filamentous fungi, frequently found in soil, which can cause infections ranging from mild to invasive (100, 101). The Order includes genera like Rhizopus, Mucor and Lichthiemia (101). Building on previous preexposure studies that showed the utility of Drosophila in Mucorales studies, Wurster and colleagues showed that exposing three Mucorales, Rhizopus arrhizus, R, pusillus, and Mucor circinelloides, to the triazoles isavuconazole and voriconazole, triggered hypervirulence in the fungi, resulting in lower survival rates in a Toll-deficient model (Tlr ⁶³² /TlI ^-RXA). This was a significant finding as it could explain infections arising in patients undergoing prophylaxis or treatment with isavuconazole (102). While the number of Mucorales tested was small, this gives some insight into this treatment challenge. This is in contrast with A. fumigatus, which often occupies the same niche and is managed in a similar way, but does not develop isavuconazole-induced hypervirulence (102).

The Fusarium solani species complex, includes F. solani sensu stricto, F. falciforme and F. keratoplasticum, and they are major opportunistic fungal pathogen, capable of causing keratitis (58). A screen of 42 environmental and clinical isolates from South India revealed that all isolates were intrinsically resistant to first-generation azoles and susceptible to imidazole, which contributes to treatment challenges (58). Survival assays comparing Oregon-R wild type and Myd88 mutant flies, infected with 6 Fusarium sp. found that MyD88 is required to mount an effective Toll defence against all Fusarium strains (58). Homa et al. (58) also showed that Fusarium virulence was distinct at strain level (58). Subsequently, Cohen et al. found that survival rates following Daisho peptide knockout also varied among Fusarium species (51).

The lower priority category has the highest number of pathogens including Scedosporium sp, Lomentospora prolificans, Coccidioides sp, Pichia kudriavzeveii (Candida krusei), Cryptococcus gattii, Talaromyces marneffei, Pneumocystis jirovecii and Paracoccidioides sp (17). These pathogens have the lowest relative global incidence and mortality rates, but still have substantial impacts (2). One of these pathogens, T. marneffei is a thermally dimorphic fungal pathogen localised to South and Southeast Asia (61, 103). It is found in the environment and Bamboo rats and can be inhaled and establish severe invasive infections in humans and animals (61). Its prevalence is not fully known due to limited surveillance and diagnostics and though mortality rates can be as high as 30%, few host-pathogen interaction studies have been performed (103, 104). Qu et al. (61) used the MyD88 −/− fly model to investigate the significance of the T. marneffei mating type on virulence in 107 clinical, Bamboo rat and environmental samples. They demonstrated that the mating type (MAT1-1 or MAT1-2) did not have an impact on flies survival upon infection, despite the fact that MAT1-2 isolates were overabundant across the entire sample population (61).