With all of the above, D. melanogaster emerges as a promising model for discovering higher-quality drug leads. However, Caenorhabditis elegans and Galleria mellonella also address time and cost concerns associated with mammal models following the principle of the 3Rs. These models have played a crucial role in improving the understanding of host-pathogen interactions and innate immune responses, and their roles in drug screening have been extensively discussed (Giacomotto and Ségalat, 2010; Pandey and Nichols, 2011; Pukkila-Worley and Ausubel, 2012; Tsai et al., 2016; Tran and Luallen, 2024). Therefore, it is important to consider the advantages and disadvantages of D. melanogaster compared with these established models when screening drugs against infectious diseases.

On the one hand, C. elegans is suitable for high-throughput screening, given its small size, rapid life cycle, and transparency, which allow for automated dispensing, image analysis, and efficient screening of thousands of compounds in 348-well plates (Giacomotto and Ségalat, 2010). It shares approximately 65% homology with human disease-causing genes; however, it might be difficult to predict mammalian bioactivity given the anatomical simplicity of C. elegans relative to mammals. In addition, its thick cuticle hinders compound absorption, and its small size complicates measuring concentrations (Pukkila-Worley and Ausubel, 2012). In this regard, the fly digestive tract is closer to mammals’ one, and treatment efficacy assays in this model also confer information about the PK-PD, gut permeability, location, and biodisponibility of the drug in a specific place in the host, mimicking closely what happens in mammals (Lee and Min, 2019). Additionally, flies drive sex-specific responses while C. elegans do not have a male/female sexual differentiation. This could be advantageous for simplicity, but it could also be a disadvantage when assessing possible sex-dependent treatment features such as dose, concentration, or time of exposure (Pandey and Nichols, 2011; Arch et al., 2023). Regarding innate immunity, C. elegans lacks key components of the mammalian immune system, such as vertebrate cytokines, cellular immunity, inflammasome, complement, and melanization pathways which D. melanogaster and G. mellonella have (Giammarino et al., 2024). It also lacks essential pattern recognition receptors (PRRs) such as Toll-like receptors (TLRs), and transcription factors like NF-κB and MYD88 (Tran and Luallen, 2024). Instead, C. elegans relies on GATA transcription factors (e.g., ELT-2) and evolutionarily conserved pathways such as p38 MAPK, β-catenin, and FOXO to regulate immune responses. The TFEB/HLH-30 transcription factor governs autophagy (Pukkila-Worley and Ausubel, 2012). Overall, the simplicity and ease of handling of C. elegans make it an excellent tool for successfully screening thousands of compounds.

On the other hand, the larval stage of Galleria mellonella is also suitable for biochemical analyses, pathogen interaction studies, and drug screening. This model is easier to maintain and handle compared with D. melanogaster. In addition, one of G. mellonella’s key advantages is its ability to survive at 37°C, which mirrors the optimal temperature for many human pathogens (Tsai et al., 2016). Yet, unlike D. melanogaster, which can survive a large inoculum of the pathogen, G. mellonella is highly susceptible to low doses of infection by numerous pathogens (Glavis-Bloom et al., 2012). G. mellonella larvae infection model is usually experimentally used at the final instar stage, which lasts approximately 8.40 days, limiting the evaluation of chronic infections and treatments that require longer exposure times (Serrano et al., 2023; Wojda et al., 2020). In this model, the most common route of drug administration is through systemic injection, which provides precise control of the dosage given to each individual but lowers the throughput (Tsai et al., 2016). This differs from the D. melanogaster and C. elegans models, where oral administration is more commonly used (Glavis-Bloom et al., 2012; Tzelepis et al., 2013). Immune responses in G. mellonella share key components with the mammalian innate immune system, such as cellular responses driven by hemocytes involved in processes like phagocytosis, nodulation, and encapsulation (Tsai et al., 2016); and humoral responses triggered when pathogens are recognized by pathogen recognition receptors (PRRs), such as Toll-like receptors, leading to the release of antimicrobial peptides (Asai et al., 2023). However, G. mellonella lacks the resources found in other models, such as stock centers, standardized procedures, transgenic strains, and full genome sequencing, making it less versatile for certain types of research (Giammarino et al., 2024). Despite all this, G. mellonella has made a space for itself as a potential model to be used in the drug screening timeline.

All in all, the selection of which model organism to use depends on the nature of the disease being studied, the scientific questions being asked, and the type of screening procedure desired. In general terms, C. elegans and G. mellonella are frequently used for early toxicity screening and have become viable high-throughput options for evaluating drug candidates’ safety and efficacy. While D. melanogaster is a more challenging model, with lower throughput, it offers a complex scenario for studying drug susceptibility due to tissue homology, sex dimorphisms, and a conservative immune system. Taking all these features into account can help to evaluate the best timeline for your drug screening.

After these considerations, we will address all the aspects to take into account in the D. melanogaster model for compound screening in the infectious diseases field.

The most common and high-throughput method to treat the flies is to dissolve the drug either in the fly’s melted standard cornmeal or in a minimal medium (made with sucrose, agarose, and water) and let the flies feed ad libitum (Kruger and Denton, 2020). This methodology has the advantage of mimicking the most common route of drug administration in humans for treating local infectious diseases and also simplifies a lot the handling of the animal model. It is generally used for long-term exposure studies where flies are exposed to the drug food for more than 24 h (Diegelmann et al., 2017; Figure 1). When to administer the compound and the exposure time of the fly to the drug-complemented media will depend on the target infectious agent and its pathogenic cycle in the fly. In addition, it is well known that adding rewarding substrates, such as sucrose or yeast paste, to the drug-food preparation increases the possibility for a successful administration of the compound to the flies in addition to masking the possible bad taste of drugs (Pandey and Nichols, 2011). In this case, it is also important to consider the thermostability of the compound. If the compound is thermosensitive, then it should be added to the media surface when it is cold. This prolongs the media preparation process and requires considering the possible dilution of the drug into the media when calculating the necessary drug concentration (DeLoriea et al., 2023).

On the other hand, when the purpose of the study is to assess the effect of acute exposure to the drug, then the strategy is to limit the fly’s capacity to reach the nutrients so they only are fed with the compound of interest. Before being transferred to the treatment vials, adult D. melanogaster should be starved to ensure the predisposition of all the flies to eat the compound. This will favor the consumption of the fly of a large dose of the drug (Pandey and Nichols, 2011). The duration of the starvation as well as the duration of the exposure to the drug will also depend on the characteristics of your study. Some examples of this methodology are to cover the standard feeding media fully with a Whatman Filter Paper disk in which you put the compound solution in a proper concentration complemented with sucrose (Needham et al., 2004; Elbi et al., 2017), or let the vials medium-free and put a pile of folded papers or cotton balls previously soaked into a 2.5 mL solution containing the compound in a proper concentration (Troha and Buchon, 2019; Figure 1). These methods show variability in the amount of drug ingested and reduce screening throughput. Moreover, since these methods, together with the minimal media, do not allow the flies to receive all the nutritional intake they need, the exposure time and the duration of the study must be reduced to a safe time limit in order to not misinterpret efficacy results with malnutrition effects (Kruger and Denton, 2020).

Survival monitoring is, by far, the most commonly used technique for treatment studies in D. melanogaster as it permits the highest throughput and reliably shows fly fitness along the experiment. It is used in the toxicological evaluation of drug candidates, dose–response experiments, and infectious disease treatment monitoring, among others. It consists of daily monitoring of live/dead flies along the duration of the experiment, which normally lasts until the infection progresses in D. melanogaster and kills the flies in the non-treated group; this could be hours for high replicative pathogens such as S. aureus (15–25 h) or days for slow-growing pathogens such as M. marinum (approximately 10–20 days) (Dionne et al., 2003; Needham et al., 2004). Nevertheless, it will depend on the characteristics of your study as the virulence of the pathogen, the infectious route (oral or systemic), and the infective dose. Every 2–3 days flies must be placed in new tubes of medium with or without compound because those get sticky and pasty, which can trap and kill flies, leading to a misinterpretation of survival results. In addition, there is a risk of flies not being able to reach the compound as it may be hindered by the eggs laid by females in the medium, impairing treatment efficacy (Troha and Buchon, 2019). Suitable treatment can improve survival curves by 50–90% (Zhang et al., 2021; Tonkin et al., 2022). However, this will depend also on the exposure time to the treatment (pre-infection, immediately post-infection, or some days after the infection), and the compound dose. Of note, treatment can contribute to infection clearance at the expense of flies’ survival. Kaynar et al., performed oral treatment with linezolid along 18 h of S. aureus-infected flies and measured survival and CFUs on several time points post-treatment. Flies did not show S. aureus CFUs in the treated group on day 20th after treatment. Nevertheless, the infection and the treatment impaired the overall survival rate in comparison with the non-infected group (Kaynar et al., 2016).

The outcome of the data generated are survival curves. They represent the probability for an individual to survive to a given time as proportions of the total of live flies. The outcome is a dichotomous event, corresponding to 1 for die or 0 for live. Those flies that die within the first 2 h after infection should be excluded from the analysis as their death can be attributed to the severe injury of the wound due to injection and not to the infection (Troha and Buchon, 2019). Analysis can be calculated using different equations. The Kaplan–Meier method is a non-parametric method that displays the probability of survival as a function of time, this means, the probability of the event occurring at time t (Lee, 2023). This method is appropriate when your study does not have additional variables (covariates), for example when studying treatment versus control groups (Kong et al., 2023; Porzio et al., 2023). On the other hand, the Cox proportional hazards model is a semi-parametric method that uses hazard functions to give a sense of the risk of the event occurring for an individual at time t (Lee, 2023). The Cox model takes into account the covariates that might have an effect on survival. In treatment experiments, this method is normally represented by comparing the ratio of hazards between the treatment and the control group (Ferreira et al., 2014; Van den Bergh, 2022).

In D. melanogaster, and in the infections field in general, host fitness (survival) not always reflects infection status in the fly. It can happen that, when treated, flies may have the same or worse survival rates in comparison with the non-treated group but clear the infection. This happened when Porzio et al. gave lactonases in combination with Tobramycin to P. aeruginosa-infected flies. Flies infected with P. aeruginosa in the presence of the enzyme had a significantly increased rate of D. melanogaster deaths. However, the same group of flies showed a reduction in the number of P. aeruginosa recovered at the end of treatment. This happened because the lactonases reduce the biofilm formed by P. aeruginosa in infection, favoring a rapid spread of the infection through the fly. At the same time, when lactonases were administered together with tobramycin, this resulted in higher susceptibility of planktonic P. aeruginosa to the antibiotic treatment (Porzio et al., 2023). In the same scenario, the trade-off between immunity and lifespan in D. melanogaster has been extensively studied. The investment by the host in immunity impairs its fitness attributes such as longevity and fecundity (Ye et al., 2009; Arch et al., 2023). In treatment, this can happen when host-directed therapies are administered, where an activation in flies’ immune system can end their life but also infection in a phenomenon known as cytokines storm, also evidenced in humans (Van De Leemput and Han, 2021). Contrarily, it may also happen that the compound does not display any antimicrobial effect in the infection but acts as host-directed therapy that favors tolerance, meaning, increases the lifespan of the fly against elevated pathogen load (Vincent and Dionne, 2021). This is why, to better elucidate the effect of a drug against infection (clearance, or tolerance and resistance induction), it is worth it to add pathogen load data to your survival data.

Pathogen load is mostly measured by counting the number of Colony Forming Units per fly at specific time points in the study (for example, pre-, and post-treatment). This is a good method when monitoring treatment because it shows the live bacteria remaining in the fly. The experimental procedure is perfectly detailed in Troha and Buchon (2019). Briefly, individual infected or infected and treated flies must be harvested, rinsed in ethanol 70%, homogenized in PBS, serially diluted, and plated in agar plates of the appropriate media until, after incubation, CFUs can be counted (Troha and Buchon, 2019). The addition of proper antibiotics on the media must be taken into account in order to get rid of the fly microbiota and detect only the pathogen of interest. From this point, tolerance and resistance can be measured in order to better elucidate treatment effect. While resistance strategies will aim at killing the pathogen or inhibit its proliferation, tolerance strategies will reduce the negative effect on fitness caused by the infection, but will not have an impact on pathogen fitness (Arch et al., 2022). Arch et al., defined tolerance as the slope of the regression between survival and pathogen load, with the flattest slopes representing increased tolerance. Resistance was measured by the authors as the Y-intercept of the regression line between the pathogen load upon death and the inoculation dose, with a lower Y-intercept as a measure of more resistance (Arch et al., 2023).

It can happen that culture-dependent methods are not suitable for your study. For example, this can occur when you cannot distinguish between the microbiota of the fly and the pathogen of interest by specific culture plate or antibiotics. Alternatively, the pathogen may be distinguishable by plating but it may require long incubation times and favorable temperatures that allow microbiota to select for antibiotic-resistant clones and grow, making it difficult to count the microorganism of interest. In these situations, pathogen load can be measured by quantification of microbial genes in fly extracts by PCR. Nevertheless, in treatment studies DNA detection is not a reliable technique to monitor treatment response as DNA is a stable molecule that survives long after cells have died (Honeyborne et al., 2011). Contrarily, 16S ribosomal RNA has a shorter half-life that has demonstrated parallelism with treatment and cell viability (Sheridan et al., 1998; Gordillo et al., 2006; Honeyborne et al., 2011; Azam et al., 2022). Some authors have already demonstrated the ability of the 16S rRNA to monitor treatment by performing a real-time Reverse Transcriptase quantitative Polymerase Chain Reaction (RT-qPCR) in tuberculosis patients’ sputum samples. Results showed a correlation between the detection of the gene and the time to positivity in MGIT and a good treatment follow-up (Honeyborne et al., 2011; Azam et al., 2022). Nevertheless, this technology has not been used yet in D. melanogaster treatment studies.

Qualitative manners of evaluating pathogen load are also present, such as the use of fluorescently labeled bacteria and their visualization in a fluorescence dissecting microscope. However, this is a lower throughput assay likely to be more relevant for validation of leads or when a more detailed analysis of the fly is required, for example, location and specific organ information. Fluorescence microscopy has been used by some authors to verify treatment efficacy against infections such as M. marinum, M. abscessus and S. aureus in D. melanogaster adult flies (Oh et al., 2013a, 2014; Nimal et al., 2016). Thus, this technique is commonly used to confirm and validate the results you are obtaining in the dose–response experiments rather than be used routinely to monitor treatment.

Extra information can be extracted from our treatment study in D. melanogaster if we assess the immune response in the host elucidated by treatment. This is an advantage of the fly over other model hosts since the expression of any gene in D. melanogaster can be studied at specific time points (Tzelepis et al., 2013). This can be achieved by comparing endogenous RNA transcript levels between infected D. melanogaster and infected and treated flies by RT-qPCR (Troha and Buchon, 2019). It is worth it to point out here that it is difficult to obtain results from individual flies. A pool of 3–10 flies will be necessary per sample and time point to reach sufficient RNA levels after the acid nucleic extraction (Troha and Buchon, 2019; Arch et al., 2023). The nucleic acid extraction can be done in multiple ways; you should check which one fits with your laboratory conditions and works well in your samples. The Trizol method is very efficient but also some commercial kits can provide high levels of purified RNA with less background noise, that can interact with PCR reagents (Livak and Schmittgen, 2001; Moskalev et al., 2015; Wu et al., 2016; Hemphill et al., 2018; Green and Sambrook, 2020; Arch et al., 2023). At this point, after a retrotranscription, a qPCR of the genes of interest can be done (Tajadini et al., 2014).

A myriad of immune-related genes of D. melanogaster can be studied. When elucidating treatment efficacy, it is of interest to study the change of expression of the typically expressed genes under infection in both flies that are infected and flies infected and treated. Genes such as drosomycin (drs), metchnikowin (mtk) and defensin (def) are known to increase their expression after a Gram-positive or fungal pathogen infects D. melanogaster by an upregulation of the Toll pathway (Lee and Edery, 2008; Hedengren-Olcott et al., 2004; Kaynar et al., 2016). Contrarily, when a Gram-negative pathogen infects the fly, AMPs such as diptericin (dpt), attacin (att), and cecropin (cec) are expressed by activation of the Imd pathway (Gottar et al., 2002; Mulcahy et al., 2011; Tapadia and Verma, 2012). In addition, the JAK/STAT pathway, involved in diverse biological processes, induces the transcription of Turandot stress genes, Upd3 (a cytokine analog of mammalian IL-6 and also related to cellular immunity), and some antiviral genes such as vir-1 (Agaisse et al., 2003; Chakrabarti et al., 2016; Lopez et al., 2018). Regarding cellular immunity, the Imaginal morphogenesis protein late 2 (Impl2) has been recently related to the activation of phagocytes (Harsh et al., 2019; Krejčová et al., 2023), and Atg8a with autophagy (Tsapras et al., 2022; Demir and Kacew, 2023). Once qPCR is performed, the relative transcript levels of target genes are calculated using the 2^−ΔΔCT method with rpl32 as the reference gene for the normalization of data (Livak and Schmittgen, 2001). A list of some relevant genes in the infectious field, with their amplification primer pairs for qPCR, is available in Table 3.

To emphasize the importance of gene expression analysis in treatment studies, a few examples are detailed here. To Bakalov et al., the measurement of different immune genes served to elucidate which were the mechanisms underlining the increase in survival curves experienced by S. aureus-infected flies treated with linezolid and DCA for 1 week in comparison with flies only treated with linezolid. The authors showed that flies on a DCA diet experienced a significant decrease in expression of drosomycin and cecropin A and an increase in defensin in comparison with those infected flies growing on a regular diet (Bakalov et al., 2020). Along the same line, it has been demonstrated that infected flies treated with linezolid showed a focused and maintained increment in expression in comparison to sham, which corresponded with the shorter lifespan of these groups (Kaynar et al., 2016). Similarly, Thomsen et al. (2016) showed that the treatment efficacy of lantibiotics nisin and NAI-107 was correlated with a decrease in expression levels of the immune response genes drosomycin, cecropin A, and attacin B in flies infected with S. aureus. If the compound is a HDT the gene expression data brings a lot of information on how is the host actually reacting to the treatment, which, in this case, is expected to have a direct effect on the progress of the infection. When D. melanogaster is previously challenged with heat-killed E. coli, then flies showed protection against the consequent infection by both Enterococcus faecalis and P. aeruginosa mainly moved by a synergistic action of Toll and Imd pathways (Wen et al., 2019). Altogether, the gene expression gave an additive informative value to what was happening in the host during and after treatment, which better helped to elucidate pathogenic phenotype.