Drosophila and its gut microbes: a model for drug-microbiome interactions

Abstract

Drosophila is an excellent system to investigate how the presence and composition of the gut microbiome influences the efficacy of therapeutic drugs and downstream consequences for host health. These opportunities derive from two key attributes of Drosophila. First, Drosophila is amenable for microbiome research, with simple, standardized methods to produce large numbers of microbiologically-sterile flies and flies with a standardized gut microbiome, thereby facilitating experimental reproducibility. Second, Drosophila is a well-established genetic model that is increasingly used to elucidate the molecular and physiological basis of lesions associated with human disease alleles; this provides the opportunity to link microbiome/drug interactions to previously-described processes shaping health and disease. In this way, Drosophila can fast-track understanding of fundamental biology to generate precise hypotheses for testing in mammalian systems. Drosophila is particularly well-suited to investigate the incidence of microbiome/drug interactions mediated by different mechanisms, including microbial drug metabolism (to active, inactive or toxic derivatives), microbial production of compounds that inhibit drug efficacy, and off-target effects of the drug on the microbiome, resulting in dysbiosis and host ill-health. Drosophila can also be used to investigate how interactions between the microbiome and host genotype may shape responses to therapeutic drugs, informing the reliability of precision medicine based exclusively on human genomic markers.

1. Introduction

The basis for this article is the largely untapped potential of simple animal models to facilitate study of drug interactions with the resident microorganisms, also known as the microbiome, in human patients. The basis for these opportunities is two-fold.

The first issue is the increasing evidence that the microbiome can influence the efficacy of drug therapies administered to human patients [1–3]. Multiple processes have been implicated. Microbial metabolism of drugs can variously increase therapeutic activity, cause toxicity or inactivate drug function (Fig. 1A). Microbes can also reduce drug efficacy by releasing metabolites that compete for host drug receptors (Fig. 1B), and by modulating physiological traits of the host, especially immune-responsiveness. Compounding these various microbial effects on the therapeutic value of drugs, some human-targeted drugs with no predicted anti-microbial activity inhibit the growth of human gut bacteria [4] and the resultant changes in the microbial composition can cause dysbiosis, i.e. have negative consequences for human health (Fig. 1C). These complex effects are exceptionally difficult to manage because the taxonomic and functional composition of the microbiota varies extensively among individuals and can also change over time within one individual [5]. The problems posed by microbiome-drug interactions are particularly acute in the context of precision medicine, i.e. treatments tailored to the individual patient. This is because precision medicine is guided by the genetic traits of the patient, but microbiome composition and function cannot be predicted accurately from host genotype [6].

Fig. 1.

Microbiome-drug interactions. A. Microbial metabolism of drug, resulting in activation, inactivation or production of toxic product from the drug. B. Microbial product as competing ligand for drug target (host receptor, enzyme etc). C. Drug-mediated changes to the composition or activities of the microbiome, causing dysbiosis and reduced host health.

The second issue relates to the need for models to conduct controlled experiments investigating causality and mechanism of drug-microbiome interactions. Fortunately, a diversity of models is available. Human cell cultures and organoid systems colonized with human-derived microbes account for the genetic capacity of humans for transport and metabolism of drugs and their products [7, 8], while animal models capture the physiological and biochemical complexity of organ systems and their interactions [9]. To date, most animal research has used the mouse model, focusing on the gut microbiome and its interactions with orally-delivered drugs, e.g. [10]. The mouse has the advantages of compatibility with many members of the human gut microbiome and physiological similarities to humans, although various anatomical and biochemical differences can limit translation of results to humans [11]. However, the scale of some experimental designs to investigate drug-microbiome interactions in the mouse can be constrained by cost, especially of germ-free mouse facilities, and the use of large numbers of mice can also raise animal welfare concerns. Alternative animal models suitable for large, high throughput experimental designs with minimal welfare issues are available, but their use to study drug-microbiome interactions has, to date, been very limited.

The focus of this review is the potential to use Drosophila for cost- and time-effective study of drug-microbiome interaction without major animal welfare concerns. As in humans and most other animals, most microorganisms associated with Drosophila are restricted to the gut lumen; and this microbial community is of greatest relevance to interactions with the many orally-delivered therapeutic drugs. This article provides a summary of Drosophila as a biomedical model and its gut microbiome, followed by an overview of the advantages and limitations of the Drosophila system as a model for the human gut microbiome. Finally, the ways that Drosophila can contribute to future study of drug/microbiome interactions are discussed.

2. Drosophila and its microbiome

Drosophila melanogaster is universally recognized as a premier genetic model, supported by superb genetic and genomic resources, as well as sophisticated molecular techniques to engineer the genome and control gene expression with exquisite spatiotemporal precision [12]. Its widespread use as a human disease model is driven primarily by the demonstration that the Drosophila genome includes orthologs for many human disease genes, and the capacity to replace the Drosophila genes by human orthologs, including disease alleles, for functional analysis [13, 14]. In addition, the Drosophila research community has developed powerful resources to facilitate translation between the fly and human disease, including the Human Disease Model Report in FlyBase (http://flybase.org/), where the disease orthologs in Drosophila are collated with links to the OMIM (Online Mendelian Inheritance in Man) database [15], and DIOPT (Drosophila RNAi Screening Center (DRSC) Integrative Ortholog Prediction Tool) for Drosophila-human ortholog searches [16]. Major biomedical Drosophila programs focus on the genetic mechanisms underlying cancers and both neurodegenerative and metabolic diseases [17–20], further facilitating the opportunity for speedy translation of discovery of drug-microbiome interactions in Drosophila to treatment of human disease.

The Drosophila microbiome is dominated by two groups of taxa: gut microorganisms localized to the gut lumen, and endosymbionts, which are transmitted with high fidelity from mother to offspring via the oocyte. Endosymbionts, especially the α-proteobacterium Wolbachia which is harbored by many laboratory strains of Drosophila, can have substantial effects on the metabolic and immune phenotype of flies [21]. Humans lack such endosymbionts and, generally, the use of endosymbiont-free Drosophila strains is recommended for study of drug-gut microbiome interactions. Molecular and microscopical methods to test for endosymbionts in Drosophila are available [21].

The gut microbiome of laboratory Drosophila is generally dominated by bacteria, usually Acetobacteraceae (α-proteobacteria) and Lactobacillales (Firmicutes) [22]. Drosophila diets with a high sugar content favor Acetobacteraceae, while lactobacilli are favored by Drosophila diets dominated by complex carbohydrates, e.g. cornmeal-based diets. Ascomycete yeasts, predominantly Sacharomycetales (e.g. Hanseniaspora, Pichia), are prevalent in wild Drosophila [23], but are suppressed by the near-universal use of antifungal agents in laboratory diets. In routine Drosophila cultures, bacterial abundance per fly varies widely even within a single vial, from 100->100,000 viable cells per fly, as quantified by number of colony-forming units on standard media [24, 25]. Consistent differences in microbial abundance and composition between male and female flies are not evident, but the microbiome shifts with fly age towards increased abundance and among-fly variability in composition [24].

For the proper design and interpretation of experiments on Drosophila-microbiome interactions, an appreciation of the nutritional ecology of Drosophila is essential (Fig. 2). The natural diet of Drosophila is dominated by the microorganisms that mediate the rotting of fruits and other decaying plant material [26]. In other words, Drosophila is a microbivore, and many of the microorganisms ingested into the gut are digested as food items, making a quantitatively important contribution to host nutrition (Fig. 2A). By contrast, microbes generally make an incidental contribution to human nutrition. Details of the transmission dynamics of gut microbes also differ between Drosophila and humans. Although microbes are acquired orally and lost by fecal shedding in both systems, the microbiota is more dynamic in Drosophila than in humans. Many of the taxa in Drosophila exploit the mobile fly as a vector for transmission between fruits at different stages of decay [26] (Fig. 2A, B) and, consequently occur as substantial free-living populations [25, 27]; but most human gut microbes are obligate anaerobes (or micro-aerobes) with no (or limited) free-living populations and with direct transmission between hosts dictated mostly by host social interactions [28]. The capacity of ingested microorganisms to persist and proliferate in the Drosophila gut varies among microbial isolates, and also among individual cells of a single isolate [25, 27, 29]. For example, in one study of the fate of Acetobacter tropicalis ingested by adult flies, one-third of the cells passed through the gut with bulk flow of food, 10–20% of the cells persisted in the gut, and the remainder were lost, presumably by digestion in the gut lumen [29]. Some microbial taxa are likely subjected to repeated transitions between food and gut by oral-fecal cycling within a single Drosophila culture in the laboratory (or fruit in the field), with unknown and potentially important effects on their metabolic traits, including interactions with administered drugs.

Fig. 2.

Drosophila-microbiome interactions. A. Mutualism founded on fly-mediated dispersal of fruit-rotting microorganisms from rotted fruit to over-ripe fruit, and microbial contributions to Drosophila nutrition, comprising both insect digestion of microbial biomass and utilization of microbial products (including microbial protein and B vitamins, as well as products of fermentative metabolism). B. The Drosophila life cycle, from sexually-produced egg, through three larval instars and pupa (with complete metamorphosis, involving the development of the adult organs and body form) to the adult fly. * the late third instar larva ceases feeding and moves to (and often away from) the fruit surface prior to pupation.

As a result of the intimate relationship between Drosophila and its food, the microbial complement of the diet is an important dimension of Drosophila-microbiome interactions. The microbial community in the diet is a major determinant of the microbiota in Drosophila and, reciprocally, the presence and activity of Drosophila influence the composition of the microbial community in the diet [30, 31]. In addition, the metabolic activity of the microbes modifies the composition of the diet ingested by the Drosophila [32, 33]. To date, microbial metabolism has focused primarily on macronutrients, especially in reduction of dietary sugar content, but there is every expectation that the availability and chemical profile of test drugs and other xenobiotics administered to Drosophila cultures may be altered by microbes in the food. For these reasons, it is essential to monitor the chemical composition, as well as the microbiological complement, of the diet in studies of Drosophila.

For some purposes, microbial modification of diet composition would confound accurate interpretation of experimental results. This difficulty can be overcome by feeding flies on sterile liquid food administered in capillary tubes, a procedure commonly referred to as CAFÉ (Capillary Feeder) [34]. Where investigated, fly feeding neither contaminates the capillary contents with microbes nor modifies the chemical composition of the liquid diet. Capillary feeding is not, however, suitable for high throughput analyses because the capillaries need to be replaced manually on a daily basis.

3. Experimental manipulation of the gut microbiome

The Drosophila gut microbiome is amenable to experimental manipulation in the sense that the insect host and its microbial partners can be maintained in isolation and associations with defined microbial partners can be synthesized (Fig. 3). Standardized protocols to generate microbe-free Drosophila are now established [35], founded on the exclusive localization of contaminating microorganisms to the outermost layer, the chorion, of deposited eggs. These microbes are eliminated by brief incubation of eggs in bleach (hypochlorite solution), which removes the chorion without damaging the eggs. The dechorionated eggs are then transferred aseptically to sterile food. The resultant axenic insects can be maintained through multiple generations, and apparently indefinitely, on nutrient-rich diets, but perform very poorly on nutrient-poor diets [36, 37]. Both the larval and adult stages feed readily on microbes (see above), enabling the synthesis of associations with single microbial taxa, defined mixtures or undefined mixtures obtained from fly feces or gut homogenates. Although Drosophila is a permissive host (meaning that it is capable of associating with diverse microorganisms) it is important to monitor the microbial complement in gnotobiotic hosts as a check for colonization of the flies by all administered taxa. For some microbial combinations, the persistence of specific taxa varies with the density of Drosophila in the vial [30], and so standardization of the number of eggs administered per vial is recommended for experimental designs involving a diverse microbiota.

Fig. 3.

Procedures for experimental dissection of Drosophila association with its gut microbiome. The insect host and its gut microbiota can be separated, and then associations with a standardized microbiota (of known composition) can be constructed. The standardized associations can include mutants or genetically-modified Drosophila and microbial taxa.

Some studies have used antibiotics as an alternative or supplementary method to egg dechorionation for generating axenic Drosophila. Generally, this is ill-advised because some antibiotic cocktails are not completely effective in eliminating all microbial cells, i.e. antibiotic-treated flies are microbe-depleted and not microbiologically sterile, and these treatments can have deleterious effects on Drosophila [38, 39]. Antibiotic treatment of insects derived from dechorionated eggs is totally unnecessary if proper aseptic technique is applied.

For reproducible results, both within and across laboratories, it is preferable to utilize gnotobiotic Drosophila with a defined microbial complement, together with axenic controls. Conventional treatments (Drosophila cultures with their native complement of microorganisms, see Fig. 3) can yield irreproducible results because of chance variation in the microbial composition [40, 41]. If it is considered important to use conventional cultures, then the microbial composition should be determined for every experiment.

The taxonomic composition and abundance of the microbiota in Drosophila can be scored by culturing on solid media, specifically quantification of the number of colony-forming units (CFUs) and by sequencing methods [35]. Each method has its limitations. CFU counts offer a rapid and inexpensive assay but this method is, self-evidently, limited to taxa that are culturable on the test media. CFU counts are sufficient for gnotobiotic flies colonized with culturable microbial taxa that are morphologically distinctive or that display differential growth on selective media. For example, Lactobacillus isolates are generally more susceptible than Acetobacteraceae to the antibiotic ampicillin but, unlike Acetobacteraceae, grow under low oxygen conditions. For complex communities, sequence-based methods are essential. Standard protocols are available for rRNA gene sequencing of bacteria (usually 16S rRNA gene fragments) and fungi (usually fragments that include the 5S rRNA gene and flanking regions of ITS1 and ITS2), supported by publicly available pipelines for sequence processing and databases for taxonomic identification. Whole genome sequencing of cultured microorganisms and shotgun metagenomics of the microbiome are also very useful, providing supplementary taxonomic information and insights into functional properties of the microorganisms [42–44]. Standard sequence-based methods cannot, however, discriminate between metabolically-active, inactive and dead microbial cells, and the output from amplicon studies can also be distorted by primer bias. An additional limitation of these methods is that they yield relative abundance data, even though host-microbiome interactions are influenced by total abundance, as well as by composition of the microbiome. Some studies use quantitative PCR assays with general 16S or ITS primers to obtain a rough index of total abundance of bacterial and fungal communities.

All experimental designs that involve manipulation of the microbiota should include axenic Drosophila to identify microbe-independent traits and, as importantly, to reveal any inadvertent microbial contamination. For many purposes, CFU counts offer a reliable check for microbial contamination but, especially during method development, endpoint PCR reactions using general bacterial or fungal primers provide an assured test for microbial sterility and, with follow-up amplicon sequencing, can establish the taxonomic source of contamination.

4. Outlook: opportunities for microbiome-drug interaction studies in Drosophila

To date, there has been very little application of Drosophila to investigate the incidence and mechanism of gut microbiome interactions with therapeutic drugs. This reflects a wider reticence in the research community to use Drosophila in drug discovery studies [45, 46], and can be attributed to two perceived limitations of the Drosophila model. First, screening strategies are, of necessity, more sophisticated for a whole animal than for cell cultures; and, second, the results obtained for an insect may correspond poorly, compared to the mouse model, to humans. For study of drug/microbiome interactions, these concerns are compounded by the difference in taxonomic and functional traits between the aerobic/micro-aerobic microbiome of Drosophila and mostly obligate anaerobic microbiome of humans. However, these issues should be treated as important considerations to assist in assessing whether and how Drosophila can contribute to a particular research question, rather than the basis to exclude the use of Drosophila altogether.

Let us start with the perceived complexity of drug screens using Drosophila. This constraint is mitigated by the long-standing tradition of high-throughput genetic screens in Drosophila. There is the opportunity to design drug screens that make use of the sophisticated genetic tools and superb repositories of mutant and transgenic Drosophila. Protocols available to Drosophila (but not the laboratory mouse and other mammalian models) include high throughput screens in 96-well format with fluorescence reporters of gene function [47, 48], and rapid imaging techniques that facilitate swift, accurate quantification of eye and wing defects in adult Drosophila with genetic lesions (or human alleles) expressed exclusively in these organs [14, 49]. Furthermore, the limitations imposed by the complexity of a whole-animal screen should be balanced against the value of the information obtained. Specifically, a Drosophila screen can address the contributions of processes (uptake, metabolism etc) in different organs on the organismal response to therapeutic drugs and drug-microbiome interactions, in a fashion that would be impossible for cell culture systems. In addition, Drosophila screens can be adapted for a diversity of high throughput read-outs of immediate biomedical relevance, e.g. lifespan, reproductive function, behavioral traits, that cannot be addressed with cell culture systems.

There is a traditional concern that the molecular and functional differences between insects and humans may undermine the value of Drosophila in drug discovery and drug-microbiome interactions, relative, for example, to the mouse model. This concern is being tempered by the growing number studies using Drosophila successfully in drug discovery. In particular, many medical drugs have activity in Drosophila that is founded on the same molecular mechanism as in humans [46]; and the biomedical relevance of Drosophila are enhanced further by studies using Drosophila engineered to code human (including disease) alleles [14]. Building on these parallels, it has been suggested that Drosophila may be particularly useful as an intermediate screen between in vitro systems and mammalian models, specifically to eliminate leads with poor efficacy in a “real” animal [20, 46].

These considerations demonstrate the feasibility and, in general terms, the biomedical relevance of Drosophila for studies of drug/microbiome interactions. Drosophila has the potential to make a significant contribution in the coming years. Of particular interest would be to establish the quantitative significance of the different modes of drug/microbiome interaction (Fig. 1) for different classes of drugs. In relation to precision medicine, Drosophila has exceptional potential as a system to investigate the relative significance of gut microbiota composition and host genotype in shaping variation in drug efficacy among individuals. The output of studies on Drosophila can then be used to construct precise hypotheses for directed testing in the laboratory mouse and other mammalian systems.