Drosophila strategies to study psychiatric disorders

Abstract

For decades, Drosophila melanogaster has been used as a model organism to study human diseases, ranging from heart disease to cancer to neurological disorders [9]. For studying neurodegenerative diseases, Drosophila has been instrumental in understanding disease mechanisms and pathways as well as being an efficient tool in drug discovery studies. For some better-understood disorders, such as Fragile X (a mental retardation syndrome), clinical trials are being run, based in part on translational work in flies and rodents. However, Drosophila is currently less used to study psychiatric disorders such as autism, schizophrenia and attention deficit and hyperactivity disorder (ADHD), despite numerous discoveries of disease susceptibility genes that could be explored by reverse genetics or miss-expression studies. This deficit might be explained by (1) a lack of reliable tests to study more complex disease (endo)phenotypes in flies, (2) difficulties in translating disease symptoms into animal models and (3) the polygenetic nature of these diseases. In this review we discuss strategies to use D. melanogaster to study complex psychiatric disorders such as schizophrenia, autism and ADHD. Two common features of these diseases may be defective sleep and attention mechanisms, hence calling for better methods for quantifying and screening arousal thresholds in flies.

Introduction

Drosophila, lovingly nicknamed ‘the Golden Bug’ by some Drosophilists [15], has been a useful model organism to study several biological processes including genetics, development, molecular and cell biology, and neuroscience. Even though the arthropod evolutionary lineage separated from the vertebrate lineage over 600 million years ago [2], [131], their genetic and molecular make up is remarkably similar to vertebrates [49], [141]. While flies and humans have very different body plans, they are remarkably similar on the level of biological processes. For example, the central nervous system of flies and humans are derived from a common evolutionary origin [77] and several neurobiological processes are similar in Drosophila and Homo sapiens, including membrane excitability, neuronal signalling and shared classes of neurotransmitters [118].

The Drosophila genome consists of approximately 16–17,000 genes [75] distributed on four pairs of chromosomes, although the 4th chromosome is so tiny it is mostly irrelevant. In comparison, the human genome has ∼27,000 genes on 23 pairs of chromosomes. There is considerable overlap, as ∼75% of all human disease genes have a recognizable match in the Drosophila genome [138]. About one third of the human disease genes, involved in a broad spectrum of diseases, have sufficiently well conserved homologues to be studied in Drosophila [138].

The extensive genetic toolbox of Drosophila melanogaster is the envy of many biologists, although psychiatrists rarely consider such tools as relevant to their field. In addition to versatile genetics allowing temporal and spatial control of any gene's expression [22], the ability to quickly generate genetic variants with altered gene expression patterns is potentially useful in studying disorders such as schizophrenia and autism, because these diseases are likely to involve complex underlying genetic effects. Lists of candidate genes involved in psychiatric disorders are growing almost weekly as more genes are discovered in genome wide association studies (GWAS) resulting from large groups of patients and matched controls. For example, an online database of schizophrenia candidate genes currently lists over 1000 genes and over 8000 polymorphisms (www.szgene.org see [3]). Screening the function of so many genes in a mammalian model can be impractical, and understanding genetic interactions among these genes is almost impossible. Yet, it is likely that heritable cognitive disorders not only result from multiple gene effects, but also multiple genetic interactions [11], [35], [96], [135]. By making simple crosses of mutant strains and phenotyping their progeny, Drosophila researchers are able to unravel genetic mechanisms in ways not yet available to most vertebrate researchers, let alone psychiatrists. In addition to these small-animal benefits, Drosophila has a relatively simple genome with less redundancy than the human genome, so mutations to single genes are more likely to yield a measurable phenotype. Importantly, the fly brain has only about 100,000 neurons [148], compared to the 100 billion neurons of the human brain. The fruit fly thus seems to fit the bill on all counts for unravelling the biology of psychiatric disorders, in a simpler reductionist model.

However, there is an obvious problem with the Drosophila approach to psychiatry, leading to some pointed questions. To what extent can psychiatrists use fly behaviours to study what their human patients are experiencing? What exactly should we measure when screening Drosophila mutants for candidate genes based on human GWAS data? Perhaps because the answers to these questions are not obvious, the history of studying psychiatric disorders in Drosophila seems to have placed greater emphasis on genes and less on behaviour, often utilizing the same set of two or three readouts to provide the requisite behavioural histogram alongside the molecular data. We challenge this approach, and will discuss alternate behavioural strategies later in this review. First, however, we will cover the traditional approaches to studying these diseases in flies, some more successful than others.

As outlined in Fig. 1, there are three strategies to initiate a fly-based approach to studying cognitive disorders. In the first strategy (forward genetic screens), random mutations are tested for behavioural phenotypes. The second strategy (reverse genetics) uses known disorder genes, derived from patient studies, and examines their roles in an animal model. The third strategy uses animal models to test more general theories about disorders, by for example manipulating environmental variables.

In the forward genetic approach, random mutations are introduced in a population with the same genetic background, after which the resulting mutants are analysed based on their gain- or loss-of-function phenotypes. This is a powerful way to dissect molecular pathways underlying biological process. Drosophila genetics was pioneered a century ago by Morgan [111] and mutagenesis approaches were subsequently developed to modify the animal's DNA, by using chemicals such as ethyl methanesulfonate [50], [86], [87]. Thus, individual genetic lesions could be associated with phenotypes. Around the time when Drosophila mutagenesis was increasingly being applied to understanding physical phenotypes, Benzer and colleagues realized that this same approach could be used to link individual genes with behaviours [12] such as circadian activity rhythms [92] and olfactory learning and memory [39], [47], [137].

Behavioural phenotypes such as circadian locomotion or olfactory memory are Drosophila-specific, and the possible relevance of any mutants to human psychiatry can only be revealed by further studies in humans and flies. One example of this approach successfully informing psychiatric research can be found the dunce mutant. The phosphodiesterase II gene (PDE4B), which is mutant in dunce animals, was found by forward-genetic olfactory memory screens [47], but eventually also associated with schizophrenia [20], [56], [132]. Forward genetic screens have also played a crucial role in identifying genes involved in the regulation of sleep in Drosophila [32], [33], [179], and it is likely that some of these will also point to genes we should be examining in human patients, as disrupted sleep is also a recurring theme in many psychiatric disorders (see Sections 1 Introduction, 2 A, 3 , 3.1 Schizophrenia, 3.1.1 , 3.2 Autism spectrum disorders, 3.3 Attention deficit hyperactivity disorder, 3.3.1 , 4 Problems with studying psychiatric diseases in, 5 Solving the problem with appropriate behavioural assays, 5.1 Modelling positive, negative and cognitive symptoms of schizophrenia in animals, 6 Discussion, 6.1 Future directions, Conflict of interest below).

Alternatively, by using reverse-genetic strategies, specific genes of interest can be targeted in two ways. When a Drosophila homolog of the gene of interest exists, expression levels or patterns can be altered or disrupted. On the other hand, when there is no Drosophila homolog, function of a human disease gene can still be studied by expressing a variant of the human gene in the fly model. This latter approach proved fruitful studying neurodegenerative disorders such as Alzheimer's disease, by expressing human tau [101] or Amyloid Beta [80] in Drosophila [76]. In both reverse-genetic approaches, flies are tested for neurological phenotypes correlating with the disease, such as age-associated locomotion defects or cellular degeneration visualized by imaging the fly brain [80].

In a third approach, theories of the disease etiology or disease progression can be tested, for example by exposing flies to altered environmental conditions that are proposed to play a role in the disease. The Drosophila Fragile X story (below) provides an excellent example, where mGLuR5 activity was linked to the disease phenotype [97], and the severity of the defect was associated with the concentration of glutamate in the diet [28]. Other potential areas of investigation could include the link between drug use and psychosis [161] or links between schizophrenia and advanced paternal age [164].

While there are theoretically three viable approaches for studying neuropsychiatric diseases in Drosophila, there are really two lynchpins for success. The first is the extent to which disease genes have a fly homolog. This has recently been reviewed by O’Kane [118] for schizophrenia, autism, and bipolar disorder. The second lynchpin is the ability to test a fly mutant for symptoms of the disorder. As can be seen in Fig. 1, all three strategies converge on a common experimental endpoint, where the effect of the manipulation is evaluated. Without a way to evaluate the effect of genetic manipulations, genetic malleability of a model organism, no matter how advanced, becomes irrelevant. This creates a problem for complex, heterogenic disorders, such as schizophrenia and autism that have symptoms that are hard to measure in animal models. So far, the usefulness of Drosophila for studying neurological and psychiatric disorders is limited by the availability of good neurophysiological or behavioural tests. This second lynchpin, the lack of relevant tests, limits the perceived usefulness of fruit flies for studying complex psychiatric disorders. To some extent, Drosophilists are to blame for underselling their model in this regard, by reverting too often to three trusted standards – locomotion, circadian rhythms, and olfactory learning – rather than adding new behavioural phenotypes to widen the scope of research in modelling human diseases of the brain.

In the following section, we will discuss Drosophila approaches to two very different categories of disease, the relatively simple case of Fragile X, contrasted to the more complex conditions such as schizophrenia, autism, or ADHD. The study of Fragile X syndrome (FXS) is a representative example that illustrates the effective use of Drosophila to understand a neurological disorder and develop potential interventions.