|Home | About | Journals | Submit | Contact Us | Français|
Fragile X-associated tremor/ataxia syndrome (FXTAS) is a progressive neurodegenerative disorder recognized in fragile X premutation carriers. Using Drosophila, we previously identified elongated non-coding CGG repeats in FMR1 allele as the pathogenic cause of FXTAS. Here, we use this same FXTAS Drosophila model to conduct a chemical screen that reveals small molecules that can ameliorate the toxic effects of fragile X premutation ribo-CGG (rCGG) repeats, among them several known phospholipase A2 (PLA2) inhibitors. We show that specific inhibition of PLA2 activity could mitigate the neuronal deficits caused by fragile X premutation rCGG repeats, including lethality and locomotion deficits. Furthermore, through a genetic screen, we identified a PLA2 Drosophila ortholog that specifically modulates rCGG repeat-mediated neuronal toxicity. Our results demonstrate the utility of Drosophila models for unbiased small molecule screens and point to PLA2 as a possible therapeutic target to treat FXTAS.
The fragile X mental retardation 1 (FMR1) gene contains a highly polymorphic CGG repeat in the 5′-untranslated region (1). Whereas normal individuals generally possess between 5 and 54 repeats, individuals with more than 200 CGG repeats, referred to as the full mutation, develop fragile X syndrome (FXS) (2). Premutation alleles (55–200 CGG repeats) of the FMR1 gene are known to contribute to the FXS phenotype through genetic instability, as these alleles can expand into the full mutation during germline transmission (3). Within the last decade, fragile X-associated tremor/ataxia syndrome (FXTAS), a late-onset neurodegenerative disorder distinct from the neurodevelopmental disorder, FXS, has been recognized mainly among many male premutation carriers in or beyond their fifth decade of life (4).
The most common clinical feature of FXTAS is a progressive tremor with ataxia. More advanced or severe cases may show a progressive cognitive decline that ranges from executive and memory deficits to dementia (5). Patients may also present with common psychiatric symptoms, such as increased anxiety, mood liability and depression (6,7). Magnetic resonance imaging of adult male patients affected with FXTAS shows mild to moderate global brain atrophy, most common in the fontal and parietal regions, as well as the pons and the cerebellum (8). Nearly, all autopsy studies on the brains of symptomatic premutation carriers show degeneration in the cerebellum, which includes Purkinje neuronal cell loss, Bergman gliosis, spongiosis of the deep cerebellar white matter and swollen axons (9,10). The major neuropathological hallmark and postmortem criterion for definitive FXTAS is eosinophilic, ubiquitin-positive intranuclear inclusions broadly distributed throughout the brain in neurons, astrocytes and in the spinal column (9). Although it is mainly male premutation carriers who are affected, female premutation carriers can also experience the same neurological symptoms, but with less severity, which is likely due to the FMR1 gene being located on the X chromosome and random X-chromosome inactivation. So far, there is no effective therapeutic intervention for FXTAS.
An RNA gain-of-function mechanism has been suggested as the culprit in FXTAS based on the observation of increased levels of CGG-containing FMR1 mRNA, along with either no detectable change in FMRP or slightly reduced FMRP levels in premutation carriers (8,10–13). The lack of FXS, which results from the loss of function of the FMR1 gene product, in FXTAS patients, along with the absence of FXTAS symptoms in older individuals with FXS, suggest a role for the expanded ribo-CGG (rCGG) repeat in FXTAS pathology. Using in situ hybridization, Tassone et al. (14) demonstrated the presence of expanded FMR1 RNA transcripts in the FXTAS inclusions of a 70-year-old male who died with FXTAS.
In addition to the increased levels of CGG-containing FMR1 mRNA seen in fragile X premutation carriers, several lines of evidence further support an RNA-mediated gain-of-function toxicity model for FXTAS. First, in a ‘knock-in’ mouse model, in which the endogenous CGG repeats (five CGG repeats in the wild-type mouse Fmr1 gene) were replaced with a ~100 CGG repeat fragment, intranuclear inclusions were found to be present throughout the brain, with the exception of cerebellar Purkinje cells (13). An increase in both the number and size of the inclusions was seen during the life course, which correlates with the progressive character of the phenotype observed in humans (9). Secondly, neuropathological studies in humans have revealed a highly significant association between length of the CGG tract and frequency of intranuclear inclusions in both neurons and astrocytes, indicating that the CGG repeat length is a powerful predictor of neurological involvement clinically (age of death), as well as neuropathologically (number of inclusions) (9). Thirdly, intranuclear inclusions can be formed in both primary neural progenitor cells and established neural cell lines, as was revealed using a reporter construct with an FMR1 5′-UTR harboring expanded (premutation) CGG repeats (15). Finally, we have described a model of FXTAS using Drosophila expressing the FMR1 untranslated-CGG repeats 5′ to the enhanced green fluorescent protein (EGFP) coding sequence and demonstrated that premutation-length rCGG repeats are both toxic and sufficient to cause neurodegeneration (11). These observations led us to propose that transcription of the CGG90 repeats leads to an RNA-mediated neurodegenerative disease.
Within the last decade, Drosophila has emerged as a premiere model system for the study of human neurodegenerative diseases, due to the realization that flies and humans share many structurally and functionally related gene families (16–18). Development of such disease models in the fly allows genetic approaches to be applied to address specific hypotheses concerning disease progression and to test candidate modifier genes or therapeutic drug compounds (16–18). Here, we used our FXTAS Drosophila model to conduct a chemical screen and identified the small molecules that can ameliorate the toxic effects of fragile X premutation rCGG repeats, among them several known phospholipase A2 (PLA2) inhibitors. We show that specific inhibition of PLA2 activity could indeed suppress the neuronal toxicity caused by fragile X premutation rCGG repeats. An additional genetic screen led us to identify a PLA2 Drosophila ortholog that could specifically modulate rCGG repeat-mediated neuronal toxicity. Our results reveal a previously unrecognized pathway and druggable target for FXTAS and highlight the general potential of using Drosophila for unbiased chemical screens in drug development.
We have previously reported a Drosophila disease model of FXTAS and provided experimental evidence that fragile X premutation rCGG repeats alone are sufficient to cause neurodegeneration in a repeat-dosage manner (11). We discovered that flies with modest expression of fragile X premutation rCGG90 repeats exclusively in the neurons do not reach adulthood. Lethality occurs primarily during embryonic development before larval formation. Taking advantage of this lethality, we designed a high-throughput strategy to screen chemical libraries for small molecules relevant to FXTAS. We first balanced the UAS-CGG90-EGFP transgene with a chromosome carrying a curly wing (Cyo) marker recombined with a heat shock-inducible apoptotic hid (hs-hid) transgene (w1118, UAS-CGG90-EGFP/TM2 Cyo:hs-hid). Next, we tested the viability of adult progeny using the Drosophila GAL4/UAS system. We performed crosses of UAS-CGG90-EGFP/TM2 Cyo:hs-hid with homozygous pan-neuronal elav-Gal4 driver. As expected, there was an absence of heterozygous non-Cyo adult progeny (elav-GAL4/+; UAS-CGG90-EGFP/+), which confirmed that rCGG90 repeat expression in neurons leads to lethality. The only living progenies would carry elav-GAL4/+; +/TM2 Cyo:hs-hid. Furthermore, upon heat shock, the progenies carrying elav-GAL4/+; +/TM2 Cyo:hs-hid also died, because the expression of hid leads to cell death and lethality at embryonic stages that are otherwise viable.
Since the rescue of the lethality was an easily scored phenotype, we were able to use this assay to conduct a chemical screen to identify the small molecules that can improve/restore the viability of the flies expressing fragile X premutation rCGG repeats. As depicted in schematic representation (Fig. 1A), we raised crosses between elav-GAL4 and UAS-CGG90-EGFP/TM2 Cyo:hs-hid as well as control flies into food supplemented with 40 μm of individual unique compounds from a library of 2000 US Food and Drug Administration (FDA)-approved drugs and natural products (The Spectrum Collection, MicroSource Discovery Systems, Inc.). At day 2, crosses were subsequently heat shocked to eliminate (elav-GAL4/+; +/TM2 Cyo:hs-hid) progenies, and the vials were then kept at 25°C for 10–15 days to score for any of the viable (elav-GAL4/+; UAS-CGG90-EGFP/+) flies that are otherwise lethal. Among the 2000 compounds initially screened, 58 were found to result in either puparium formation or the emergence of adult non-Cyo progeny (elav-GAL4/+; UAS-CGG90-EGFP/+). Since the puparium formation could also represent dead (elav-GAL4/+; +/TM2 Cyo:hs-hid) progeny simply due to heat shock, we selected the top 35% (20 of 58 compounds) based on the percentage of recovered flies or pupae from the crosses for further validation. Among these 20 compounds, 11 were validated using large-scale viability assays (Fig. 1B). The confirmed compounds belong to several biochemically distinct pathways, with several of them having the potential to target inflammatory pathways (Fig. 1B).
To further validate these compounds, we developed a secondary behavioral assay. In this assay, UAS-CGG90-EGFP transgenic flies were crossed with nervana-GAL4, which could drive the expression of fragile X rCGG repeats in CNS, albeit weakly, allowing us to analyze adult flies for locomotion deficits. The expression of rCGG repeats leads to locomotion defects (Fig. 2). As shown in Figure 2A, we set up the cross between UAS-CGG90-EGFP/TM2 Cyo and nervana-GAL4 in food that was supplemented with each compound identified above. The progeny flies (UAS-CGG90-EGFP/nervana-GAL4) were collected to monitor locomotion. To continuously monitor the locomotion for a given fly, we have established a Drosophila Activity Monitor (DAM) system for this purpose (19). For each individual compound, we simultaneously monitored the locomotion of 32 flies that were fed with fly food containing the compound each time. The data were collected and analyzed for locomotor activity (Fig. 2B). Each drug treatment was repeated at least three times. Through this secondary testing, we found that fluocinolone acetonide and xylazine could significantly ameliorate the locomotion defects induced by fragile X premutation rCGG repeats (Fig. 2B). The other compounds had either mild effects or no effect on locomotion deficits induced by rCGG repeats.
Intriguingly, fluocinolone acetonide is a well-known inhibitor of phospholipase A2 (PLA2) (20). The PLA2 family includes secretory phospholipase A2, cytosolic phospholipase A2, plasmalogen-selective phospholipase A2 and calcium-independent phospholipase A2 (21,22). It is generally thought that the release of arachidonic acid by cytosolic PLA2 is the rate-limiting step in the generation of eicosanoids and platelet-activating factor. These lipid mediators play critical roles in the initiation and modulation of inflammation and oxidative stress (21,22). The tight regulation of PLA2 activity is necessary for maintaining basal levels of arachidonic acid and eicosanoids for performing normal function (21,22). Our finding that fluocinolone acetonide could suppress both lethality and locomotion deficits caused by fragile X premutation rCGG repeats raised the possibility that PLA2 inhibitors in general could suppress rCGG-mediated neurodegeneration. To test this hypothesis, we used other known PLA2 inhibitors for locomotion assays. Cytidine 5'diphospho-choline (citicoline) inhibits cPLA2 activity and lowers the concentration of free fatty acids in a dose- and time-dependent manner (23). Likewise, release of arachidonic acid can be blocked by quinacrine and arachidonyl trifluoromethylketone (Fig. 3) (24,25). We observed significant improvements in locomotion in flies expressing rCGG repeats for all the PLA2 inhibitors, whereas these compounds had no significant impact on the locomotor activity of wild-type flies (Fig. 3). These observations strongly suggest that PLA2 activity is altered by fragile X premutation rCGG repeats, and PLA2 inhibitors could suppress rCGG-mediated neurodegeneration.
Given our findings from the chemical screen, we sought to harness the power of Drosophila genetics to explore the role of PLA2 isoforms in mediating fragile X premutation rCGG repeat-mediated neuronal toxicity. In Drosophila, the genes for some PLA2 isoforms have been mapped, but none has been cloned and fully characterized; therefore, to examine the role of the PLA2 pathway in FXTAS, we used the Kyoto encyclopedia of genes and genomes (KEGG) human database to identify and collect known pathways that are related to lipid metabolism and PLA2 activity (26). We then created reference maps for the well-known PLA2 pathways and looked for the genes and pathways that are related in flies (Table 1). We identified mutations in some of the genes of the pathway and conducted a genetic screen based on the fragile X premutation rCGG repeat-mediated neurodegenerative eye phenotype that we saw previously. The screen involved directing the expression of fragile X premutation-length rCGG repeats to the eye with the gmr-GAL4 driver. This was followed by crossing gmr-GAL4, UAS-(CGG)90-EGFP transgenic flies with flies mutant in genes coding for lipid metabolism and PLA2 activity identified from the reference KEGG map. The progenies were examined for potential suppression or enhancement of the disorganized eye phenotype by comparison with control rCGG flies. Through this screen, we identified an EP line, CG1583EP1516 (a Drosophila homolog of PLA2 gene), that could modulate the rCGG-mediated neurodegeneration (Fig. 4A). Since CG1583EP1516 is an EP insertion that is supposed to activate the transcription of the adjacent gene, we performed quantitative reverse transcription polymerase chain reaction (RT–PCR) using the RNAs isolated from age-matched brains to determine the expression of CG1583 in both CG1583EP1516 and control (w1118) flies. We found that CG1583EP1516 specifically disrupted the CG1583 gene and led to the partial loss of gene expression (Fig. 4B). We also found that an adjacent gene, Traf 6, did not interact genetically with rCGG-mediated neurodegeneration. This modulation was further confirmed using an RNAi UAS line (GIIIsPLA2v50353) that could express dsRNAs against CG1583 in the presence of a GAL4 driver (Table 1).
Since PLA2 has been linked to several neurological disorders (21,22,27–30), we sought to determine whether CG1583 is a specific modifier of rCGG-mediated toxicity. We crossed CG1583EP151 with fly models of other neurodegenerative disorders. Like fragile X premutation rCGG repeats, transgenic flies expressing mutant Huntington's disease (HD), cytosine uracil guanine (CUG) repeats or SCA1 show rough eye phenotype when crossed with gmr-Gal4 driver (18,31,32). As shown in Figure 4A, the partial loss of CG1583 had no effect on the neuronal eye degeneration caused by mutant HD, CUG repeats or SCA1. We further determined the expression of CG1583 using RNAs isolated from the age- and sex-matched brains of w1118 and flies expressing CGG60 repeats in neurons (elav/+; UAS-CGG60-EGFP). We found that fragile X premutation rCGG repeats, but not SCA1, could cause increased expression of CG1583 in fly brains (Fig. 4B). These results together suggest that CG1583, a Drosophila ortholog of PLA2, is a specific modulator of rCGG-mediated neurodegeneration.
FXTAS, a late age of onset neurodegenerative disorder uncoupled from the neurodevelopmental disorder, FXS, has been recognized in older males of FXS families. Several lines of evidence, including ours, have led to the proposal of an RNA-mediated gain-of-function toxicity model for FXTAS, in which rCGG repeat-binding proteins could become functionally limited by their sequestration to lengthy rCGG repeats (33–35). There is currently no effective therapeutic intervention available for FXTAS (36). In recent years, Drosophila has emerged as a premiere model system for the study of human neurodegenerative diseases, due to that flies and humans share many structurally and functionally related gene families (16–18,31,32). Development of fly disease models allows us to address specific hypotheses concerning disease progression, test candidate modifier genes or therapeutic drug compounds and potentially identify novel small molecules via unbiased chemical screens. We have previously established a Drosophila model and shown that fragile X premutation rCGG repeats are sufficient to cause neuronal toxicity (11). Here, we used this same model to conduct an unbiased chemical screen and identified the small molecules that can ameliorate the toxic effects of fragile X premutation rCGG repeats. Specifically, we found that inhibition of PLA2 activity could suppress neuronal toxicity caused by fragile X premutation rCGG repeats. Our results reveal a previously unrecognized pathway and druggable target for FXTAS and highlight the general potential of using Drosophila for unbiased chemical screens in drug development.
The discovery of lead compounds that have the potential to elicit a pharmacological effect is often achieved via the screening of small molecule libraries for interaction with purified proteins, or for an ability to induce a desired physiological response in cultured cells. Since complex biological processes cannot be recapitulated in cell or organ culture, efforts are now underway to perform chemical screens on whole animals. One of the most commonly used animal models in biology is mouse; however, its cost limits its use in large-scale therapeutic screening. Fortunately, extensive research during the past two decades, coupled with the elucidation of annotated genome sequences for human and other eukaryotic organisms, has revealed a remarkable degree of conservation of biochemical pathways. This conservation has made it feasible to consider using invertebrate models as tools for drug discovery and validation (37). Indeed, fly models of human neurodegenerative disorders, such as Huntington's disease, Parkinson's disease (PD) and Alzheimer's disease (AD), have been used to validate candidate compounds for therapeutic interventions (17). However, whether fly models could be used for unbiased chemical screens or to reveal new biological pathway(s) has not been explored. Our results here indicate that a fly model of a human neurological disorder could indeed be used for an unbiased chemical screen. Through our screen, we revealed altered PLA2 activity in the presence of fragile X premutation rCGG repeats, which is particularly interesting, since many PLA2 inhibitors have been or are being developed for therapeutic use.
The biological functions of the compounds identified through our initial screen implicate various mechanisms of action (Fig. 1B). In particular, our results suggest that PLA2 inhibitors could be potent suppressors of rCGG-mediated neuronal toxicity. Indeed, altered PLA2 activity has been linked to several neurodegenerative diseases and brain trauma. A number of neurodegenerative disorders, including AD, PD and multiple sclerosis, as well as acute neural trauma, such as ischemia, spinal cord injury and head injury, have been characterized by inflammatory reactions, oxidative stress, altered phospholipid metabolism, accumulation of lipid peroxides and increased phospholipase A (21,22,27,30). Interestingly, several PLA2 inhibitors active in our assays are known to exert their neurodegeneration-preventive effects by suppressing or lowering transcription of genes for PLA2 isozymes (22). In our FXTAS fly model, we found that CG1583, a PLA2 ortholog, could modulate the neuronal toxicity caused by fragile X premutation rCGG repeats. Furthermore, the total steady-state level of CG1583 mRNA was also elevated in the fly brain expressing fragile X rCGG repeats. These findings suggest that the altered PLA2 activity could directly or indirectly contribute to the molecular pathogenesis of FXTAS, which could make an important therapeutic target for FXTAS.
The remaining compounds identified through our screen may also be viable candidate compounds for drug development. In particular, acetylcarnitine, a cholinergic drug, has been used to prevent neuronal cell death. It has been reported as neuroprotective in instances of cerebral ischemia in rats and may be useful in treating peripheral nerve injury; it may also be of benefit in the treatment of PD (38).
In summary, we have conducted an unbiased chemical screen using the FXTAS fly model that we developed previously. Through this screen, we identified the small molecules that can ameliorate the toxic effects of fragile X premutation rCGG repeats. In particular, we show that specific inhibition of PLA2 activity could indeed suppress neuronal toxicity caused by fragile X premutation rCGG repeats. Our results reveal a previously unrecognized pathway and druggable target for FXTAS and highlight the general potential of using Drosophila for unbiased chemical screens in drug development.
Transgenic flies expressing rCGG90 and rCGG60 repeats were described previously. The Cyo:hs-hid, elav-Gal4, TM2 chromosome balancer fly line and CG1583EP1516 insertions were obtained from Bloomington Stock Centre (Bloomington, IN, USA). Flies of other neurological disease models, including UAS-HD128G/+, UAS-CUG480/+ and UAS-SCA1 82Q/+, were obtained from Juan Botas at Baylor College of Medicine. Other lesions used for genetic interactions as depicted in Table 1 were obtained from either Bloomington or Vienna Drosophila RNAi Center, Austria. All crosses were grown on standard medium at 25°C in 50–70% humidity.
Flies carrying the UAS-CGG90-EGFP transgene over a heat shock-inducible hid-balancer chromosome (Cyo:hs-hid, TM2) were generated for viability assay. Progeny embryos from the crosses between UAS-CGG90-EGFP/Cyo:hs-hid and homozygous pan-neuronal elav-Gal4 driver were subjected to heat shock to ensure hid-induced lethality. The Spectrum Collection compound library was obtained from MicroSource Discovery Systems, Inc. In the initial screen, crosses were performed in small vials with 40 μm of individual drugs. Crosses were kept at 25°C, and the percentages of pupae and adults recovered were scored. Drugs used in larger quantities for confirmation and subsequent analyses were obtained from Sigma. Drugs supplemented with Jazz-Mix commercial food (CF) for the screen were added to the final concentration of 40 μm at 50°C for uniform drug dispersion in the food. For the confirmation of the 11 compounds from the initial screen, viability assays were performed on a larger scale using 5 μm and 40 μm doses of the drug.
UAS-CGG90-EGFP/Cyo transgenic flies were crossed with nervana-Gal4/Cyo in a CF and in CF plus drug from the top 11 compounds. Drugs were used at their optimum dose and mixed with CF at 50°C for homogenous distribution. Homozygous progeny females (UAS-CGG90-EGFP/nervana-Gal4) 48 h of age were collected for locomotor activity assays. To continuously monitor the locomotion, a DAM system (TriKinetics) was used. For each individual compound, we simultaneously monitored the locomotion of 32 flies that were treated. The readings were recorded each time the fly crossed the infrared beam. These interruptions were collected as a bin of 1 h. The activity of flies was calculated as the average number of movements in a given time recorded together with control flies. Throughout the assay, UAS-CGG90-EGFP/nervana-Gal4 females were maintained in a drug in which they were initially raised. Since behavior is often sexually dimorphic, only 48 h old virgin female flies were used in the assay. Data were collected and analyzed for locomotor activity.
For the RNA isolation, adult heads of the required genotype were mashed in Trizol (Gibco BRL Life Technologies) using plastic kontes. After precipitation, RNA was further cleaned with an RNeasy Kit (QIAGEN). For RT–PCR, RNA was reverse-transcribed with random primers using the high-capacity cDNA reverse transcription kit (Applied Biosystems) according to the manufacturer's instructions. Real-time PCR was performed with commercial TaqMan probes obtained against CG1583, and TaqMan Master Mix (Applied Biosystems) using a 7500 Fast Real-Time PCR system (Applied Biosystems). TaqMan probes Dm_RpL32 (Applied Biosystems) were used as internal control.
We used DAM Scan software (TriKinetics) to detect any outliers in the readings obtained. The readings obtained in a bin of 1 h were averaged over 24 h. Recordings for a period of 7 days for each genotype per drug were plotted. Analyses to show significant differences between genotypes/treatments were performed using analysis of variance, post hoc t-tests (two sample assuming equal variances). All data are shown as mean with standard error of mean.
A.Q. was partially supported and is a recipient of a National Ataxia Foundation Postdoctoral Award. D.L.N. and P.J. are supported by NIH grants (R01 NS051630 and R21 NS067461). P.J. is a recipient of the Beckman Young Investigator Award and the Basil O'Connor Scholar Research Award, as well as an Alfred P Sloan Research Fellow in Neuroscience.
The authors would like to thank J. Taylor of The Integrated Microscopy and Microanalytical Facility for her help with SEM, Juan Botas for sharing the fly lines, the members of the Jin lab for their assistance and C. Strauss for critical reading of the manuscript.
Conflict of Interest statement. None declared.