Argonaute-2-dependent rescue of a Drosophila model of FXTAS by FRAXE premutation repeat

Transcription

1 Argonaute-2-dependent rescue of a Drosophila model of FXTAS by FRAXE premutation repeat Oyinkan A. Sofola 1, Peng Jin 2, Juan Botas 1,{ and David L. Nelson 1, *,{ 1 Department of Molecular and Human Genetics, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA and 2 Department of Human Genetics, Emory University School of Medicine, Atlanta, GA 30322, USA Received May 10, 2007; Revised and Accepted July 8, 2007 Human Molecular Genetics, 2007, Vol. 16, No doi: /hmg/ddm186 Advance Access published on July 17, 2007 Fragile X Syndrome is the most common form of hereditary mental retardation. It is caused by a large expansion of the CGG trinucleotide repeat (>200 repeats) in the 5 0 -untranslated region (UTR) of the FMR1 gene that leads to silencing of its transcript. Individuals with CGG repeat expansions approximately between 60 and 200 are referred to as premutation carriers. Fragile X-associated tremor and ataxia syndrome (FXTAS), an RNA-mediated neurodegenerative disease has been described in up to 50% of males carrying premutation alleles. FRAXE, the most common form of non-syndromic X-linked mental retardation, is caused by expansion of a CCG trinucleotide repeat (>200) in the 5 0 -UTR of the FMR2 gene. While the FRAXE premutation length repeat is observed in the general population, there has not yet been a report of a neurodegenerative phenotype associated with these alleles. In this study, we show that the CCG premutation length repeat leads to an RNA-mediated neurodegenerative phenotype in a Drosophila model. Furthermore, we show that co-expression of both the CCG and CGG-containing RNAs suppresses their independent toxicity and is dependent on the RNAi pathway. These data support the concept that RNA toxicity is the mechanism of neuronal toxicity and suggests potential reversal of RNA-mediated phenotypes with complementary RNA molecules. INTRODUCTION Mental retardation is a common disorder that affects about 1.5 3% of the human population worldwide (1). Two folatesensitive fragile sites in the chromosomal region Xq27.3- Xq28, FRAXA and FRAXE result in mental retardation (1,2). In the FRAXA site, large expansions of the CGG trinucleotide repeat (.200 repeats) in the 5 0 -untranslated region (UTR) of the FMR1 gene lead to hypermethylation that prevents transcription and results in the subsequent loss of the FMR1 product, FMRP (3). Fragile X Syndrome (FXS) occurs in individuals lacking the fragile X mental retardation protein (FMRP). It is the most common form of hereditary mental retardation, with a prevalence of 1 in 4500 males and 1 in 9000 females (3). In addition to the learning impairment, these patients are hyperactive, have a decreased attention span and speech impairment (4,5). The FMR1 CGG repeat is polymorphic in the human population, ranging from 5 to nearly 60 repeats in the general population, while individuals with repeat expansions between 60 and 200 are referred to as premutation carriers (3). These premutation carriers are phenotypically normal with respect to the features of FXS but are predisposed to passing a full mutation allele to their offspring (3). However, premature ovarian failure and a neurodegenerative disorder, fragile X-associated tremor/ ataxia syndrome (FXTAS) have been described in premutation adult carriers (6,7). FXTAS is characterized by tremor, gait problems, cerebellar dysfunction, cognitive decline and parkinsonism associated with generalized brain atrophy (8). Patients with FXS do not display FXTAS, and there is an increased expression of the FMR1 mrna in premutation carriers (9,10). These findings led to the proposal that FXTAS results from elevated levels of FMR1 transcripts with extended CGG repeats (6). Subsequently, a transgenic fly model expressing the 5 0 -UTR of the human FMR1 gene with 90 CGG repeats showed that the premutation length repeat in the context of a heterologous transcript could cause neuronal degeneration (11). The FMR2 gene is located approximately 600 kb distal to the FMR1 gene and is implicated in non-syndromic X-linked mental retardation, FRAXE (1,2,12,13). FRAXE is the most common form of non-syndromic X-linked mental retardation; *To whom correspondence should be addressed. Tel: þ ; Fax: þ ; nelson@bcm.tmc.edu { Juan Botas and David L. Nelson should be considered co-senior Authors. # The Author Published by Oxford University Press. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 Human Molecular Genetics, 2007, Vol. 16, No patients with FRAXE have a much milder form of mental retardation in comparison to FRAXA/FXS, and the incidence of disease is much lower, approximately 1 in males (12). Large expansions of a trinucleotide repeat, CCG (.200 repeats) in the 5 0 -UTR of FMR2 lead to transcriptional silencing of the gene and consequently the loss of FMR2 protein in FRAXE (1). Patients with other lesions in FMR2 exhibit similar phenotypes, although controversy remains regarding the role of FMR2 in disease (14). FMR2 knockout mice have impaired learning and memory performance, supporting the protein s role in the human disorder (12). FRAXE patients have a wide array of phenotypes that are similar to FXS patients that include attention-deficit hyperactive disorder and speech impairment (15), and microdeletions within the FMR2 gene may also be implicated in premature ovarian failure (16). To our knowledge, there has been no report of a neurodegenerative phenotype associated with the CCG premutation repeat. This is in contrast to the CGG premutation. However, a correlation between intermediate length alleles (approximately 40 CCG repeats) at the FRAXE locus and Parkinson s disease has been made in an Italian population (17). Given the similarities in phenotype between the FMR1 and FMR2 disease-associated loci, such as mental retardation and premature ovarian failure, and the association of FMR2 intermediate alleles with Parkinson s disease, we hypothesized that the CCG premutation length repeat leads to an RNA-mediated neurodegenerative phenotype. Based on the accumulating data on the potential of small interfering RNAs (sirnas) as therapeutic agents in neurodegenerative disorders (18), we hypothesized that co-expression of both the CCG and CGG RNA would suppress the FRAXA CGG toxicity. The antisense RNA could mask, edit or degrade its target sense RNA (19). The mechanism(s) of antisense regulation is unknown, however, due to the role of RNA interference (RNAi) in double-stranded RNA (dsrna) regulation, we were interested in testing the role of the RNAi pathway in the suppression of the phenotype observed in our fly model. In Drosophila, evidence suggests that the sirna-triggered mrna degradation makes use of Dicer-1/ Dicer-2 and Argonaute-2 while the mirna makes use of Dicer-1 and Argonaute-1 (20,21), providing the potential to distinguish pathways via genetic analyses. Here, we describe a transgenic Drosophila model of the FRAXE CCG90 premutation length repeat. We show that expression of the CCG premutation allele leads to a neurodegenerative phenotype similar to that found in FRAXA CGG premutation models. We further show that co-expression of both the CCG90 RNA and CGG90 RNA in Drosophila rescues their independent neuronal degeneration and demonstrate that the RNAi machinery is required for this rescue. RESULTS Generation of CCG90-EGFP transgenic flies We generated a transgenic Drosophila model expressing ribo (CCG) 90. The CCG90-EGFP construct was made by reversing the previously described FMR UTR CGG90 construct which contains two interrupting AGG triplets [(11), Fig. 1A]. This fragment was cloned into the puast vector upstream of enhanced green fluorescent protein (EGFP). PCR and sequencing were performed with DNA isolated from the transgenic flies to ensure that the construct was integrated into the Drosophila genome, that the repeat size and composition (including two interrupting CCT triplets) was maintained and that no upstream alternative translation start site was present (data not shown). The expression of the 90 CCG repeats can be directed in a spatial and temporal manner using the UAS/GAL4 system (22). We directed expression of the CCG90-EGFP transgene to the fly eye using the GMR-GAL4 driver. Six transgenic lines were generated and the three lines that presented with an external eye phenotype were used for further study. RT PCR analysis demonstrates that the transgenic flies express the 90 CCG repeats and EGFP (Fig. 1B); and western blot data show EGFP levels, with line 14A having the lowest expression and lines 14J and 20J having comparable and higher levels of expression (Fig. 1C), expression levels were also confirmed with semi-quantitative RT PCR (data not shown). Translation of the repeats was not detected as EGFP produced from repeat-containing lines and EGFP alone migrate to the same gel position, indicating an identical molecular weight. rccg repeats are toxic in a dosage-dependent manner To determine whether the rccg repeat constructs cause toxicity, we drove the expression of the different constructs with various GAL4 drivers. We found that in accordance with the levels of GFP expression observed on the western blots, all three lines presented with different degrees of phenotypes, as described in Table 1. Expression of the 90 CCG repeats with Actin5C-GAL4, a ubiquitous driver at 298C causes lethality in all lines (Table 1). Expression with GMR-GAL4, an eye driver, results in a disorganized eye phenotype with line 14A having the mildest phenotype and 14J and 20J exhibiting a more severe eye phenotype (Table 1 and Fig. 2A). The scanning electron microscope (SEM) images reveal compressed ommatidia in all the CCG expressing lines, and in the higher expressing lines (14J and 20J) there is also ommatidia fusion and disorganized or missing inter-ommatidia bristles (Table 1 and Fig. 2A). Transverse sections of the eyes show severe disruption and detachment of the retina and loss of photoreceptor cells, a phenotype that was most evident in the higher expressing lines (Table 1 and Fig. 2A). We note that this eye phenotype is not modified by heat shock protein 70 (Supplementary Material, Figure S1). Expression of the CCG repeats with MS1096-GAL4, which drives expression in the wing pouch, causes a crumpled wing phenotype in the two higher expressing lines only, and is observed only in the male flies of line 20J (Table 1 and Fig. 2B). The control flies, which express the EGFP construct without the CCG repeats do not show pathology with any of the drivers tested. To investigate whether expression of the rccg repeats leads to a neurodegenerative phenotype, we used c164-gal4 to direct expression to motor neurons. Expression of the repeats with this driver causes lethality in the higher expressing lines but 14A, the lower expressing line is viable. This line was used to study the motor performance of flies following expression of the rccg repeats in motor neurons. We used the climbing assay which allows a quantitative measure

3 2328 Human Molecular Genetics, 2007, Vol. 16, No. 19 Figure 1. Expression of (CCG) 90 -EGFP. (A) Schematic representation of the construct generated to express a non-coding CCG90 repeat upstream of EGFP. (B) RT-PCR analysis reveals the presence of transcripts containing both CCG90 (top) and EGFP (bottom) sequences. (C) Western blot analysis confirms protein expression and translation of EGFP. Note same apparent molecular weight as control GFP confirming the absence of a translational start site upstream of the repeats. GMR-GAL4 was used to drive expression at 278C. of the motor ability of flies as a function of age. As shown in Figure 2C, flies expressing GFP alone (control flies) maintain normal motor performance ability until very late (22 days) in their life span. In contrast, the flies expressing rccg repeats (line 14A) decline in their motor performance ability earlier (16 days). These data indicate that the expression of rccg repeats leads to a progressive motor dysfunction phenotype in these flies. rccg and rcgg-independent toxicity is reversed and is RNAi-dependent The rcgg transgenic fly model for FXTAS was previously described and has been shown to display a neurodegenerative eye phenotype (11). Here we show that the rccg repeats can cause a neurodegenerative phenotype in Drosophila. While each construct alone results in eye pathology, co-expression of the two transgenes suppresses the disorganized external eye phenotype and the degeneration of the photoreceptor cells suggesting a mechanism of rescue by the two RNAs together (Fig. 3A). To determine the mechanism of suppression, we investigated whether co-expression of the rcgg and rccg repeats either hindered translation or caused transcript degradation. We confirmed that the CCG and CGG lines had comparable levels of expression by RT PCR (data not shown). Consequently, we examined the expression levels of CCG transcript by RT PCR and EGFP protein by Western blot analysis. We found that both transcript and protein levels are reduced in flies co-expressing rcgg and rccg in comparison to flies expressing rccg alone (Fig. 3B and 3C). The observed reduction of EGFP protein and transcript levels suggests that co-expression of rcgg and rccg may cause blockage of transcription or result in the formation of double stranded RNA (dsrna) that is then degraded. We hypothesized that it led to RNA degradation, and therefore investigated whether the sirna pathway was being utilized. Using the genetic tools, the Drosophila system provides, we generated flies co-expressing both rcgg and rccg in an Ago2 null background to inhibit the sirna pathway and thus dsrna degradation. Interestingly,

4 Human Molecular Genetics, 2007, Vol. 16, No Table 1. Phenotypic characterization of CCG repeats in different tissues of Drosophila GAL4 line Expression pattern EGFP CCG 90 -EGFP14A CCG 90 -EGFP14J CCG 90 -EGFP20J gmr-gal4 Eye cells No effect Compressed ommatidia Disorganized bristles compressed ommatidia Disorganized bristles fused ommatidia actin5c-gal4 Ubiquitous No effect Lethal Lethal Lethal c164-gal4 Motor neurons No effect Motor dysfunction phenotype Lethal Lethal ms1096-gal4 Dorsal wing disc No effect No effect Crumpled wings Crumpled wings Figure 2. Expression of (CCG) 90 -EGFP causes a neurodegenerative eye phenotype. (A) Top SEM images (magnification is 200 for whole eye and 600 for close-up ommatidia images) and bottom transverse retinal sections of eyes (magnification is 250). From left to right, GMR-GAL4/UAS:EGFP, GMR-GAL4/ UAS:CCG 90 -EGFP #14A, GMR-GAL4/UAS:CCG 90 -EGFP #14J, GMR-GAL4/UAS:CCG 90 -EGFP #20J. Note compressed and fused ommatidia, disorganized and/or missing inter-ommatidial bristles, detached retinas and tissue loss in CCG expressing flies in comparison with control flies (UAS:EGFP). Flies were grown at 278C. (B) Over-expression of rccg 90 also causes a wing phenotype (magnification is 40). Crumpled wings are seen in CCG-expressing flies, lines 14J and 20J. The crosses were performed at 278C, and MS1096-GAL4 was used. (C) Over-expression of rccg 90 causes a motor dysfunction phenotype, compare flies expressing (CCG) 90 -EGFP14A to flies expressing EGFP only. The graph shows two sets of experiments (each set consisting of control flies which express EGFP and experimental flies which express 90 CCG repeats), which were carried out simultaneously. The crosses were performed at 298C, and c164-gal4 was the driver used. Indicates that comparing time-points from days 18 to 24 using a t-test results in P-values, for each pair. We compared 1aGFP to 1aCCG and 1a2GFP to 1a2CCG. Bars represent standard deviations of 10 trials per time-point. we found that in an Ago2 null background, flies co-expressing both rcgg and rccg do not revert their independent disorganized eye phenotypes. In fact, these flies were lethal in the pupal stage and had to be dissected from the pupal case (genotype was confirmed by PCR, data not shown) in order to examine the eyes of these flies. They were found to exhibit a much more severe eye phenotype than flies expressing either rcgg or rccg alone (Fig. 3D). To control for the possibility of the Ago2 null condition enhancing the repeat phenotypes, we characterized flies with complete loss

5 2330 Human Molecular Genetics, 2007, Vol. 16, No. 19 Figure 3. CCG 90 and CGG 90 independent eye phenotypes are rescued via an antisense mechanism that is dependent on the RNAi machinery. (A) Light microscopy images, SEM images and transverse retinal sections of eyes from flies expressing transgenes. From left to right, GMR-GAL4 UAS:CGG 90 - EGFP/UAS:EGFP, GMR-GAL4 UAS:CCG 90 -EGFP/ UAS:EGFP, GMR-GAL4 UAS:CGG 90 -EGFP/ UAS:CCG 90 -EGFP. Note that co-expression of rcgg 90 and rccg 90 reverses their independent-mediated toxicity. The crosses were performed at 278C. (B) RT-PCR and (C) western analysis show a reduction in CCG transcript and EGFP expression, respectively, in flies co-expressing rccg 90 -EGFP and rcgg 90 -EGFP in comparison with flies expressing rccg 90 -EGFP alone; we do not compare with CGG-expressing flies due to the nature of these repeats to impede translation in cis (32). The crosses were performed at 278C. (D) Fly co-expressing rccg 90 -EGFP and rcgg 90 -EGFP in an Ago2 heterozygous background, GMR-GAL4 UAS:CGG 90 -EGFP/ UAS:CCG 90 -EGFP; Ago2 414 /Tm6 has a wild-type eye but in an Ago2 homozygous null background, GMR-GAL4 UAS:CGG 90 -EGFP/UAS:CCG 90 -EGFP Ago2 414 /Ago2 414, the eye is disorganized. Control Ago2 null, GMR-GAL4/þ; Ago2 414 /Ago2 414 flies have a wild-type eye. of Ago2. This genetic background does not alter the phenotypes of CGG or CCG expressing flies (Supplementary Material, Figure S2). DISCUSSION Numerous neurodegenerative diseases result from expansion of triplet repeats including Huntington s disease and numerous spinocerebellar ataxias. In most of the cases, the pathogenic mechanisms result from gain of function of either the protein or RNA. FXTAS pathogenesis has been described to result from the RNA-mediated gain-of-function mechanism. Here, we show that the expression of rccg leads to a variety of phenotypes when expressed in different tissues of the fly. We show that expression of the rccg repeats can cause neuronal degeneration similar to the rcgg repeats. Even though individuals carrying the CCG premutation-length repeat have not been reported to present with a neurodegenerative phenotype, it is evident that the rccg repeats are toxic in

6 Human Molecular Genetics, 2007, Vol. 16, No a manner similar to rcgg repeats in our Drosophila model. These data suggest that it might be valuable to investigate pathology in these premutation carriers and in general, carriers of long CCG containing transcripts that are expressed in neurons (potentially from loci other than FMR2). This notion is underscored by the correlation that has been made with the FMR2 intermediate CCG alleles and Parkinson s disease (17). Significantly, we also show that co-expression of rcgg and rccg reverses their independent toxicities through a reduction of their transcript levels. It is important to note that only a portion of the transcripts (the repeats and 300 of flanking base pairs) are complementary while the rest of the transcripts are not, as they each carry the coding sequence for EGFP. A plausible explanation for the observed phenotypic suppression is that the complementary sequences in the 5 0 -UTRs of each RNA lead to dsrna that is processed/cleaved, which results in reduced RNA and protein levels. It is remarkable that these partially complementary RNAs can stimulate this mechanism. It has been reported that over 20% of human genome transcripts might possess naturally occurring antisense transcripts (23) which might be involved in their regulation. Antisense transcripts may also be responsible for other biological processes such as X-inactivation (24). They are also implicated in diseases such as spinocerebellar ataxia type 8 (25). The mechanisms of antisense regulation are mostly unknown and the role of RNAi in this mode of regulation has been debated. It has recently been proposed that natural antisense regulation does not depend on dicer-associated RNAi in mammalian cells (26). These authors were unable to detect RNA duplexes in the cytoplasm of Hela cells over-expressing both sense and antisense vectors. The canonical Dicer products of 21nt were not present in these cells, suggesting that the RNAi pathway is not activated (26). We show in the Drosophila model that the RNAi machinery is involved in the phenotypic suppression observed when both complementary transcripts are expressed. We observe a reduction in both protein and RNA levels and we also demonstrate that depletion of Ago2 from the system abolishes suppression and leads to a more severe phenotype. Interestingly, in this context, it was recently reported that there is an active RNAi machinery in the nucleus (27) opening the possibility of cytoplasmic and/or nuclear degradation of the transcripts. In this study, we show that the CCG RNA causes neurodegeneration, which suggests that the premutation CCG alleles could contribute to the unknown causes of ataxia. We also provide further evidence for the role of CGG RNA in FXTAS since knocking down the CGG transcript with its complementary transcript leads to a reversal of its neurodegenerative eye phenotype. In addition, we demonstrate the dependence of Argonaute-2, an RNAi component in mediating this transcript knock-down. Strategies to reduce target mutant mrna and protein levels have been developed for other neurodegenerative disease models. For example, short hairpin RNAs (shrnas) against ataxin-1 and huntingtin have partially ameliorated pathology and overt motor dysfunction in the SCA1 and HD models (18,28,29). Also, synthetic oligonucleotides composed of repeats are capable of specifically silencing mutant transcripts of complementary repeats in cells (18,28,29). Our data suggest potential therapeutics with complementary RNAs that target mutant mrnas carrying aberrantly sized repeats. METHODS Drosophila genetics EGFP was cloned into the puast vector and the 90 CCG repeats (human FMR1 containing 90 CGG repeats cloned in the opposite orientation) with flanking sequences (300 bp) were cloned upstream of the EGFP cdna. The constructs were confirmed by sequencing and injected by standard methods. The repeat size was confirmed by PCR and sequencing the DNA extracted from the transgenic flies. UAS-CGG90-EGFP was previously described (11), Ago 2 null (Ago2 414 ) stock was a kind gift from Dr Siomi. c164-gal4 is described in this reference (30), all other fly stocks were obtained from the Bloomington Drosophila stock center. All experiments were either carried out at 228C, 278C or at 298C as indicated in the text. PCR, RT PCR and western blot analysis DNA extraction was by standard protocols and PCR was carried out for the CCG repeats with two primer sets, (212F5 0 -GGAACAGCGTTGATCACGTGACGTGGTTTC-3 0 / 571R5 0 -GGGGCCTGCCCTAGAGCCAAGTACCTTGT-3 0 ) and primers (212R5 0 -GAAACCACGTCACGTGATCAACG TTGTTCC/pUASTR5 0 TGCTCCCATTCATCAGTTCC-3 0 ); EGFP-specific primers and rp49specific primers were used. GMR-GAL4 was used to drive the expression of the transgenic flies and adult fly heads were collected to extract RNA and protein. RNA extraction was done using Trizol (Invitrogen) and protein extraction with laemili buffer (Biorad) and betamecarpthoethanol (Sigma). Primary antibody, rabbit anti-g FP was used at 1:5000 (Sigma) and secondary antibody goat anti-rabbit was used at 1:5000 (Biorad). GMR-CGG90/CCG 90;AGO2 414 /AGO2 414 is lethal and was dissected at the pupal stage and genotype confirmed by PCR. Histology For SEM images, whole flies were dehydrated in ethanol and analyzed. For adult eye transverse sections, adult heads were fixed, dehydrated, sectioned and stained with Harry s hematoxylin. Climbing behavioral assay Thirty flies expressing either EGFP only or CCG90-EGFP in the motor neurons (c164-gal4) were used for the climbing assay adapted from Ganetzky et al. (31). Climbing performance (ability to climb past a 5 cm mark in 18 s) was assessed every other day from day 2 to day 28 after eclosion, and at the same time of the day to minimize circadian rhythm effects. SUPPLEMENTARY MATERIAL Supplementary Material is available at HMG Online.

7 2332 Human Molecular Genetics, 2007, Vol. 16, No. 19 ACKNOWLEDGEMENTS The authors would like to thank Dr Jim Barrish for the help with SEM and members of Jin, Nelson and Botas laboratories for assistance. D.L.N. is supported by NIH grant RO1 HD and the BCM-Emory Fragile X Research Center, J.B. is supported by NIH grant NS42179 and both by the BCM Mental Retardation and Developmental Disabilities Research Center P50 HD Conflict of Interest statement. None declared. REFERENCES 1. Gu, Y. and Nelson, D.L. (2003) FMR2 function: insight from a mouse knockout model. Cytogenet. Genome Res., 100, Gu, Y., Shen, Y., Gibbs, R.A. and Nelson, D.L. (1996) Identification of FMR2, a novel gene associated with the FRAXE CCG repeat and CpG island. Nat. Genet., 13, O Donnell, W.T. and Warren, S.T. (2002) A decade of molecular studies of fragile X syndrome. Annu. Rev. Neurosci., 25, Bakker, C.E. and Oostra, B.A. (2003) Understanding fragile X syndrome: insights from animal models. Cytogenet. Genome Res., 100, Cornish, K., Sudhalter, V. and Turk, J. (2004) Attention and language in fragile X. Ment. Retard. Dev. Disabil. Res. Rev., 10, Hagerman, R.J. and Hagerman, P.J. (2002) The fragile X premutation: into the phenotypic fold. Curr. Opin. Genet. Dev., 12, Oostra, B.A. and Willemsen, R. (2003) A fragile balance: FMR1 expression levels. Hum. Mol. Genet., 12 (Spec No. 2), R249 R Hagerman, R.J., Leehey, M., Heinrichs, W., Tassone, F., Wilson, R., Hills, J., Grigsby, J., Gage, B. and Hagerman, P.J. (2001) Intention tremor, parkinsonism, and generalized brain atrophy in male carriers of fragile X. Neurology, 57, Kenneson, A., Zhang, F., Hagedorn, C.H. and Warren, S.T. (2001) Reduced FMRP and increased FMR1 transcription is proportionally associated with CGG repeat number in intermediate-length and premutation carriers. Hum. Mol. Genet., 10, Tassone, F., Hagerman, R.J., Taylor, A.K., Gane, L.W., Godfrey, T.E. and Hagerman, P.J. (2000) Elevated levels of FMR1 mrna in carrier males: a new mechanism of involvement in the fragile-x syndrome. Am. J. Hum. Genet., 66, Jin, P., Zarnescu, D.C., Zhang, F., Pearson, C.E., Lucchesi, J.C., Moses, K. and Warren, S.T. (2003) RNA-mediated neurodegeneration caused by the fragile X premutation rcgg repeats in Drosophila. Neuron, 39, Gecz, J. (2000) The FMR2 gene, FRAXE and non-specific X-linked mental retardation: clinical and molecular aspects. Ann. Hum. Genet., 64, Gecz, J., Gedeon, A.K., Sutherland, G.R. and Mulley, J.C. (1996) Identification of the gene FMR2, associated with FRAXE mental retardation. Nat. Genet., 13, Gecz, J. (2000) FMR3 is a novel gene associated with FRAXE CpG island and transcriptionally silent in FRAXE full mutations. J. Med. Genet., 37, Gu, Y., McIlwain, K.L., Weeber, E.J., Yamagata, T., Xu, B., Antalffy, B.A., Reyes, C., Yuva-Paylor, L., Armstrong, D., Zoghbi, H. et al. (2002) Impaired conditioned fear and enhanced long-term potentiation in Fmr2 knock-out mice. J. Neurosci., 22, Murray, A., Webb, J., Dennis, N., Conway, G. and Morton, N. (1999) Microdeletions in FMR2 may be a significant cause of premature ovarian failure. J. Med. Genet., 36, Annesi, G., Nicoletti, G., Tarantino, P., Cutuli, N., Annesi, F., Marco, E.V., Zappia, M., Morgante, L., Arabia, G., Pugliese, P. et al. (2004) FRAXE intermediate alleles are associated with Parkinson s disease. Neurosci. Lett., 368, Denovan-Wright, E.M. and Davidson, B.L. (2006) RNAi: a potential therapy for the dominantly inherited nucleotide repeat diseases. Gene Ther., 13, Good, L. (2003) Diverse antisense mechanisms and applications. Cell. Mol. Life Sci., 60, Lee, Y.S., Nakahara, K., Pham, J.W., Kim, K., He, Z., Sontheimer, E.J. and Carthew, R.W. (2004) Distinct roles for Drosophila Dicer-1 and Dicer-2 in the sirna/mirna silencing pathways. Cell, 117, Okamura, K., Ishizuka, A., Siomi, H. and Siomi, M.C. (2004) Distinct roles for Argonaute proteins in small RNA-directed RNA cleavage pathways. Genes Dev., 18, Brand, A.H. and Perrimon, N. (1993) Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development, 118, Chen, J., Sun, M., Kent, W.J., Huang, X., Xie, H., Wang, W., Zhou, G., Shi, R.Z. and Rowley, J.D. (2004) Over 20% of human transcripts might form sense-antisense pairs. Nucleic Acids Res., 32, Ogawa, Y. and Lee, J.T. (2002) Antisense regulation in X inactivation and autosomal imprinting. Cytogenet. Genome Res., 99, Nemes, J.P., Benzow, K.A., Moseley, M.L., Ranum, L.P. and Koob, M.D. (2000) The SCA8 transcript is an antisense RNA to a brain-specific transcript encoding a novel actin-binding protein (KLHL1). Hum. Mol. Genet., 9, Faghihi, M.A. and Wahlestedt, C. (2006) RNA interference is not involved in natural antisense mediated regulation of gene expression in mammals. Genome Biol., 7, R Robb, G.B., Brown, K.M., Khurana, J. and Rana, T.M. (2005) Specific and potent RNAi in the nucleus of human cells. Nat. Struct. Mol. Biol., 12, Harper, S.Q., Staber, P.D., He, X., Eliason, S.L., Martins, I.H., Mao, Q., Yang, L., Kotin, R.M., Paulson, H.L. and Davidson, B.L. (2005) RNA interference improves motor and neuropathological abnormalities in a Huntington s disease mouse model. Proc. Natl Acad. Sci. USA, 102, Xia, H., Mao, Q., Eliason, S.L., Harper, S.Q., Martins, I.H., Orr, H.T., Paulson, H.L., Yang, L., Kotin, R.M. and Davidson, B.L. (2004) RNAi suppresses polyglutamine-induced neurodegeneration in a model of spinocerebellar ataxia. Nat. Med., 10, Torroja, L., Packard, M., Gorczyca, M., White, K. and Budnik, V. (1999) The Drosophila beta-amyloid precursor protein homolog promotes synapse differentiation at the neuromuscular junction. J. Neurosci., 19, Ganetzky, B. and Flanagan, J.R. (1978) On the relationship between senescence and age-related changes in two wild-type strains of Drosophila melanogaster. Exp. Gerontol., 13, Primerano, B., Tassone, F., Hagerman, R.J., Hagerman, P., Amaldi, F. and Bagni, C. (2002) Reduced FMR1 mrna translation efficiency in fragile X patients with premutations. RNA, 8,