A Drosophila model for TDP-43 proteinopathy.
Journal: 2010/May - Proceedings of the National Academy of Sciences of the United States of America
ISSN: 1091-6490
Neuropathology involving TAR DNA binding protein-43 (TDP-43) has been identified in a wide spectrum of neurodegenerative diseases collectively named as TDP-43 proteinopathy, including amyotrophic lateral sclerosis (ALS) and frontotemporal lobar dementia (FTLD). To test whether increased expression of wide-type human TDP-43 (hTDP-43) may cause neurotoxicity in vivo, we generated transgenic flies expressing hTDP-43 in various neuronal subpopulations. Expression in the fly eyes of the full-length hTDP-43, but not a mutant lacking its amino-terminal domain, led to progressive loss of ommatidia with remarkable signs of neurodegeneration. Expressing hTDP-43 in mushroom bodies (MBs) resulted in dramatic axon losses and neuronal death. Furthermore, hTDP-43 expression in motor neurons led to axon swelling, reduction in axon branches and bouton numbers, and motor neuron loss together with functional deficits. Thus, our transgenic flies expressing hTDP-43 recapitulate important neuropathological and clinical features of human TDP-43 proteinopathy, providing a powerful animal model for this group of devastating diseases. Our study indicates that simply increasing hTDP-43 expression is sufficient to cause neurotoxicity in vivo, suggesting that aberrant regulation of TDP-43 expression or decreased clearance of hTDP-43 may contribute to the pathogenesis of TDP-43 proteinopathy.
Similar articles
Articles by the same authors
Discussion board
Proc Natl Acad Sci U S A 107(7): 3169-3174

A <em>Drosophila</em> model for TDP-43 proteinopathy


Generation of Transgenic Flies Expressing Human TDP-43.

To study human TDP-43 (hTDP-43) in vivo, we used Drosophila, a powerful genetic model widely used to study neurodegeneration (17, 18). We generated transgenic flies expressing monomeric red fluorescent protein (RFP) as a control or hTDP-43 fused to RFP in different populations of neurons using UAS/Gal4 system (19) (Fig. S1C). We also generated transgenic flies expressing a mutant hTDP-43, T202, containing the carboxyl-terminal glycine-rich domain but lacking the amino-terminal RNA recognition motif, because expression of this mutant led to neuronal death in vitro, consistent with a recently published study (16). Both RT-PCR and Western analyses confirmed hTDP-43 expression, consistent with RFP signals detected in the eye discs of hTDP-43-RFP larvae (Fig. S1).

To test whether TDP-43 proteinopathy can be caused by insufficient TDP-43 expression, we also generated flies in which the expression of the endogenous fly homolog, Tar DNA binding protein homolog (TBPH), was reduced by RNA interference (RNAi) with two fly lines (numbers 38377 and 38379). Using TBPH-specific primers, we demonstrated that TBPH was significantly reduced in TBPH-RNAi-expressing flies as compared to the control flies (Fig. S1).

Expression of hTDP-43 in the Drosophila Eye Leads to Progressive Retinal Degeneration.

We used a GMR-Gal4 driver to express hTDP-43 specifically in Drosophila eyes beginning at the third instar larval stage. RFP control flies showed normal eye morphology (Fig. 1A), whereas flies expressing hTDP-43 began to show obvious ommatidia loss at eclosion with different levels of severity (Fig. 1B). Although there were significant variations in the eye defects among individual flies of the same age and the same genetic background, hTDP-43-expressing flies showed progressive degeneration with aging (Fig. 1C). Such retinal degeneration phenotype was not detectable in flies expressing the hTDP-43 mutant T202. We also investigated effects of decreasing expression of the endogenous fly TBPH on eye morphology using TBPH-RNAi flies. No obvious eye defect was found in these TBPH RNAi flies, consistent with a recent study in TBPH knockout flies (20).

An external file that holds a picture, illustration, etc.
Object name is pnas.0913602107fig01.jpg

Expression of hTDP-43 in fly eyes leads to progressive eye defects in an age-dependent manner. (A and B) Eye morphology of one control fly (A, GMR-Gal4/UAS-RFP) and hTDP-43 expressing flies (B, GMR-Gal4/UAS-hTDP-43-RFP). GMR-Gal4/UAS-hTDP-43-RFP eyes exhibit ommatidia loss and necrotic lesions. B1B4 show hTDP-43-expressing fly eyes with ommatidia loss at different levels. (C) Quantification of the eye defects in hTDP-43-expressing flies during aging, showing significant difference from that in the control RFP-expressing flies (P < 0.0001 in χ test). Eye defects were classified into four levels: 1, <25% ommatidia loss; 2, 25–50% ommatidia loss; 3, 50–75% ommatidia loss with small regions of necrosis (appearing as black dots); 4, >75% ommatidia loss with massive regions of necrosis. The percentages of flies with each level of eye defects are shown in different colors (1, blue; 2, red; 3, yellow; 4, cyan). By day 6, most flies expressing hTDP-43 showed moderate to severe eye defects. (DF) Retinal sections stained with toluidine blue showed intact ommatidial and rhabdomere structure. (D) GMR-Gal4/UAS-RFP control fly eye (high magnification shown in Inset). Rhabdomere structures were disintegrated, and ommatidia arrangement was disrupted in GMR-Gal4/UAS-hTDP-43-RFP fly eyes (E) but not in GMR-Gal4/UAS-hTDP-43T202-RFP fly eyes (F). The arrow in E marks a large vacuole in the eye. (Scale bars: 100 μm.)

We examined eye structures of hTDP-43 transgenic flies in frontal sections after toluidine blue staining. Both the RFP-expressing control flies and flies expressing hTDP-43 mutant T202 showed intact rhabdomere structure throughout their lifespan (Fig. 1 D and F). In contrast, the ommatidia organization was completely lost in hTDP-43 transgenic flies, with large vacuoles frequently detected (arrow in Fig. 1E), although their cone cells remained largely normal. Using electron microscopy, we compared the ultrastructure of fly eyes expressing hTDP-43 with controls (Fig. 2). Well organized rhabdomeres (Rh) were seen in the control group of both young (1 day) and old (30 day) flies (Fig. 2 A and B). However, the normal rhabdomere arrangement was disrupted in hTDP-43 flies as early as day 1, with very few rhabdomeres detected (Fig. 2C). By day 30, the rhabdomere structure was completely lost in hTDP-43 flies with morphological features of neurodegeneration including degenerating mitochondria (as marked by the asterisk), multilamellar bodies (MLB), multivesicular bodies (MVB), and autophagic vacuoles (AV) (Fig. 2 DF). However, flies expressing mutant hTDP-43, T202, show the normal ultrastructure (Fig. 2 G and H). These results indicate that expressing hTDP-43 in Drosophila eyes leads to age-dependent progressive degeneration of photoreceptors. The expression of the carboxyl-terminal fragment of hTDP-43 alone is not sufficient to cause neurodegeneration in flies.

An external file that holds a picture, illustration, etc.
Object name is pnas.0913602107fig02.jpg

Transmission EM reveals ultrastructural defects in the hTDP-43-expressing fly eyes. (A and B) Control flies (GMR-Gal4/UAS-RFP) showed normal rhabdomere (Rh) arrangement and ultrastructure at day 1 (A), and the intact rhabdomere structure is well maintained by day 30 (B). (CF) At day 1, GMR-Gal4/UAS-hTDP-43-RFP flies showed severe losses of ommatidial structure with disrupted rhabdomeres (C) and the presence of multilamellar body (MLB). In 30-day-old flies expressing hTDP-43, the ommatidial arrangement under the cone (marked by “c”) structure was completely disrupted (D), with the formation of autophagic vacuoles (AV) and multivesicular bodies (MVB) (F). The asterisk marks a degenerating mitochondrion. (G and H) The ommatidia were normal in transgenic flies expressing T202 mutant hTDP-43. (Scale bars: AD and G, 10 μm; E, F, and H, 2 μm.)

Expression of hTDP-43 in Mushroom Bodies (MBs) Causes Axonal Loss and Neuronal Death.

To test whether hTDP-43 expression had similar pathogenic effects in the central nervous system, we prepared flies expressing hTDP-43 in MBs under the control of OK107-Gal4. Axon projections from three subtypes of MB cells form the α/β, α′/β′, and γ lobes, respectively. These axon bundles were visualized by signals of the coexpressed membrane-localized GFP (mGFP) from UAS-mCD8::GFP construct. Three-dimensional structure of MB lobes was imaged by using confocal microscopy (Fig. 3). MBs in control flies showed typical lobe structure (Fig. 3A), whereas axon bundles were much thinner or completely lost as early as day one in hTDP-43-expressing MBs (Fig. 3C). There were remarkable variations in axonal defects among different flies with the same genetic background, involving different MB lobes. Interestingly, the remaining axons in each individual brain were often derived from one subtype of MB neurons (for example, α/β in C1, α′/β′ in C2, and γ in C3 in Fig. 3). This suggests that the axon loss induced by hTDP-43 expression was not specific to one specific neuronal subpopulation, although it often occurred in a group manner. Furthermore, in hTDP-43-expressing MBs, these CNS neurons showed an age-dependent neurodegeneration, similar to that in hTDP-43-expressing photoreceptors cells. As compared with the normal lobe structures observed in the control MBs expressing RFP at day 30 (Fig. 3B), MBs expressing hTDP-43 showed only residual axon bundles with mGFP signals barely detectable (Fig. 3D). In addition, these hTDP-43-expressing MB neurons in aged flies showed obvious signs of axon degeneration, including a loss of axon integrity and increased abnormal axonal varicosities with uneven distribution of mGFP signals (marked by thin arrows in Fig. 3 D1 and D3). Such axonal pathology is not detectable in transgenic flies expressing T202 mutant TDP-43, suggesting the requirement of the amino-terminal domain for TDP-43 neurotoxicity in vivo.

An external file that holds a picture, illustration, etc.
Object name is pnas.0913602107fig03.jpg

Expression of hTDP-43 in mushroom bodies causes axonal loss. Axon bundles of MB neurons are visualized by membrane GFP (CD8::mGFP). Confocal images include all lobes in each brain obtained in Z-stacks and projected into single images. (A and B) Control flies (OK107-Gal4/UAS-mGFP/UAS-RFP) showed the normal MBs lobe structure at both day 1 and day 30. Arrow, α/α′ lobes; arrowhead, β/β′ lobes; star, γ lobe. (C) OK107-Gal4/UAS-mGFP/UAS-hTDP-43-RFP flies began to lose MB lobes at day 1. C1, only parts of α and β lobes were left; C2, only α′ and β′ lobes were left; C3, only γ and a few α′/ β′ lobes were left. (D) By day 30, flies expressing hTDP-43 had lost most axons, with few remaining in α and β lobes (D1 and D2) or in γ lobe (D3). The thin arrows in D1 and D3 mark the abnormal axonal varicosities with uneven distribution of mGFP signals.

We further tested whether reducing the expression of the endogenous Drosophila TBPH could affect the CNS neurons in MBs. Using the same TBPH RNAi lines, we showed that decreased TBPH expression in the MBs also resulted in axonal loss and neuronal death (Fig. S2), indicating that TBPH gene is essential for normal development of Drosophila CNS neurons. However, axonal loss in TBPH RNAi flies was less severe and did not occur in the “group” manner as that in hTDP-43 overexpressing MBs (Fig. 3 C and D). No obvious abnormal axonal varicosities were detected in TBPH RNAi flies (Fig. S2 C and D).

We also examined morphological changes in cell bodies using confocal microscopy (Fig. 4). In normal flies, mushroom bodies have ∼2,500 neurons in each calyx, organized in clusters. Such calyx structures were well preserved in control flies expressing RFP protein (Fig. 4A). In contrast, hTDP-43-expressing brains showed dramatic neuronal loss with significant variations in severity, similar to that in the fly eyes. For example, among 20 samples examined, 8 had severe neuronal loss with only a few neurons left, 10 lost more than half of their MB neurons (Fig. 4B), and 2 showed little neuronal defects. Interestingly, in some neurons, the mGFP signal was completely lost with only hTDP-43-RFP signal detectable (arrows in Fig. 4B1B3). Consistent with our observation in MB axons, MB neuronal death also occurred in groups, rather than in a random fashion, with regions of cell body loss and only residual axonal mGFP signals detectable (arrowhead in Fig. 4B).

An external file that holds a picture, illustration, etc.
Object name is pnas.0913602107fig04.jpg

Expression of hTDP-43 induces neuronal death in mushroom bodies. (A and B) Confocal sections of MB calyces in control flies (A, OK107-Gal4/UAS-mGFP/UAS-RFP) or hTDP-43-expressing flies (B, OK107-Gal4/UAS-mGFP/UAS-hTDP-43-RFP). MBs were well organized in the control, whereas hTDP-43-expressing brains showed much fewer neurons with underneath axons of residual neurons (arrowhead) visible in the area lacking cells. Note some neurons (arrows in B1+2) with hTDP-43-RFP signal did not have mGFP signal, whereas all control neurons showed normal morphology with both RFP and mGFP detected (A1+2). Nuclei were revealed by Hoechst dye staining in blue. (Scale bars: 50 μm.) (C) TUNEL staining revealed cell death in hTDP-43-expressing cells. C1, hTDP-43-RFP; C2, TUNEL staining; C3, Hoechst staining; C1+2 and C, superimposed images. (Scale bars: 20 μm.) The dying neurons show significant variations. Cell 1 showed relatively normal nuclear hTDP-43 distribution, whereas cell 3 had cytoplasmic hTDP-43-RFP signals with abnormal nuclear morphology as well as reduced nuclear TDP-43 signals. Cell 4 showed the condensed nucleus.

We performed terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay to examine the cell death. No TUNEL-positive cells were detected in control MBs expressing RFP (Fig. S3A) in 10-day-old flies. However, three to six TUNEL-positive cells were frequently found in each calyx of hTDP-43-expressing MBs from flies at the same age as control flies (Fig. 4C2), indicating that overexpression of hTDP-43 in these CNS neurons induced neuronal death. These dying neurons showed significant variations in the nuclear morphology as well as TDP-43-RFP fluorescence level and expression patterns. This may reflect dynamic changes of hTDP-43 expression and nuclear morphology during different stages of cell death.

Expression of hTDP-43 in Motor Neurons (MNs) Causes Aggregate Formation in Cell Bodies and Axons, Together with Axon Swelling.

To investigate potential roles of TDP-43 in MNs in vivo, we examined flies expressing hTDP-43 in MNs using a MN-specific driver, OK371-Gal4. Such transgenic flies expressing hTDP-43 showed apparently normal morphology at the larval stage but failed to hatch in the late pupa stage. We first examined the expression and localization of hTDP-43 in the transgenic flies using fluorescent microscopy. In most MNs, hTDP-43 was predominantly localized in the nucleus. When hTDP-43 was detected in the cytoplasm, neurons often showed signs of cell death: condensation and/or fragmentation of nuclei (marked by the arrowhead in Fig. 5A). TDP-43 protein aggregation was observed in both cell bodies (arrow in Fig. 5A) and axons (arrows in Fig. 5 B and C) in these MNs. Strikingly, axon swelling was frequently detected in MNs expressing hTDP-43 (Fig. 5C), similar in morphology to spheroid structures found in ALS patients (21, 22).

An external file that holds a picture, illustration, etc.
Object name is pnas.0913602107fig05.jpg

Overexpression of hTDP-43 in motor neurons leads to protein aggregation in cytoplasm and axons, together with axonal swelling. (A and B) Overexpressing hTDP-43 using the motor neuron-specific driver (OK371-Gal4) resulted in protein aggregates detectable in the cytoplasm (A) and axons (B). Arrowheads in A1A3, A1+2, A1+3, and A1+2+3 mark cells containing condensed and fragmented nucleus. Arrows in B1 and B1+2 mark the hTDP-43-containing abnormal axonal varicosities. (C) All hTDP-43-overexpressing flies exhibited axon swelling with TDP-43-positive inclusions, which was absent in the control flies. Some regions with TDP-43 protein aggregates show reduced mGFP signals (marked by arrows), and other regions exhibit significant axonal swelling as marked by arrowheads. In all images: red, hTDP-43-RFP; green, mGFP; blue, Hoechst staining. (Scale bars: 50 μm.)

Expression of hTDP-43 in MNs Causes Neuronal Death and Functional Deficits.

We systematically studied our transgenic flies, focusing on MNs of late third instar larvae. Similar to the wild-type flies, MNs in the control RFP-expressing flies were well organized into clusters in different segments, as visualized by RFP and mGFP signals (Fig. 6 A1 and A2). In contrast, MNs in hTDP-43-expressing flies were not organized in their normal clusters, showing cell loss. Such MN loss was more severe in the posterior abdominal segments as compared with the anterior ones. For example, in Fig. 6B1, two neurons expressing hTDP-43-RFP marked by arrowheads had lost mGFP signals, and another neuron (marked by the arrow) showed cell body swelling, suggesting of neuronal death. Such neuronal death was confirmed by TUNEL staining. There were no TUNEL-positive neurons in the controls, whereas TUNEL-positive cells were frequently detected among MNs expressing hTDP-43 (Fig. S4).

An external file that holds a picture, illustration, etc.
Object name is pnas.0913602107fig06.jpg

Expression of hTDP-43 causes morphological and functional defects of motor neurons (MNs). (A1A3) MNs in the dorsal clusters were well organized in the ventral nerve cord (VNC) in the control flies. Arrows and arrowheads mark three representative MNs showing healthy morphology. (Scale bars: 20 μm.) (B1B3) MNs expressing hTDP-43 showed cell death and morphological defects in dorsal clusters, especially in the last three segments of VNC. The arrow marks a swollen neuron with an enlarged mGFP area, and arrowheads mark MNs with fragmented or condensed nuclei and reduced mGFP signals. (C and D) Examples of axon branches and boutons at NMJ in the abdominal segment A2 of control or hTDP-43 transgenic flies, as revealed by mGFP signals. (Scale bars: 100 μm.) (E) In the abdominal segments 2–4, hTDP-43 expression led to a significant reduction in the number of axon branches and boutons, as compared with the control flies. ***, P < 0.0001; **, P < 0.001. (F) The larval locomotion was significantly impaired in hTDP-43-expressing flies but not in TBPH-RNAi flies (labeled as “TDP,” “RNAi” groups, in black and gray bars, respectively), as compared to the control (Ctrl) flies. When hTDP-43-expressing flies were crossed to TBPH-RNAi flies, their larval movement was improved as shown in the cross-hatched bar. More than 30 larvae were tested in each group. ***, P < 0.0001; **, P < 0.001. (G) The climbing ability of adult flies showed progressive decrease during aging in hTDP-43-expressing flies, as compared with the controls, with the difference between the two groups significant beginning from day 15. The genotype of flies used in controls: AE, OK371-Gal4/UAS-mGFP/UAS-RFP or UAS-TDP-43-RFP; F, OK371-Gal4/UAS-hTDP-43-RFP or UAS-RFP, or these flies crossed with TBPH-RNAi; G, RN2-Gal4/UAS-hTDP-43-RFP or UAS-RFP.

Axonal and synaptic morphology of hTDP-43-expressing motor neurons were examined, especially at neuromuscular junctions (NMJ). Nerves from motor neurons project to their target muscles, branch a few times, and form synaptic boutons to innervate the muscle. MNs overexpressing hTDP-43 showed a significant decrease in the numbers of both big and small boutons, and of the axon branches (Fig. 6 D and E) as compared to the control flies (Fig. 6C). Consistently, the locomotive ability of hTDP-43 flies was also significantly impaired at both larval and adult stages. The larval motility was examined from late third instar larvae, and the motility index was quantified as the number of movement waves during a period of 2 min. As shown in Fig. 6F, the motility index of control flies was 86.1 ± 3.1. This index was reduced to 46.0 ± 2.3 in hTDP-43-expressing flies. Interestingly, the locomotive ability of TBPH-RNAi fly larvae was normal, indicating that reducing the endogenous TBPH expression did not affect larval movement. When hTDP-43 transgenic flies were crossed onto the TBPH-RNAi background, however, their larval movement was improved. The change in motility index was statistically significant (Fig. 6F). This suggests that the phenotype caused by hTDP-43 expression is not due to a loss of function of TDP-43. Otherwise, we would expect exacerbation of the phenotype by the further reduction of the endogenous fly TDP-43 gene expression.

Most flies expressing hTDP-43 under the OK371-Gal4 driver died at or before eclosion. To examine adult fly motility, we used RN2-Gal4 to drive hTDP-43 expression in a small subset of MNs at the adult stage (23). Adult fly locomotive ability was examined at different ages by using a climbing assay as described in ref. 24. As compared to the control flies, hTDP-43-expressing flies showed a reduction in their motility, becoming progressively worse with aging and with the difference significant from day 10 after eclosion (Fig. 6G).

We further investigated whether hTDP-43 expression in fly neurons led to the formation of insoluble TDP-43 species by fractionating protein lysates prepared from heads of adult flies with the GMR driver. Sarkosyl-insoluble high-molecular-weight TDP-43 immunoreactive protein aggregates were detected in these flies (Fig. S5, lane 6), again mimicking the pathological TDP-43 species detected in patient tissues (6).


The concept of TDP-43 proteinopathy has emerged as TDP-43-positive lesions have been discovered in a range of neurodegenerative diseases including ALS. TDP-43 proteinopathy is characterized by TDP-43 immunoreactive neuronal and glial inclusion bodies together with a progressive loss of function of the affected neurons. A recent study using an adeno-associated virus vector to express hTDP-43 in substantia nigra (SN) in rats revealed a loss of SN neurons, gliosis, and changes in amphetamine-stimulated rotational behavior (25). Here, we report the generation and characterization of transgenic flies expressing human TDP-43. Our hTDP-43 transgenic flies have amorphologic features strikingly similar to neuropathological characteristics of TDP-43 proteinopathy, and functional defects in these flies mimic those in TDP-43 proteinopathy patients. Several findings make our study particularly intriguing (SI Discussion). First, this animal model displays a wide range of neuropathological, biochemical, and functional features of human TDP-43 proteinopathy. Other currently available animal models recapitulate a limited repertoire of features of this syndrome. Second, our data suggest that overexpression of hTDP-43 or accumulation of hTDP-43 gene products may lead to morphological changes and functional deficits of TDP-43 proteinopathy. Third, similar to ALS and FTLD cases with TDP-43 pathology, our transgenic flies show significant variations in both neuropathology and functional changes. Fourth, our model may be useful for studying both sporadic and familial forms of TDP-43 proteinopathy, especially those without detectable mutations in the TDP-43 gene. Fifth, the neuropathological and functional features of our transgenic flies make them a powerful model for identifying both genetic modifier genes and chemical compounds that can either improve or exacerbate phenotypes in TDP-43 proteinopathy patients.

Progressive and age-dependent neuronal loss is a major pathologic hallmark of TDP-43 proteinopathy, including FTLD and ALS. In our study, when wild-type hTDP-43 protein was expressed in specific neuronal subpopulations in vivo, progressive and age-dependent neuronal loss was found in photoreceptors (Figs. 1 and and2),2), MBs (Figs. 3 and and44 and Fig. S2), and motor neurons (Figs. 5 and and66 and Fig. S4), suggesting that simply increasing TDP-43 expression is sufficient to initiate progressive neuronal death and functional deficiencies. ALS neuropathology is characterized by progressive degeneration of motor neurons, with synaptic loss and the presence of chromatolytic neurons (“balloon-like cells”) together with axonal spheroids (swellings) (26, 27). The synaptic reduction in anterior horns of the spinal cord in ALS patients is correlated with the degree of neuronal loss (27). In some cases, axonopathy with giant axon swellings in the corticospinal tracts was detected, suggesting possible distal axonal degeneration during the early stages of this disease (21, 22, 26). In our transgenic flies, motor neurons expressing hTDP-43 show both synapse reduction and abnormal axon swelling, mimicking the synaptic loss and axonal spheroids/axon swellings found in ALS patients. In addition, these transgenic flies exhibited age-dependent deficits in locomotion, similar to the clinical disability in ALS patients. Interestingly, a very recent study published while this article was under review reported the comparison of nontransgenic mice and transgenic mice expressing A315T mutant form of hTDP-43, showing a reduction in motor neuron axon numbers in A315T-TDP-43 transgenic mice (28). It remains unclear whether overexpressing wild-type hTDP-43 in mice could lead to similar pathology. However, deleting fly TBPH gene or reducing its expression by RNAi has been reported to decrease dendritic branching in sensory neurons, uncovering an important role of TBPH in dendritic development (29).

It is unclear whether dysfunction of TDP-43 is the major cause of TDP-43 proteinopathy and whether gain or loss of function of TDP-43 protein is the underlying pathogenic mechanism. We examined the effects of down-regulating fly TDP-43 homolog TBPH on the formation and function of the nervous system. As a result of alternative splicing, Drosophila TBPH gene is predicted to produce four RNA transcripts encoding two protein isoforms (Fig. S1). With TBPH-specific primers, the long isoform (TBPHl) was detected as the predominant isoform in the fly heads, and it was significantly reduced in TBPH-RNAi-expressing flies as compared to that in the control flies (Fig. S1B). These RNAi flies show axonal and synapse losses as well as impaired locomotion in the adult stage, suggesting that TBPH is required for function and maintenance of the nervous system in Drosophila, consistent with a recent study (29). It will be interesting to test whether overexpressing either long or short isoform of fly TBPH gene may also cause neurodegeneration phenotypes. Our data show that motor neuron phenotype caused by hTDP-43 expression was partially rescued by knocking down the expression of the endogenous fly TDP-43 homolog (Fig. 6F), consistent with the possibility of gain-of-function toxicity of hTDP-43 overexpression. If loss-of-function of TDP-43 is a major cause of MN phenotype caused by hTDP-43 expression in flies, it would be anticipated that reducing the endogenous fly TDP-43 expression by RNAi should further exacerbate, not partially rescue, MN loss. Therefore, our results support the gain-of-function model for TDP-43 proteinopathy. It is worth noting that the carboxyl-terminal domains of Drosophila TBPH (in both splicing isoforms) show little similarity to that of hTDP-43, a region where most human disease mutations have been located. Accumulation of carboxyl-terminal fragments of TDP-43 in affected tissue has been reported (6). One proteolytic cleavage site has been mapped to arginine residue 208, and expression of the carboxyl-terminal fragment in transfected cells leads to the formation of cytoplasmic aggregates in vitro (16). In our parallel study, overexpression of a similar carboxyl-terminal fragment, T202, did not affect either neuronal morphology or neuronal function (Figs. 1 and and2).2). This indicates that neurotoxicity caused by hTDP-43 expression requires its amino terminus, including the first RNA recognition motif.

Taken together, our hTDP-43 transgenic fly model recapitulates critical features of human TDP-43 proteinopathy and should serve as a powerful animal model for TDP-43 proteinopathy. In addition, our hTDP-43 transgenic fly data indicate that simply increasing hTDP-43 expression may be sufficient to cause neurotoxicity in vivo. This suggests that aberrant regulation of TDP-43 gene expression or decreased clearance of hTDP-43 in the adult stage may contribute to the pathogenesis of TDP-43 proteinopathy. Consistent with this is the observation that TDP-43 pathology is indistinguishable between sporadic and familial ALS cases and that TDP-43 proteinopathy can occur in the absence of TDP-43 mutation (ref. 30, reviewed in refs. 31 and 32). Our hTDP-43 transgenic flies can be used for studying both sporadic and familial forms of TDP-43 proteinopathy without detectable mutations in TDP-43 gene. Furthermore, our fly model will be useful for identifying genetic interacting factors that may modify the disease phenotypes and screening for chemical compounds that could delay or reduce neuromuscular defects caused by aberrant TDP-43 protein expression, thus facilitating future development of effective therapy for TDP-43 proteinopathy.

Materials and Methods

Constructs, Fly Stains, Histology, Immunostaining, Fluorescent Microscopy, and Transmission Electron Microscopy.

For a detailed description, see SI Text.

Fly Motility Assays.

The larval motility was measured as the number of peristaltic waves in 2 min in the late third instar larvae expressing control or hTDP-43 under the driver of OK371-Gal4, as described in ref. 20. The adult motility was determined by using the climbing assay (24). Briefly, 35 virgin female flies were collected in each group. Fly climbing ability was examined every 5 days as the number of flies climbing above a 5-cm line in 18 seconds after they were tapped to the bottom of an empty vial. The experiment was repeated 10 times for each group.

Department of Neurology, Lurie Cancer Center, Center for Genetic Medicine, Northwestern University School of Medicine, Chicago, IL 60611;
State Key Laboratory of Brain and Cognitive Science &amp; System Biology Center, Institute of Biophysics, Chinese Academy of Sciences, Beijing 100101, China;
Lincoln Park High School, Chicago, IL 60614;
Biological Sciences and Chemistry, University of Illinois, at Chicago, Chicago, IL 60607; and
Magnify Inc., Nashville, TN 37216-3330
To whom correspondence may be addressed. E-mail: ude.nretsewhtron@nay-l or ude.nretsewhtron@uw-enaj.
Communicated by Xiaodong Wang, University of Texas Southwestern Medical Center, Dallas, TX, December 12, 2009 (received for review July 27, 2009)

Author contributions: E.J.R. and J.Y.W. designed research; Y.L., P.R., E.J.R., C.S., W.G., X.C., K.F., and J.Y.W. performed research; E.A.W. and J.Y.W. contributed new reagents/analytic tools; Y.L., P.R., E.J.R., C.S., W.G., X.C., E.A.W., K.F., and J.Y.W. analyzed data; and Y.L., P.R., E.J.R., and J.Y.W. wrote the paper.

Present address: Trumbull College, Yale University, New Haven, CT 06511.
Communicated by Xiaodong Wang, University of Texas Southwestern Medical Center, Dallas, TX, December 12, 2009 (received for review July 27, 2009)


Neuropathology involving TAR DNA binding protein-43 (TDP-43) has been identified in a wide spectrum of neurodegenerative diseases collectively named as TDP-43 proteinopathy, including amyotrophic lateral sclerosis (ALS) and frontotemporal lobar dementia (FTLD). To test whether increased expression of wide-type human TDP-43 (hTDP-43) may cause neurotoxicity in vivo, we generated transgenic flies expressing hTDP-43 in various neuronal subpopulations. Expression in the fly eyes of the full-length hTDP-43, but not a mutant lacking its amino-terminal domain, led to progressive loss of ommatidia with remarkable signs of neurodegeneration. Expressing hTDP-43 in mushroom bodies (MBs) resulted in dramatic axon losses and neuronal death. Furthermore, hTDP-43 expression in motor neurons led to axon swelling, reduction in axon branches and bouton numbers, and motor neuron loss together with functional deficits. Thus, our transgenic flies expressing hTDP-43 recapitulate important neuropathological and clinical features of human TDP-43 proteinopathy, providing a powerful animal model for this group of devastating diseases. Our study indicates that simply increasing hTDP-43 expression is sufficient to cause neurotoxicity in vivo, suggesting that aberrant regulation of TDP-43 expression or decreased clearance of hTDP-43 may contribute to the pathogenesis of TDP-43 proteinopathy.

Keywords: amyotrophic lateral sclerosis, animal model, RNA binding protein

TDP-43 (TAR DNA binding protein of 43 kDa) is encoded by TARDBP, a highly conserved gene on chromosome 1 of the human genome (13). Initially identified as a transcriptional repressor of HIV-1 gene expression (4), TDP-43 is a multifunction protein involved in transcription, splicing, and mRNA stabilization (reviewed in ref. 5). Recent studies show that TDP-43 is a major protein component of neuronal inclusion bodies in the affected tissues in a range of neurodegenerative disorders, including amyotrophic lateral sclerosis (ALS), frontotemporal lobar dementia (FTLD) (6, 7), Alzheimer’s disease (AD) (810), and other types of dementia (1013). Decreased protein solubility, hyperphosphorylation, abnormal cleavage, and cytoplasmic mis-localization of TDP-43 have been associated with TDP-43 proteinopathy (1416). It is not clear whether TDP-43 proteinopathy is caused by loss-of-function of TDP-43 or gain-of-function neurotoxicity. Here, we report the generation and characterization of transgenic flies expressing human TDP-43. In different types of neurons, including photoreceptors, mushroom bodies, or motor neurons, simply overexpressing hTDP-43 by itself is sufficient to cause protein aggregate formation and neuronal loss in an age-dependent manner, suggesting that increased hTDP-43 expression or aberrant accumulation of hTDP-43 may lead to TDP-43 proteinopathy. Our transgenic flies recapitulate important pathological and clinical features of ALS, representing a powerful animal model for TDP-43 proteinopathy.

Click here to view.


We thank Drs. E. Bigio, A. Kolodkin, L. M. Li, L. Q. Luo, M. Mesulam, and T. Siddique for generously providing invaluable suggestions, reagents, and support and critical reading of the manuscript; J. Wu for kind help; A. Sibtain for technical assistance; and the members of the Wu laboratory for helpful discussion and critical reading of the manuscript. W.G. is supported by MOST China (2009CB825402). This work was supported by National Institutes of Health Grants GM070967; and EY014576 (to J.Y.W.) and by the Searle Foundation (to J.Y.W.).



The authors declare no conflict of interest.

This article contains supporting information online at www.pnas.org/cgi/content/full/0913602107/DCSupplemental.



  • 1. Stover C, et al The human gene for mannan-binding lectin-associated serine protease-2 (MASP-2), the effector component of the lectin route of complement activation, is part of a tightly linked gene cluster on chromosome 1p36.2-3. Genes Immun. 2001;2:119–127.[PubMed][Google Scholar]
  • 2. Wang HY, Wang IF, Bose J, Shen CKStructural diversity and functional implications of the eukaryotic TDP gene family. Genomics. 2004;83:130–139.[PubMed][Google Scholar]
  • 3. Ayala YM, et al Human, Drosophila, and C.elegans TDP43: nucleic acid binding properties and splicing regulatory function. J Mol Biol. 2005;348:575–588.[PubMed][Google Scholar]
  • 4. Ou SH, Wu F, Harrich D, García-Martínez LF, Gaynor RBCloning and characterization of a novel cellular protein, TDP-43, that binds to human immunodeficiency virus type 1 TAR DNA sequence motifs. J Virol. 1995;69:3584–3596.[Google Scholar]
  • 5. Buratti E, Baralle FEMultiple roles of TDP-43 in gene expression, splicing regulation, and human disease. Front Biosci. 2008;13:867–878.[PubMed][Google Scholar]
  • 6. Neumann M, et al Ubiquitinated TDP-43 in frontotemporal lobar degeneration and amyotrophic lateral sclerosis. Science. 2006;314:130–133.[PubMed][Google Scholar]
  • 7. Arai T, et al TDP-43 is a component of ubiquitin-positive tau-negative inclusions in frontotemporal lobar degeneration and amyotrophic lateral sclerosis. Biochem Biophys Res Commun. 2006;351:602–611.[PubMed][Google Scholar]
  • 8. Amador-Ortiz C, et al TDP-43 immunoreactivity in hippocampal sclerosis and Alzheimer’s disease. Ann Neurol. 2007;61:435–445.[Google Scholar]
  • 9. Uryu K, et al Concomitant TAR-DNA-binding protein 43 pathology is present in Alzheimer disease and corticobasal degeneration but not in other tauopathies. J Neuropathol Exp Neurol. 2008;67:555–564.[Google Scholar]
  • 10. Higashi S, et al Concurrence of TDP-43, tau and alpha-synuclein pathology in brains of Alzheimer’s disease and dementia with Lewy bodies. Brain Res. 2007;1184:284–294.[PubMed][Google Scholar]
  • 11. Nakashima-Yasuda H, et al Co-morbidity of TDP-43 proteinopathy in Lewy body related diseases. Acta Neuropathol. 2007;114:221–229.[PubMed][Google Scholar]
  • 12. Geser F, et al Pathological TDP-43 in parkinsonism-dementia complex and amyotrophic lateral sclerosis of Guam. Acta Neuropathol. 2008;115:133–145.[PubMed][Google Scholar]
  • 13. Neumann M, et al TDP-43 in the ubiquitin pathology of frontotemporal dementia with VCP gene mutations. J Neuropathol Exp Neurol. 2007;66:152–157.[PubMed][Google Scholar]
  • 14. Zhang YJ, et al Progranulin mediates caspase-dependent cleavage of TAR DNA binding protein-43. J Neurosci. 2007;27:10530–10534.[Google Scholar]
  • 15. Sreedharan J, et al TDP-43 mutations in familial and sporadic amyotrophic lateral sclerosis. Science. 2008;319:1668–1672.[PubMed][Google Scholar]
  • 16. Igaz LM, et al Expression of TDP-43 C-terminal Fragments in Vitro Recapitulates Pathological Features of TDP-43 Proteinopathies. J Biol Chem. 2009;284:8516–8524.[Google Scholar]
  • 17. Feany MB, Bender WWA Drosophila model of Parkinson’s disease. Nature. 2000;404:394–398.[PubMed][Google Scholar]
  • 18. Bilen J, Bonini NMDrosophila as a model for human neurodegenerative disease. Annu Rev Genet. 2005;39:153–171.[PubMed][Google Scholar]
  • 19. Brand AH, Perrimon NTargeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development. 1993;118:401–415.[PubMed][Google Scholar]
  • 20. Feiguin F, et al Depletion of TDP-43 affects Drosophila motoneurons terminal synapsis and locomotive behavior. FEBS Lett. 2009;583:1586–1592.[PubMed][Google Scholar]
  • 21. Delisle MB, Carpenter SNeurofibrillary axonal swellings and amyotrophic lateral sclerosis. J Neurol Sci. 1984;63:241–250.[PubMed][Google Scholar]
  • 22. Gold BGThe pathophysiology of proximal neurofilamentous giant axonal swellings: implications for the pathogenesis of amyotrophic lateral sclerosis. Toxicology. 1987;46:125–139.[PubMed][Google Scholar]
  • 23. Fujioka M, et al Even-skipped, acting as a repressor, regulates axonal projections in Drosophila. Development. 2003;130:5385–5400.[Google Scholar]
  • 24. Romero E, et al Suppression of neurodegeneration and increased neurotransmission caused by expanded full-length huntingtin accumulating in the cytoplasm. Neuron. 2008;57:27–40.[Google Scholar]
  • 25. Tatom JB, et al Mimicking aspects of frontotemporal lobar degeneration and Lou Gehrig’s disease in rats via TDP-43 overexpression. Mol Ther. 2009;17:607–613.[Google Scholar]
  • 26. Okamoto K, Hirai S, Shoji M, Senoh Y, Yamazaki TAxonal swellings in the corticospinal tracts in amyotrophic lateral sclerosis. Acta Neuropathol. 1990;80:222–226.[PubMed][Google Scholar]
  • 27. Sasaki S, Maruyama SDecreased synaptophysin immunoreactivity of the anterior horns in motor neuron disease. Acta Neuropathol. 1994;87:125–128.[PubMed][Google Scholar]
  • 28. Wegorzewska I, Bell S, Cairns NJ, Miller TM, Baloh RHTDP-43 mutant transgenic mice develop features of ALS and frontotemporal lobar degeneration. Proc Natl Acad Sci USA. 2009;106:18809–18814.[Google Scholar]
  • 29. Lu Y, Ferris J, Gao FBFrontotemporal dementia and amyotrophic lateral sclerosis-associated disease protein TDP-43 promotes dendritic branching. Mol Brain. 2009;2:30.[Google Scholar]
  • 30. Geser F, et al Clinical and pathological continuum of multisystem TDP-43 proteinopathies. Arch Neurol. 2009;66:180–189.[Google Scholar]
  • 31. Van Deerlin VM, et al TARDBP mutations in amyotrophic lateral sclerosis with TDP-43 neuropathology: a genetic and histopathological analysis. Lancet Neurol. 2008;7:409–416.[Google Scholar]
  • 32. Pesiridis GS, Lee VM, Trojanowski JQMutations in TDP-43 link glycine-rich domain functions to amyotrophic lateral sclerosis. Hum Mol Genet. 2009;18(R2, R2):R156–R162.[Google Scholar]
Collaboration tool especially designed for Life Science professionals.Drag-and-drop any entity to your messages.