Introduction
The abnormal aggregation of proteins into inclusions in the central nervous system is a common pathological hallmark of neurodegenerative diseases, including Alzheimer’s disease, Parkinson’s disease, amyotrophic lateral sclerosis (ALS), and frontotemporal dementia (FTD). TAR DNA-binding protein 43 (TDP-43), a DNA/RNA-binding protein, is a major component of the cytoplasmic inclusions formed in the affected neurons of patients with ALS and FTD [
1,
31]. Almost all ALS patients and half of all FTD patients are thought to have TDP-43 pathology; ALS and FTD are currently recognized as a spectrum disorder with common genetic, clinical, and pathological aspects [
22]. TDP-43 is mainly localized in the nucleus, and TDP-43 in cytoplasmic inclusions is known to undergo various post-translational modifications, such as ubiquitination, phosphorylation, and truncation [
41]. Cytoplasmic inclusions of TDP-43 are accompanied with the depletion of TDP-43 in the nucleus of affected neurons of ALS/FTD patients, and therefore, two major pathogenic mechanisms have been suggested; a gain of toxic function triggered by TDP-43 aggregation, and loss of the physiological functions of TDP-43 through its mislocalization [
19,
20]. Because TDP-43 plays essential roles in cellular activities, including nucleocytoplasmic transport, RNA processing, and stress granule metabolism, the aggregation and mislocalization of TDP-43 are likely to cause abnormalities in various cellular functions, leading to detrimental consequences for cell survival [
37].
Despite the central roles of TDP-43 in the pathogenesis of ALS/FTD, the mechanism of TDP-43 aggregation remains unclear. Upon stress, TDP-43 is known to physiologically form cytoplasmic condensates called stress granules [
41]. Stress granules are cytoplasmic membraneless organelles that are transiently formed under stress conditions through the liquid–liquid phase separation of RNA-binding proteins, RNA, and other components, which are readily disassembled during recovery after stress. The carboxy-terminal domain of TDP-43 is intrinsically disordered with low complexity, and plays an important role in the formation of stress granules [
4,
5]. Intriguingly, missense mutations in TDP-43 that are associated with familial cases of ALS/FTD are mostly located in the carboxy-terminal low-complexity domain, and are found to potentially increase the aggregation propensity of TDP-43. Therefore, these familial mutations may disrupt the physiological liquid–liquid phase separation and/or stress granule dynamics, triggering the pathological aggregation of TDP-43 in the cytoplasm [
21]. However, in most sporadic ALS/FTD patients, wild-type TDP-43 without mutations is aggregated into inclusions, and thus, it is unclear as to what drives the aggregation of wild-type TDP-43 in sporadic ALS/FTD.
The abnormal aggregation of TDP-43 is also observed in other neurodegenerative diseases, including Perry syndrome [
27,
44] and inclusion body myopathy associated with Paget disease of bone and frontotemporal dementia (IBMPFD) [
30,
33], which are referred to as secondary TDP-43 proteinopathies. Perry syndrome is an autosomal dominant neurodegenerative disease, and TDP-43-positive neuronal inclusions have been observed in the substantia nigra, globus pallidus, and brainstem of patient brains [
39]. Several missense mutations in DCTN1, a subunit of the microtubule-associated motor protein complex dynactin, have been identified as the genetic cause of Perry syndrome [
9]. Because these mutations are mostly located in the microtubule-binding domain at the amino terminus, the functional loss of DCTN1 may play a key role in the formation of cytoplasmic TDP-43 inclusions [
43]. Intriguingly, DCTN1 expression is markedly downregulated in motor neurons of sporadic ALS patients [
18], and the missense mutation in DCTN1 has also been linked to the autosomal dominant distal hereditary motor neuronopathy type VIIB (HMN7B) [
32], indicating the crucial role of DCTN1 not only in Perry syndrome, but also in motor neuron diseases, including ALS. Because intracellular transport along microtubules has been implicated in the formation and disassembly of stress granules [
24], we hypothesized that DCTN1 dysfunction and/or impaired microtubule transport may cause abnormalities in stress granule dynamics, which potentially leads to the pathological aggregation of TDP-43 in the cytoplasm.
To test this hypothesis, in this study we investigated whether deficiency in DCTN1 and intracellular transport along microtubules drives the formation of cytoplasmic TDP-43 aggregates in vivo. We demonstrated using a Drosophila model of ALS/FTD that genetic knockdown of DCTN1, as well as other components of microtubule-associated motor protein complexes, accelerates the formation of ubiquitin-positive cytoplasmic inclusions of TDP-43, leading to the exacerbation of neurodegeneration. Notably, DCTN1 knockdown delayed the disassembly of stress granules in stressed cells, leading to an increase in the formation of pathological cytoplasmic inclusions of TDP-43. These results demonstrate that DCTN1 and other microtubule-associated motor proteins are modifiers that drive the aggregation of wild-type TDP-43 through the dysregulation of stress granule dynamics, indicating the crucial role of intracellular transport along microtubules in the pathological development of TDP-43 proteinopathies, including ALS/FTD.
Discussion
Although abnormal aggregation of TDP-43 into inclusions is a common pathological hallmark of ALS/FTD, what drives the aggregation of TDP-43 remains unclear. In the present study, we addressed this issue by investigating the role of DCTN1, a causative gene of Perry syndrome in which patients develop TDP-43 pathology, in the formation of TDP-43 aggregation using a Drosophila model of ALS/FTD. We demonstrated in vivo that the knockdown of DCTN1, as well as other components of microtubule-associated motor protein complexes, including dynactin, dynein, and kinesin, exacerbates pathological TDP-43 aggregation in the cytoplasm. Cell culture experiments demonstrated that DCTN1 knockdown delays the disassembly of stress granules and promotes the formation of ubiquitin-positive TDP-43 inclusions in stressed cells. These results demonstrate that a deficiency in DCTN1 and other microtubule-associated motor proteins drives the aggregation of wild-type TDP-43 through the dysregulation of stress granule dynamics, indicating the crucial role of intracellular transport along microtubules in the pathological development of the TDP-43 proteinopathies, including ALS/FTD.
Our work provides important insights into the mechanisms of TDP-43 aggregation, and a potential therapeutic target for the TDP-43 proteinopathies. First, dysfunction of microtubule-dependent processes drives TDP-43 pathology in vivo. We demonstrated using ALS/FTD flies that DCTN1 knockdown not only exacerbates the neurodegeneration caused by TDP-43, but also promotes the formation of ubiquitin-positive inclusions of TDP-43 in the cytoplasm (Figs.
1,
2). DCTN1 loss and disease-linked mutations in DCTN1 alone were previously shown to cause neuronal dysfunction and eventual neurodegeneration in cultured neurons [
29],
C. elegans [
16],
Drosophila [
2,
15,
23], and mice [
7,
26,
45]. Our data showing the appearance of abnormal small vacuole-like structures in the retina of DCTN1-IR flies (Fig.
1d) is in line with these previous reports; the toxicity of DCTN1 loss was not detectable on light microscopy (Fig.
1b) or the climbing assay (Fig.
1e) in the flies expressing DCTN1-IR alone, possibly owing to the sensitivity of the experimental methods and the low knockdown efficiency in our flies. Although several reports have suggested that DCTN1 deficiency itself potentially causes detrimental effects on cell survival, it was unknown whether DCTN1 genetically interacts with TDP-43 pathology; therefore, our study is the first report to our knowledge demonstrating that DCTN1 is a genetic modifier that promotes TDP-43 aggregation in vivo. DCTN1 is a subunit of dynactin, a motor protein complex functioning in microtubules. Intriguingly, knockdown of other components of microtubule-associated motor protein complexes, such as dynein and kinesin, also exacerbated the cytoplasmic aggregation of TDP-43 in ALS/FTD flies (Fig.
5). These data indicate that intracellular transport along microtubules plays a crucial role in the development of TDP-43 pathology. In support of this, mutations in
TUBA4A were recently identified as the genetic cause of familial FTD with TDP-43 pathology [
28].
TUBA4A encodes α-tubulin, a major component of the microtubule network, and FTD patients with
TUBA4A mutations demonstrated a decreased trend of α-tubulin levels and impairment of microtubule network reformation, implying that dysfunction of microtubule-dependent cellular activities triggers TDP-43 pathology in patients. Considering that microtubule-dependent transport is likely impaired by various stresses and aging [
11,
38], our results highlight the importance of abnormalities in microtubule functions in the TDP-43 proteinopathies, including ALS/FTD.
Second, stress granule dynamics is regulated via microtubule-dependent processes in heat-stressed cells. We showed that DCTN1 deficiency does not affect the formation of but delays the disassembly of stress granules during recovery after stress (Fig.
3). Nocodazole treatment also delayed stress granule disassembly (Additional file
1: Fig. SI–2). These results suggest that intracellular transport along microtubules is necessary for the disassembly of stress granules. Importantly, DCTN1 knockdown facilitated the formation of ubiquitin-positive cytoplasmic TDP-43 aggregates in the heat-stressed cells (Fig.
4). Taken together, we propose the following model for TDP-43 aggregation: upon stress, TDP-43, together with other RNA-binding proteins and RNAs, transiently forms stress granules in the cytoplasm, which are readily disassembled in a manner that depends on intracellular transport along microtubules. Mutations in DCTN1, or dysfunction in intracellular transport along microtubules, cause the dysregulation of stress granule dynamics, which leads to the aberrant aggregation of stress granule components, including TDP-43 (Additional file
1: Fig. SI–3). Previous studies reported that degradation machineries, such as autophagy and the proteasome, the molecular chaperone Hsp70, and ubiquitination are required for the disassembly of stress granules [
3,
12,
25,
40]. Therefore, functional decline not only in microtubule-dependent transport, but also in multiple cellular processes that regulate the formation and disassembly of stress granules, may lead to the formation of TDP-43 aggregates through the impairment of stress granule dynamics. Although it remains to be clarified as to how stress granule components are sorted and cleared during recovery after stress, and how DCTN1 deficiency delays the disassembly of stress granules and promotes pathological TDP-43 aggregation, abnormalities in microtubule functions may cause impaired transport of stress granule components, leading to aberrant aggregation of TDP-43 and other components.
Finally, intracellular transport along microtubules might be a therapeutic target for the TDP-43 proteinopathies. Although it remains unclear as to how TDP-43 aggregation causes neurodegeneration in patients with ALS/FTD, suppressing TDP-43 aggregation will potentially rescue neurons from TDP-43 toxicity. Indeed, we recently showed that a transient reduction in TDP-43 levels by antisense oligonucleotides leads to the suppression of cytoplasmic TDP-43 aggregation, and long-lasting improvement in behavioral abnormalities in ALS/FTD mice [
36]. In the present study, we showed that knockdown of DCTN1 and other microtubule-dependent motor proteins facilitates TDP-43 aggregation and exacerbates neurodegeneration in vivo (Figs.
1,
2,
5). Our data indicate that the impairment of intracellular transport along microtubules is a key modifier that drives TDP-43 aggregation in the TDP-43 proteinopathies, indicating the possibility that the restoration of microtubule functions may prevent aberrant TDP-43 aggregation and neurodegeneration. Microtubule stability and functions are regulated by the post-translational modifications of tubulins, including acetylation. Promoting the acetylation of α-tubulin by the inhibition of histone deacetylase 6 (HDAC6), a major deacetylating enzyme of α-tubulin, was shown to normalize microtubule-dependent axonal transport in mouse models of Charcot-Marie-Tooth disease with HSPB1 mutations [
6], and in motor neurons derived from ALS patients with familial mutations in FUS and TDP-43 [
10,
14]. Thus, the reactivation of microtubule functions through HDAC6 inhibition might reverse not only defects in microtubule-dependent transport, but also the accumulation of TDP-43 aggregates, providing a potential therapeutic approach for the TDP-43 proteinopathies, including ALS/FTD.
Materials and methods
Fly stocks
Flies were cultured and crossed under standard conditions at 25 °C. Transgenic flies bearing the UAS-TDP-43 transgene have been previously described [
17]. Transgenic flies bearing the UAS-dDCTN1-IR (#1, 24760; #2, 24761) and UAS-grim (9922) transgenes were obtained from the Bloomington
Drosophila Stock Center, and transgenic flies bearing the UAS-DCTN2-IR (#1, 23726; #2, 110741), UAS-Dhc64C-IR (#1, 28053; #2, 28054), UAS-Dlc-IR (#1, 22760; #2, 105760), UAS-Khc-IR (#1, 44337; #2, 44338), and UAS-Klc-IR (#1, 22125; #2, 39583) transgenes were obtained from the Vienna
Drosophila Resource Center. Female flies were used in all experiments.
Fly eye imaging
The eyes of one-day-old flies were analyzed and imaged using the stereomicroscope SZX10 equipped with the digital camera DP22 (Olympus). Areas of eye pigmentation and necrotic patches were analyzed using ImageJ software (NIH). For the histological analysis of retinal structures, fly heads (1-day-old) were fixed in Carnoy’s solution, and embedded in paraffin. Eye sections of 3-µm thickness were stained with hematoxylin and eosin, and analyzed using the microscope BX51 with a CCD camera DP71 (Olympus).
Quantitative RT-PCR
The expression of DCTN1 in neurons and other tissues in
Drosophila has been reported previously [
8]. Total RNA was isolated from adult flies (five flies per sample) using the RNeasy Lipid Tissue Mini Kit (Qiagen). cDNA was synthesized from total RNA using the PrimeScript RT Master Mix (Qiagen), and real-time PCR was performed using the Mx3000P qPCR System (Agilent) and Premix Ex Taq (Takara Bio). The primer sequences were as follows:
dDCTN1 forward: 5′-AGCCGTGCCAGGTTTG-3′
dDCTN1 reverse: 5′-CGTTCGCCCTCATACA-3′
rp49 forward: 5′-AGCGCACCAAGCACTTCATCCGCCA-3′
rp49 reverse: 5′-GCGCACGTTGTGCACCAGGAACTTC-3′
Climbing assay
The climbing assay was performed as previously described [
35]. Briefly, flies at 7 days of age in a glass vial were gently tapped to the bottom of the vial, and the height of each fly after 10 s of climbing was scored as follows: 0 (less than 2 cm), 1 (between 2 and 3.9 cm), 2 (between 4 and 5.9 cm), 3 (between 6 and 7.9 cm), 4 (between 8 and 9.9 cm), and 5 (more than 10 cm). In each experiment, the trial was repeated five times at 20 s intervals, and the scores were averaged (> 30 flies).
Immunohistochemistry
Immunohistochemistry was performed as previously described [
17]. Briefly, eye imaginal discs of third instar larvae were fixed with 4% paraformaldehyde (PFA) and then incubated overnight at 4 °C with the following antibodies: anti-TDP-43 (12892–1-AP rabbit polyclonal, Proteintech), and anti-ubiquitin (FK2 mouse monoclonal, MBL). As secondary antibodies, Alexa Fluor-labeled IgGs (Thermo Fisher Scientific) were used. Hoechst 33342 was used for nuclear staining. Images were taken with the confocal laser-scanning microscope FV3000 (Olympus). Cytoplasmic inclusions of TDP-43 were analyzed using ImageJ software (NIH). Cytoplasmic inclusions were defined as inclusions with strong fluorescence signals that are localized around Hoechst 33342 nuclear staining, but do not overlap with the Hoechst staining.
Detergent solubility assay
Five adult fly heads were lysed with a lysis buffer containing 1% Triton X-100 supplemented with protease inhibitors (Nacalai Tesque), and centrifuged at 20,000 × g for 10 min at 4 °C to separate the supernatant (S1) and pellet fractions. The pellets (insoluble in Triton X-100) were suspended in a buffer containing 1% sarkosyl, sonicated, and centrifuged at 20,000 × g for 10 min at 4 °C. The supernatants were removed as the sarkosyl-soluble fraction S2, and the pellets were solubilized in 9 M urea (the sarkosyl-insoluble fraction P2).
FCS measurements
Ten adult fly heads were lysed in a lysis buffer containing 0.5% Triton X-100 supplemented with protease inhibitors (Nacalai Tesque), and were centrifuged at 10,000 ×
g for 30 min at 4 °C. The supernatant fractions were subjected to FCS measurements. FCS measurements were performed at room temperature using the FlucDEUX system with a × 40 objective lens (NA1.15, water immersion) (MBL). The autocorrelation function G(τ) was calculated and fitted by FCCS Editor software (MBL) using Eq.
1,
$$\text{G(}\tau \text{) = }\frac{\langle I\left(t\right)I(t+\tau )\rangle }{{\langle I(t)\rangle }^{2}}=1+\frac{1}{N}{(1+\frac{\tau }{{\tau }_{i}})}^{-1}{(1+\frac{\tau }{{s}^{2}{\tau }_{i}})}^{-\frac{1}{2}}$$
(1)
where <
I(t) > is the average fluorescence intensity (photons per second),
N is the number of particles in the detection area, and
s is the structural parameter based on the dimensions of the confocal volume. The correlation time (τ
i), which is the average time for diffusion of fluorescent particles across the detection area, correlates with the size of the fluorescent particles.
Cell culture, transfection, and immunocytochemistry
Human embryonic kidney 293 (HEK293) cells were cultured in DMEM supplemented with 10% fetal bovine serum at 37 °C under 5% CO2. For heat shock, cells were incubated at 42 °C for 60 min. For RNAi-mediated knockdown, cells were transfected with siRNA against the human DCTN1 gene or nontargeting control siRNA (Dharmacon) using Lipofectamine RNAiMAX (Invitrogen), and incubated for 48 h. For immunocytochemistry, cells were fixed with 4% PFA, permeabilized with 0.3% Triton X-100, and incubated with the following antibodies: anti-TDP-43 (12892–1-AP rabbit polyclonal, Proteintech), anti-G3BP1 (611127 mouse monoclonal, BD), and anti-ubiquitin (FK2 mouse monoclonal, MBL). As secondary antibodies, Alexa Fluor-labeled IgGs were used. Hoechst 33342 was used for nuclear staining. Images were taken with the confocal laser-scanning microscope FV3000. Cells with TDP-43- and G3BP1-positive dots (stress granules), and TDP-43 inclusions with ubiquitin staining (TDP-43 aggregates) were analyzed using ImageJ software.
Western blot analysis
Proteins were separated using 5–20% gradient sodium dodecyl sulfate–polyacrylamide gel electrophoresis gels (Atto), and transferred onto polyvinylidene fluoride membranes (Bio-Rad). Membranes were incubated overnight at 4 °C with the following antibodies: anti-TDP-43 (10782–2-AP rabbit polyclonal, Proteintech), anti-DCTN1 (sc-365274 mouse monoclonal, Santa Cruz), and anti-actin (AC-40 mouse monoclonal, Sigma). As secondary antibodies, horseradish peroxidase (HRP)-conjugated IgGs (Jackson Immuno Research Laboratory) were used. The HRP signal was visualized using ImmunoStar Zeta (Wako), and captured using Amersham Imager 600 (Cytiva). Images were analyzed using ImageJ software.
Statistical analyses
Data in the fly experiments, including data from quantitative RT-PCR, eye pigmentation and necrotic patch analyses, the climbing assay, and quantification of TDP-43 inclusions by immunohistochemistry, were analyzed by one-way analysis of variance (ANOVA) followed by the Dunnett multiple comparison test (Figs.
1a and c,
2b,
5b and d) or by the Tukey multiple comparison test (Fig.
1f). The data in the cell culture experiments, including densitometric analysis of the immunoblot images (Fig.
3c), quantification of stress granules (Fig.
3f and i), and TDP-43 aggregation (Fig.
4d), were analyzed by the Student
t-test. For all analyses, GraphPad Prism software was used. A
p-value of less than 0.05 was considered to indicate a statistically significant difference between groups.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.