• depolarization;
  • neuronal dendrites;
  • processing body;
  • RNA granules;
  • RNA-binding proteins;
  • TDP-43


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

TDP-43, recently identified as a signature protein of the pathogenic inclusions in the brains cells of frontotemporal lobar degeneration patients, is a 43 kDa RNA-binding protein. It has been known mainly as a nuclear factor capable of repressing transcription and promoting exon exclusion. TDP-43 also forms distinct nuclear substructures linking different types of nuclear bodies. In this study, we provide the first evidence supporting TDP-43 as a neuronal activity-responsive factor in the dendrites of hippocampal neurons. In particular, TDP-43 resides in the somatodendrites mainly in the form of RNA granules colocalized with the post-synaptic protein PSD-95. These granules also contain RNAs including at least the β-actin mRNA and CaMKIIα mRNA. Furthermore, TDP-43 is localized in the dendritic processing (P) body and it behaves as a translational repressor in an in vitro assay. Related to this, repetitive stimuli by KCl greatly enhance the colocalization of TDP-43 granules with FMRP and Staufen 1, two RNA-binding proteins known to regulate mRNA transport and local translation in neurons. These data together suggest that TDP-43 is a neuronal activity-responsive factor functioning in the regulation of neuronal plasticity, the impairment of which would lead to the development of certain forms of neurodegenerative diseases including frontotemporal lobar degeneration.

Abbreviations used

calcium calmodulin dependent kinase


frontotemporal lobar degeneration


glutathione S-Transferase




PML oncogenic domains


RNA recognition motif


TAR DNA binding protein

TAR DNA binding protein (TDP)-43 is a 43 kDa protein containing two RNA recognition motifs (RRM), RRM I and RRM II. It was originally identified as a human immunodeficiency virus TAR DNA-binding protein (Ou et al. 1995) behaving like a strong transcription repressor in vitro (Ou et al. 1995) and in transfected cells (Wang et al. 2002). TDP-43 has also been shown to be involved in an exon-exclusion reaction (Buratti et al. 2001a; Wang et al. 2004) of the cystic fibrosis transmembrane receptor gene transcript through preferential binding to segment(s) with UG-rich sequences on the RNA (Buratti and Baralle 2001b). The RRM I motif and the glycine-rich domain are required for the DNA-binding, RNA-binding, and/or exon exclusion activities of TDP-43 (Buratti and Baralle 2001b; Buratti et al. 2001a; Wang et al. 2004). Notably, proteomic analysis has identified TDP-43 as one of the proteins in the RNA granules (Elvira et al. 2006). The involvement of TDP-43 in splicing is consistent with the observation that nuclear TDP-43 formed distinct substructures linking other types of nuclear bodies including the speckles, CB and PML oncogenic domains (POD) (Wang et al. 2002). TDP-43 is encoded by the TDP gene well conserved in the mammals, flies, and Caenorhabditis  elegans (Wang et al. 2004). In mammals, the TDP gene is transcribed and spliced to generate 11 alternatively spliced transcripts, among which the one encoding TDP-43 is the major mRNA species. TDP-43 is a widely expressed protein, suggesting it is an essential gene functioning in different tissues. More recently, TDP-43 was identified as the major component of the ubiquitinated cytoplasmic, nuclear, and neuritic inclusion in the brains of the sporadic and familial frontotemporal lobar degeneration (FTLD-U), and of the sporadic amyotrophic lateral sclerosis (Neumann et al. 2006, 2007; Seelaar et al. 2007; Tan et al. 2007). This finding is highly suggestive of the TDP-43 functioning in important neuronal activities, the impairment of which would lead to the pathological progression of specific neuronal degeneration.

The molecular and cellular basis of the functions, e.g. learning and memory, of the CNS has been studied using various rodent-based laboratory systems. For example, cultured hippocampal neuron cultures derived from the rodent brains have been used to examine the events leading to changes of the shape, structure, and number of the dendrites upon different electrical and chemical stimulations (Zhou et al. 2004; Bingol and Schuman 2006), some of which were repetitive and spaced, mimicking the repetitive learning and experience storage (Wu et al. 2001; Li et al. 2004). Among the events is the nuclear transcriptional activation as well as local translation in the dendrites and synapses (Martin et al. 2000). Increases of dendritic transport of different mRNAs and RNA-binding protein complexes were observed upon neuronal activation (Miller et al. 2002; Tiruchinapalli et al. 2003; Antar et al. 2004), suggesting a role of the RNA-binding proteins in the reciprocal interactions between trafficking and local protein synthesis needed for long-lasting plasticity. (Schratt et al. 2004; Sutton and Schuman 2006) Newly synthesized dendritic proteins including the cytoskeleton proteins, e.g. β-actin, and synaptic proteins are required for the stabilization of new synapses (Kiebler and DesGroseillers 2000; Steward and Worley 2001). In the neuronal cells, the local translation occurs upon transport of specific mRNAs and RNA–protein complexes (Huang et al. 2005) in the dendrites and to the spines (Smart et al. 2003; Antar et al. 2004).

In consistency with the above scenario of regulation of neuro-plasticity, loss of the RNA-binding protein(s) have been shown to affect memory. For example, fragile X mental retardation protein (FMRP)-knockout mice as well as human patients with the fragile X syndrome and carrying defective FMR1 gene showed memory disorders and abnormalities in their dendritic spine maturation (Nimchinsky et al. 2001; Miyashiro and Eberwine 2004). The Drosophila Staufen is a double-stranded RNA-binding protein important for long-term memory of the fruit flies (Dubnau et al. 2003). The mammalian orthologs of the fly Staufen, Staufen 1 and Staufen 2, have demonstrated roles in mRNA transport in the neurons (Tang et al. 2001). They are also associated with the polysomes (Dugre-Brisson et al. 2005). Nevertheless, how the RNA-binding proteins such as Staufen and FMRP regulate translation and activity-dependent formation and storage of memory have remained unclear.

In the following, we show that, in striking similarity to other known RNA-binding proteins functioning in the neurons, TDP-43 resides in the neuronal somatodendrites mainly in the form of RNA granules containing specific mRNAs. It also behaved as a translation repressor in an in vitro system. Interestingly, the number of the TDP granules as well as its dendritic colocalization with FMRP and Staufen 1 increased upon stimulation by different neuronal activities, thus establishing TDP-43 as a neuronal activity-responsive factor.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Primary neuronal culture and drug treatment

Hippocampal neuron cultures were prepared from E18–19 rat embryos (Goslin and Banker 1998). Neurons were cultured at low density (10 000 cells/cm2) in the neurobasal medium/B27 (Invitrogen Co., Carlsbad, CA, USA) in dish coated with poly-l-lysine (1 mg/mL). Neurons were plated on the coverslides for immunocytochemisrty. Staining by SYTO 14 (Invitrogen-Molecular Probe, Carlsbad, CA, USA) was for 5 min. Rnase treatment for 1 h at 37°C after fixation was carried out as a control as described (Knowles et al. 1996). For depolarization by single KCl stimulus, the cultured neurons (days in vitro 8–9 or 21, see text) were incubated with 50 mmol/L KCl for 5 min. For repetitive stimulation by KCl (Wu et al. 2001), the neurons (days in vitro 8–9) were pre-incubated with 1 μmol/L tetrodotoxin for 2 h and then exposed to repetitive stimuli by KCl. Each stimulus consisted of a treatment with 90 mmol/L KCl for 3 min and a spaced recovery for 10 min. Forskolin and NMDA were obtained from Sigma Co (St. Louis, MO, USA).

Immunohistochemistry and antibodies

For immunostaining with antibodies, the cells were fixed with 3.7% formaldehyde in phosphate-buffered saline at 4°C for 15 min and then permeabilized with 0.1% Triton X-100. The antigen localization was determined after incubation of the cells with appropriate antibodies at 4°C overnight. The anti-post synaptic density-95, anti-FMRP, and anti-Staufen 1 antibodies were from Upstate Biotech (Charlottesville, VA, USA). Anti-GW182 antibody was from Santa Cruz Inc (Santa Cruz, CA, USA). The secondary antibodies conjugated to the fluorescein, Cy5 or Cy3, were applied for 2 h at 22–25°C. The coverslides were mounted with Fluoromount G (Fischer Scientific, Pittsburgh, PA, USA) and the fluorescent images were analyzed in a Zeiss laser confocal microscope (LSM510) (Zeiss, Oberkochen, Germany). Each confocal image is from single optical section.

Image analysis

The lengths of the dendrites and the numbers of puncta were traced and measured with the MetaMorph software. This software was also used to identify the ‘bright’ TDP-43 granules. For a granule to be ‘bright’, we define it as having a relative intensity of fluorescence above 100 within the total range of 250. The TDP-43-, FMRP-, and Staufen 1 puncta were outlined in Fig. 5, and their sizes were measured by the MetaMorph. To analyze colocalization in the dendrites, the confocal images were quantitatively defined using the LSM510 program.


Figure 5.  Repetitive stimuli-induced assembly of multi-RNA binding proteins in the same granules. The primary hippocampal neurons (days in vitro 8–9) were repetitively stimulated with KCl and then double-stained with anti-TDP-43 (red)/anti-FMRP (green) (a) or with anti-Staufen 1 (red)/anti-FMRP (green) (b). Examples of the fluorescent patterns are shown for the untreated neurons (top row), and neurons treated one time (middle row) or four time (bottom row) with KCl. The bar in the top left panel of (a) is 30 μm. Examples of granules with colocalized pair of RNA-binding proteins in the bottom panels are indicated with arrows. In both (a and b), the changes of the numbers of the granules and the colocalization coefficients during the stimuli are shown in the upper and lower diagrams, respectively. Note the significantly large increases of the colocalization coefficients in the four time stimulated neurons. Experiments were performed in triplicate; *p < 0.05 versus control.

Download figure to PowerPoint

Immunoelectronic cytochemistry

The procedures followed those of Ostroff et al. (2002). Male mouse brain were perfused and then fixed by incubation at 4°C overnight with cold 0.1 mol/L phosphate–citrate buffer (pH 7.2) containing 1% glutaraldehyde and 1% formaldehyde. Immunogold-labeling was processed by a modified procedure (Wang et al. 2006). The ultrathin sections were blocked with 3% normal goat serum and then incubated with the anti-TDP-43 antibody. The sections were finally treated with 2% uranyl acetate and 30 mmol/L lead citrate. The samples were examined using a Zeiss EM109 electron microscope.


For characterization of TDP-43 associated RNAs in the dendrites, the cultured rat hippocampal neurons were first separated into nuclear and cytosolic fractions and then lysed. The RNA–protein complexes in the lysates were UV cross-linked and then pulled down by the anti-TDP-43 antibody using the conditions of Ule et al. (2003). The immunoprecipitates bound to the protein A/G beads were washed and the proteins were analyzed by western blotting using antibodies against TDP-43, GAP-43 (anti-GAP-43; Chemicon, Pittsburgh, PA, USA), and RNA polymerase II (8WG16; Cold Spring Harbor Lab., NY, USA), respectively, as the probes. The hybridizing bands were detected with use of the enhanced chemiluminescence detection system (Perkin-Elmer, Walthan, MA, USA). The RNAs in the immunoprecipitates were purified as described (Ule et al. 2003) and then precipitated with glycogen (Roche, Nutley, NJ, USA). The RNAs were then reverse transcribed using the oligo dT primers (SuperScriptII first strand synthesis system; Invitrogen Co.). PCR was then performed using specific primers for 25–33 cycles of 30 s at 94°C/30 s at 56°C/1 min at 72°C. The primers used were: for β-actin (from exons 3 and 6), 5′-GCCATCCTGCGTCTGGACCT-3′ and 5′-GCCACCAATCCACACAGAGT-3′; for calcium calmodulin dependent kinase (CaMKIIα) (from exons 4 and 13), 5′-GAACTTCTCCGGAGGGAAG-3′ and 5′-CGCATCCAGGTACTGAGTG-3′; for heat-shock protein70-1a, 5′-GTGTGCAACCCGATCATCAG-3′ and 5′-TTAAGTTACAAAAGATAACT-3′. The RNA-immunoprecipitation (IP) experiments were repeated three times with similar results.

In vitro translation assay

The assay followed the procedures of Lin et al. (2007) with some modifications. Plasmid pFR-SV40-Luc (a kind gift from Dr Woan-Yuh Tarn) was linearized and used as the template for in vitro transcription to generate diguanosine-capped transcripts without the poly(A) tail. Each in vitro translation reaction (25 μL) contained 10 ng of purified, in vitro-transcribed RNA, 0.5–20 pmol of glutathione S-transferase (GST) or GST-tagged mouse TDP-43, 25 μmol/L amino acids, and 12.5 μL of the reticulocyte lysate (Promega). The reactions were incubated at 30°C for 1 h, and 10% of each reaction mixture was subjected to the luciferase assay. Each reaction was repeated three times.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Depolarization-driven translocation of TDP-43-containing granules into the dendritic spines of mature hippocampal neurons

We first examined the distribution patterns of TDP-43 expression in different mouse tissues. For this, we generated a rabbit antibody against the highly conserved amino acid 79–96 (KDNKRKMDETDASSAVKV) of the N terminal of mouse TDP-43. This sequence is identical between mouse and human, and differs by only one amino between mouse and rat. When the antibody was used as the probe for western blotting analysis, a 43 kDa band was detected in the mouse brain lysate while use of the pre-immune serum showed no signal (compare the left two panels of Fig. 1a). TDP-43 appeared to be a widely expressed protein detectable in the mouse brain, heart, kidney, lung, uterus (right panel of Fig. 1a), and other tissues including the muscle (data not shown).


Figure 1.  Depolarization driven translocation of TDP-43 into the dendritic spines of the mature hippocampal neurons. (a) Western blot analysis of the expression patterns of TDP-43 in different mouse tissues. In the left two panels, the total lysate from adult mouse brain was probed with pre-immune serum and anti-TDP-43, respectively. The fainter 52 kDa band was also seen by others but its origin is unknown. For the tissue comparison (the right panels), β-actin was used as the control. (b) Immunostaining pattern of TDP-43 in the hippocampal neuron layers of adult mouse brain. Both a lower and a higher magnification pictures are shown. DG, dentate gyrus. (c–d) Colocalization of TDP-43 and PSD-95 in the mature rat hippocampal neurons (days in vitro 21) in culture. The colocalization coefficients before and after depolarization with 50 mmol/L KCl are compared in (c). The colocalization (yellow) patterns of TDP-43 (green) and PSD-95 (red) without (a and c) and with (b and d) depolarization are exemplified in (d). (c and d) are magnified photographs of the boxed areas in (a and b), respectively. The bars are 20 μm long. (e) TDP-43 clusters in the adult mouse brain as analyzed by immunoelectronic chemistry. The low magnification pictures of staining with the pre-immune serum and anti-TDP-43 are shown in the upper left and upper middle panels, respectively. Shown in the upper right panel is a high magnification picture of TDP-43 clusters in the dendrites. Two examples of the spine localization of TDP-43 are shown in the lower left and lower middle panels. The shaded lines in the two cartoons indicate the post-synaptic density areas. The length bars are indicated in each of the panels. Experiments were performed in triplicate; *p < 0.05 versus control.

Download figure to PowerPoint

Because of its close association with the neurodegenerative disease FTLD-U (Neumann et al. 2006), we have analyzed the expression patterns of the TDP-43 in the mouse brain by immunohistochemistry. The signals of TDP-43 were enriched in the hippocampus, especially in the neuronal layer (Fig. 1b). This result was consistent with observation from the in situ hybridization analyses (I.-F. Wang, data not shown; also see the Allen Brain Atlas described in Lein et al. 2007). The hippocampus-enriched distribution of TDP-43 suggested its potential role in neuronal plasticity. We thus analyzed the distribution of TDP-43 in the mouse hippocampal neurons by immunostaining with or without stimulation by 50 mmol/L KCl. In comparison to the staining pattern of the pre-immune serum (see Fig. 2a in a latter section), use of anti-TDP-43 gave unique immunostaining patterns in the cultured hippocampal neurons, with the signals in the nucleus much stronger than those in the cytosol (Fig. 1d). This was consistent with the western blot data which showed that approximately 15% of TDP-43 molecules were located in the cytosol (see Fig. 2e in a latter section). Furthermore, as shown in Fig. 1c and d, the majority of TDP-43 in the dendrites appeared as punctuated clusters/granules before depolarization, and a fivefold increase in its colocalization with post-synaptic density-95, a dendritic spine marker, was observed upon stimulation by KCl. Exogenously expressed TDP-43 tagged with green fluorescent protein exhibited a similar pattern of distribution in transfected rat hippocampal neurons (data not shown). These results together indicated that TDP-43 in the dendrites existed as clusters/granules, which could be driven by neuronal activities, such as the depolarization process, into the spines.


Figure 2.  Regulation of the TDP-43 RNA granule formation by neuronal activities. (A-C) Immunostaining patterns of the hippocampal neurons with the use of SYTO 14 (red) and TDP-43 (green). (a) Staining patterns of pre-immune serum (i) and anti-TDP-43 (ii). The blue color is from DAPI. (b) The patterns of SYTO 14 staining without (i) and with (ii) Rnase digestion. In (c), the neurons without (i–iv) and with the treatment of 50 mmol/L KCl (v–viii), 25 mmol/L NMDA (ix–xii), and 10 μmol/L forskolin (xiii–xvi), respectively, were analyzed. The photographs in (iv, viii, xiii, and xvi) are magnified pictures of the boxed areas in (iii, vii, xi, and xv), respectively. As examples, one each of the colocalization areas (yellow) in (iv and viii) are indicated by the arrows. (d) Statistical counts of the colocalization coefficients of SYTO 14/TDP-43 clusters as analyzed by a LSM program. (e) RNA-IP analysis of the TDP-43 granules. (i) The primary hippocampal neurons (days in vitro 8–9), before and after 50 mmol/L KCl treatment, were separated into the cytosol (Cyto) and nuclear (Nu) fractions. The purities of the fractions were validated by western blot (WB) showing the presence of GAP-43 in Cyto and RNA polymerase II (RNAP II) in Nu (lower two panels). The RNA–protein complexes were then immunoprecipitated with anti-TDP-43. The immunoprecipitates were analyzed by western blotting with anti-TDP-43 (third panel from bottom), and by RT-PCR analysis of the bound RNAs using primers specific for mRNAs of β-actin, CaMKII, and heat-shock protein70-1a (HSP) genes. Note the enrichment of the β-actin mRNA and CaMKII mRNAs in the cytosol after the 50 mmol/L KCl treatment (compare lanes 3 and 4 of the top three panels). The conclusion was derived from three sets of independent experiments. (ii) Comparison of RT-PCR analysis of the CamKII mRNA levels in samples precipitated with IgG and with anti-TDP-43. (iii) Semi-quantitative RT-PCR analysis of actin mRNA and CamKII mRNA in the cytosolic RNAs from neurons without (control) and with 50 mmol/L KCl treatment. Experiments were performed in triplicate; *p < 0.05 versus control.

Download figure to PowerPoint

In order to see if the association of TDP-43 with the dendritic spines also existed under physiological conditions, we have carried out immunohistochemistry analysis of ultrathin-sections of the mouse brain in the electron microscope. While use of pre-immune serum gave few signals (upper left panel, Fig. 1e), significant number of gold particles could be seen in the soma and dendrites when anti-TDP-43 was used, suggesting the clustering of the TDP-43 molecules in these areas (upper middle and right panels, Fig. 1e). Approximately 2% of the gold particles in the soma and dendrities were found in the dendritic spines with the clusters of TDP-43 located near the post-synaptic density, as exemplified in the lower panels of Fig. 1e. The data of Fig. 1 together suggested that TDP-43 very likely played a regulatory role in the neuronal plasticity, and especially those occurring during the processes of learning and memory.

Neuronal activity-regulated association between TDP-43 and dendritic RNAs

As TDP-43 is a RNA-binding protein containing well-conserved RRM I motif, we have examined whether the dendritic TDP-43 clusters contained RNAs by immunostaining of the cultured hippocampal neurons with anti-TDP-43 and simultaneous staining with SYTO 14, the latter of which specifically marked the RNAs (Bassell et al. 1998). As exemplified in Fig. 2a, while immunostaining with anti-TDP-43 gave specific patterns, the use of pre-immune serum showed few signals [compare Fig. 2a(ii) with a(i)]. Also, the signals of the SYTO 14 staining, in particular those in the cytosol, declined significantly upon Rnase treatment, suggesting that SYTO 14 specifically labeled the RNAs [compare Fig. 2b(i and ii)]. As exemplified in Fig. 2c, 70% of the TDP-43-containing clusters were colocalized with SYTO 14 [Fig. 2c(i–iv)]. Interestingly, the percentage of colocalization increased to nearly 90% upon the stimulation with KCl [Fig. 2c(v–viii) and d]. The result was consistent with the enhancement of TDP-43 binding with specific cytosolic mRNA upon KCl stimulation (see below). In contrast, the colocalization was significantly decreased to 40% in neurons treated with 25 mmol/L of NMDA [Fig. 2c(ix–xii) and d] or 10 μmol/L of forskolin [Fig. 2c(xiii–xvi) and d]. Taken together, these results showed that the association between TDP-43 and RNAs in the neuronal dendrites was dynamic in response to different neuronal activities.

In view of the significant colocalization of the dendritic TDP-43 clusters with RNAs, as described above, we have examined whether TDP-43 was associated with specific mRNA species. For this, we used anti-TDP-43 and the RNA-IP procedure to immunoprecipitate the cytosolic and nuclear extracts prepared from the rat hippocampal neuronal cultures with or without depolarization by 50 mmol/L KCl (Fig. 2e). The fractionation of the nuclei and cytosol was validated by western blotting with antibodies against the dendritic marker GAP-43 and the nuclear marker RNA polymerase II [lower two panels, Fig. 2e(i)]. The western blotting result also showed that the nuclear TDP-43 amount was approximately 8- to 10-fold higher than that of the cytosolic TDP-43 [third panel from the bottom, Fig. 2e(i)].

The RNAs in the immunoprecipitates were then analyzed by RT-PCR as described in the Materials and methods. As exemplified in the upper three panels of Fig. 2e(i), weak signals of actin mRNA could be detected in the immunoprecipitates of both the nuclear and cytosolic fractions of the cultured neurons [lanes 1 and 3 of the top panel, Fig. 2e(i)]. The relative amount of the actin mRNA in the cytosolic fraction increased upon 50 mmol/L KCl treatment of the hippocampal cultures [lane 4 of the top panel, Fig. 2e(i)]. In addition, the mRNA of CaMKIIα could be detected in the same cytosolic immunoprecipitate upon depolarization, but not in the nuclear immunoprecipitate [compare lane 4 with lanes 1, 2, and 3 of the second panel, Fig. 2e(i)]. As the controls, no RT-PCR signal could be detected when pre-immune serum was used for RNA-IP [Fig. 2e(ii)]. Also, the quantities of the CamKII mRNA in the cytosol from cultured neurons with or without depolarization did not differ much [Fig. 2e(iii)]. The data of Fig. 2c, d, and e together suggested that dendritic ribonucleoproteins containing TDP-43 became more abundantly associated with specific RNAs in response to neuronal activation processes such as the depolarization.

Translational repression by TDP-43 in vitro and localization of TDP-43 in the dendritic P bodies

The characteristics and behaviors of TDP-43 in the neuronal dendrites, as described in Figs 1 and 2, were similar to two other RNA-binding proteins, FMRP and Staufen (Kiebler et al. 1999; Antar et al. 2004). In particular, these two proteins and several others implicated in the regulation of translation process were found to be concentrated in the processing body or P body (Barbee et al. 2006), a cytosolic substructure involved in RNA induced silencing complex-mediated translation repression and mRNA degradation/storage (Parker and Sheth 2007). Therefore, we have examined the effect of TDP-43 on translation by an in vitro assay, as well as the relationship between the dendritic TDP-43 granules and P bodies.

As shown in Fig. 3a, in comparison with GST, the GST-TDP-43 fusion protein significantly inhibited the translation in vitro in a dose-dependent manner. In addition, we have searched the protein-interaction databank (Database: Binding) and found that TDP-43 also could directly interact with the elongation factor ElF3s61p and ribosomal proteins. In addition, we have carried out co-immunostaining experiments of the cultured neurons with the use of anti-TDP-43 and GW182, a marker of the P body (Liu et al. 2005). As shown in Fig. 3b, partial colocalization of the TDP-43 and GW182 signals in the dendrites was observed. Similarly, colocalization between TDP-43 and eIF4E, another P body component, could be detected (data not shown). The results of Fig. 3 suggested that TDP-43, through association with the translational apparatus, could participate in the regulation of the local translation processes in the neuronal dendrites.


Figure 3.  Association of TDP-43 with the translation apparatus. (a) Translational repression in vitro by TDP-43. The in vitro translation reaction was performed as described in the Materials and methods. The effects of GST-TDP-43 and GST, respectively, on the translation efficiency are plotted as a function of the amounts (pmol) of proteins used in the reaction mixtures. The error bars are the plus SD values. (b) Depolarization-induced colocalization of TDP-43 with GW182. The hippocampal neurons (days in vitro 21) were double-immunostained with anti-TDP-43 (green) and anti-GW182 (red).

Download figure to PowerPoint

Step-wise increase in the TDP-43 clusters in dendrites upon repetitive stimulation with KCl

Using the model system established by Wu et al. (2001), we then tested the behavior of TDP-43 clusters/granules during memory strengthening in primary cultured hippocampal neurons treated with repetitive and spaced stimulations by 90 mmol/L KCl. Similar to the findings by others (Wu et al. 2001; Li et al. 2004), depolarization indeed induced the long-lasting extracellular signal regulated kinase 2 in the cytosol. This was coupled with alterations of the structure of the spines protruding from the shaft, with an increase in elongated mushroom-spines after four times of stimuli but not single stimulation (data not shown). Interestingly, as exemplified in Fig. 4a, there were two types of TDP-43 clusters existing in the dendrites, one of which was relatively weaker in their fluorescence intensities and also looser in the structural appearance (arrowheads, Fig. 4a). The TDP-43 clusters of the other type were ‘bright’ and compact granule-like, and their diameters were mostly in the range of 0.25–0.75 μm (the arrows, Fig. 4a). Significantly, step-wise increases of the number of TDP-43 clusters of either class were prominent in the hippocampal dendrites during the repetitive depolarization (Fig. 4a and b). Although the molecular differences between the two types of TDP-43 granules are unclear at the moment, the data of Fig. 4 suggested that the induced memory of the neurons, as caused by the repetitive stimulations, was coupled with the stepwise accumulation of the TDP-43 clusters in the dendrites.


Figure 4.  Increase in the number of TDP-43 clusters by repetitive KCl stimulation. The primary cultured hippocampal neurons were stimulated repetitively with KCl. Each stimulus consisted of exposure to 90 mmol/L KCl for 3 min and a recovery time of 10 min before the next stimulus. The neurons were then stained with anti-TDP-43. The bar in (a) is 10 μm. Two examples each of the ‘bright’ (arrows) and the ‘weak’ clusters (arrowheads) are indicated in the lower panels. The numbers of the total TDP-43 dots, or granules, and those ‘bright’ ones, respectively, located in the dendrites at a distance 50–100 μm from the soma were counted with the MetaMorph software. Each bar in (b) indicates the average number of TDP-43 dots/clusters per micrometer, as calculated from counting of 10 different dendrites. Experiments were performed in triplicate; *p < 0.05 versus control.

Download figure to PowerPoint

Repetitive stimuli induced reassembly of granules containing multiple RNA-binding proteins including TDP-43

The data of Fig. 4 showed that the number of TDP-43 granules significantly increased after repetitive KCl stimuli. We have thus tested whether similar increases could be observed for RNA granules containing FMRP and Staufen 1, respectively. As shown in Fig. 5, the numbers of the FMRP granules (upper right graph, Fig. 5a) and Staufen 1 granules (upper right graph, Fig. 5b) also increased significantly in a stepwise manner during the repetitive KCl stimuli. Remarkably, the colocalization of different RNA-binding proteins in the same granules also significantly increased, as shown for the pairs of TDP-43/FMRP (Fig. 5a) and FMRP/Staufen 1 (Fig. 5b). The data of Fig. 5 suggested that repetitive stimuli were accompanied by a structural reassembly of the RNA granules in the dendrites, with multiple RNA-binding proteins getting assembled into the same granule(s). This reassembly could be one of the important steps responsible for the functional consequences, e.g. long term potentiation, of the repetitive stimuli of the neuronal network.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

In this study, we have analyzed several characteristics of the RNA-binding protein TDP-43, in particular its behaviors in dendrites in response to neuronal stimuli. Previous reports on TDP-43 were mostly on its nuclear properties and functions, which included transcriptional repression (Ou et al. 1995; Wang et al. 2002), enhancement of exon-exclusion (Buratti and Baralle 2001b; Buratti et al. 2001a; Wang et al. 2004), and formation of distinct nuclear substructures linking other nuclear bodies (Wang et al. 2002). The data presented above showed that TDP-43 could also carry important functions as a protein in the cytosol. More importantly, the study has suggested, for the first time, the role of TDP-43 as a new player in the modulation of neuronal plasticity, similar to the well-known RNA-binding proteins including FMRP and Staufen 1 (Jin et al. 2004).

While TDP-43 is a ubiquitous protein expressed in different tissues, in brain it is preferentially expressed in the hippocampal neurons in addition to the cerebellum (Fig. 1b; Lein et al. 2007). Furthermore, the TDP-43 granules were observed in the neuronal dendrites including the post-synapses (Fig. 1c–e). This pattern of dendritic distribution is very similar to other RNA-binding proteins known to be essential components for the regulation of the neuronal plasticity, including FMRP, Staufen 1, and zip code-binding protein 1/IGF-II mRNA binding protein 1 (Eom et al. 2003; Mallardo et al. 2003; Antar et al. 2004). Indeed, similar to these other proteins (Antar et al. 2004), depolarization by KCl stimulation increased the number of TDP-43 granules in the dendrites of the cultured hippocampal neurons. In particular, stepwise increase in the dendritic TDP-43 granules was observed upon the repetitive and spaced stimuli of the cultured neurons with KCl (Fig. 4), a system mimicking the repetitive learning and experience storage (Wu et al. 2001). These data imply that TDP-43 participates in the processes of learning/memory and experience storage.

What might the mechanisms underlying the neuronal function of TDP-43 be? Our data suggested that, similar to FMRP, Staufen and ZBP1, TDP-43 might play a role in the regulation of transport of mRNAs and local translation in the dendrites upon neuronal stimuli. This was implicated by the observation that a majority of the TDP-43 granules contained RNAs (Fig. 2c) and the number of these RNA granules changed under different neuronal activities (Fig. 2d). More interestingly, like other RNA-binding proteins, e.g. FMRP with FMR1 mRNA (Antar et al. 2004), Staufen with Drosophila Propero mRNA (Broadus et al. 1998), ZBP1 and FMRP with β-actin mRNA (Rackham and Brown 2004), FMRP with CaMKIIα mRNA (Zalfa et al. 2003), TDP-43 was associated with specific mRNAs, including at least the β-actin mRNA and CaMKIIα mRNA, in a depolarization-dependent manner (Fig. 2e). Both mRNAs were previously found to be increased upon depolarization, and the increase resulted in local new protein synthesis that was required for restructuring of the synapses (Sutton and Schuman 2006). In interesting connection with the above, TDP-43 behaved in vitro like a repressor of translation (Fig. 3a) and the colocalization of TDP-43 granules with GW182, a P body component (Andrei et al. 2005), was observed in dendrites of the cultured neurons (Fig. 3b). As the known functions of P-body included, besides mRNA degradation and mRNA storage, microRNA-mediated repression of translation (Fillman and Lykke-Andersen 2005; Bhattacharyya and Filipowicz 2007), it is worthy in the future to examine whether, upon neuronal activation, TDP-43 is involved in the targeting of specific mRNAs to the P bodies for switching-off their translation. In any case, TDP-43 appears to belong to the same category of RNA-binding proteins, such as FMRP and Staufen 1, regulating the transport of mRNAs and/or local translation in response to different neuronal activities.

Interestingly, we have uncovered that increase as well as redistribution of dendritic granules containing different RNA-binding proteins, including TDP-43, FMRP, and Staufen 1, occurred during the repetitive and spaced stimuli of the cultured hippocampal neurons (Fig. 5). Related to this observation, colocalization of FMRP and Staufen 1 in P body like-granular structures in the dendrites has been noticed recently (Barbee et al. 2006). FMRP also colocalized with Pur α in the dendrites of day 21 neurons (Kanai et al. 2004). The data of Fig. 5 thus suggested that during the repetitive stimuli, TDP-43 might function synergistically with FMRP and Staufen 1 in the formation of dendritic granules which very likely were required for the transport of mRNAs and/or reassembly functioning of the translation apparatus. The details of the co-operative functioning of TDP-43 with these other RNA-binding proteins in the dendrites await further investigation.

In summary, we have provided evidence for TDP-43 participating in the modulation of neuronal functions as a dynamically regulated granular protein in the dendrites. Further experimentation will better identify the molecular and cellular basis of the role of TDP-43 in the regulation of the neural plasticity, and the consequences of impairment of its functions in patients with FTLD-U and amyotrophic lateral sclerosis.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

We thank Dr M. M. Poo (U. C. Berkeley) for helpful suggestions at the beginning stage of this work, and members of Dr W.-Y. Tarn’s laboratory for their help with the in vitro translation assay. We also thank members of the Shen laboratory for discussions and comments on the manuscript. This research was supported by the National Health Research Institute, National Science Council and Academia Sinica, Taipei, Taiwan. I-FW was an Academia Sinica Post-doctoral Fellow, and C-KJS is an Academia Sinica Investigator Awardee.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  • Andrei M. A., Ingelfinger D., Heintzmann R., Achsel T., Rivera-Pomar R. and Luhrmann R. (2005) A role for eIF4E and eIF4E-transporter in targeting mRNPs to mammalian processing bodies. RNA 11, 717727.
  • Antar L. N., Afroz R., Dictenberg J. B., Carroll R. C. and Bassell G. J. (2004) Metabotropic glutamate receptor activation regulates fragile x mental retardation protein and FMR1 mRNA localization differentially in dendrites and at synapses. J. Neurosci. 24, 26482655.
  • Barbee S. A., Estes P. S., Cziko A. M. et al. (2006) Staufen- and FMRP-containing neuronal RNPs are structurally and functionally related to somatic P bodies. Neuron 52, 9971009.
  • Bassell G. J., Zhang H., Byrd A. L., Femino A. M., Singer R. H., Taneja K. L., Lifshitz L. M., Herman I. M. and Kosik K. S. (1998) Sorting of beta-actin mRNA and protein to neurites and growth cones in culture. J. Neurosci. 18, 251265.
  • Bhattacharyya S. N. and Filipowicz W. (2007) Argonautes and company: sailing against the wind. Cell 128, 10271028.
  • Bingol B. and Schuman E. M. (2006) Activity-dependent dynamics and sequestration of proteasomes in dendritic spines. Nature 441, 11441148.
  • Broadus J., Fuerstenberg S. and Doe C. Q. (1998) Staufen-dependent localization of Prospero mRNA contributes to neuroblast daughter-cell fate. Nature 391, 792795.
  • Buratti E. and Baralle F. E. (2001b) Characterization and functional implications of the RNA binding properties of nuclear factor TDP-43, a novel splicing regulator of CFTR exon 9. J. Biol. Chem. 276, 3633736343.
  • Buratti E., Dork T., Zuccato E., Pagani F., Romano M. and Baralle F. E. (2001a) Nuclear factor TDP-43 and SR proteins promote in vitro and in vivo CFTR exon 9 skipping. EMBO J. 20, 17741784.
  • Dubnau J., Chiang A. S., Grady L. et al. (2003) The Staufen/Pumilio pathway is involved in Drosophila long-term memory. Curr. Biol. 13, 286296.
  • Dugre-Brisson S., Elvira G., Boulay K., Chatel- Chaix L., Mouland A. J. and DesGroseillers L. (2005) Interaction of Staufen1 with the 5′ end of mRNA facilitates translation of theses RNAs. Nucleic Acids Res. 15, 47974812.
  • Elvira G., Wasiak S., Blandford V. et al. (2006) Characterization of an RNA granule from developing brain. Mol. Cell Proteomics 5, 635651.
  • Eom T., Antar L. N., Singer R. H. and Bassell G. J. (2003) Localization of a beta-actin messenger ribonucleoprotein complex with zipcode-binding protein modulates the density of dendritic filopodia and filopodial synapses. J. Neurosci. 23, 1043310444.
  • Fillman C. and Lykke-Andersen J. (2005) RNA decapping inside and outside of processing bodies. Curr. Opin. Cell Biol. 17, 326331.
  • Goslin K. and Banker G. (1998) Rat hippocampal neurons in low density culture, in Culturing Nerve Cells, Edn. 2 (BankerG. and GosliK., eds), pp. 339371. MIT, Cambridge. MA.
  • Huang F., Chotiner J. K. and Steward O. (2005) The mRNA for elongation factor 1alpha is localized in dendrites and translated in response to treatments that induce long-term depression. J. Neurosci. 25, 71997209.
  • Jin P., Alisch R. S. and Warren S. T. (2004) RNA and microRNAs in fragile X mental retardation. Nat. Cell Biol. 6, 10481053.
  • Kanai Y., Dohmae N. and Hirokawa N. (2004) Kinesin transports RNA: isolation and characterization of an RNA-transporting granule. Neuron 43, 513525.
  • Kiebler M. A. and DesGroseillers L. (2000) Molecular insights into mRNA transport and local translation in the mammalian nervous system. Neuron 25, 1928.
  • Kiebler M. A., Hemraj I., Verkade P., Kohrmann M., Fortes P., Marion R. M., Ortin J. and Dotti C. G. (1999) The mammalian Staufen protein localizes to the somatodendritic domain of cultured hippocampal neurons: implications for its involvement in mRNA transport. J. Neurosci. 19, 288297.
  • Knowles R. B., Sabry J. H., Martone M. E., Deerinck T. J., Ellisman M. H., Bassell G. J. and Kosik K. S. (1996) Translocation of RNA granules in living neurons. J. Neurosci. 16, 78127820.
  • Lein E. S., Hawrylycz M. J., Ao N. et al. (2007) Genome-wide atlas of gene expression in the adult mouse brain. Nature 445, 168176.
  • Li Z., Okamoto K., Hayashi Y. and Sheng M. (2004) The importance of dendritic mitochondria in the morphogensis and plasticity of spines and synapase. Cell 119, 873887.
  • Lin J. C., Hsu M. and Tarn W. Y. (2007) Cell stress modulates the function of splicing regulatory protein RBM4 in translation control. Proc. Natl Acad. Sci. USA 104, 22352240.
  • Liu J., Rivas F. V., Wohlschlegel J., Yates III J. R., Parker R. and Hannon G. J. (2005) A role for the P-body component GW182 in microRNA function. Nat. Cell Biol. 7, 12611266.
  • Mallardo M., Deitinghoff A., Muller J., Goetze B., Macchi P., Peters C. and Kiebler M. A. (2003) Isolation and characterization of Staufen-containing ribonucleoprotein particles from rat brain. Proc. Natl Acad. Sci. USA 100, 21002105.
  • Martin K. C., Barad M. and Kandel E. R. (2000) Local protein synthesis and its role in synapse-specific plasticity. Curr. Opin. Neurobiol. 10, 587592.
  • Miller S., Yasuda M., Coats J. K., Jones Y., Martone M. E. and Mayford M. (2002) Disruption of dendritic translation of CaMKIIalpha impairs stabilization of synaptic plasticity and memory consolidation. Neuron. 36, 507519.
  • Miyashiro K. and Eberwine J. (2004) Fragile X syndrome: (what’s) lost in translation? Proc. Natl Acad. Sci. USA 101, 1732917330.
  • Neumann M., Sampathu D. M., Kwong L. K. et al. (2006) Ubiquitinated TDP-43 in frontotemporal lobar degeneration and amyotrophic lateral sclerosis. Science 314, 130133.
  • Neumann M., Mackenzie I. R., Cairns N. J., Boyer P. J., Markesbery W. R., Smith C. D., Taylor J. P., Kretzschmar H. A., Kimonis V. E. and Forman M. S. (2007) TDP-43 in the ubiquitin pathology of frontotemporal dementia with VCP gene mutations. J. Neuropathol. Exp. Neurol. 66, 152157.
  • Nimchinsky E. A., Oberlander A. M. and Svoboda K. (2001) Abnormal development of dendritic spines in FMR1 knock-out mice. J. Neurosci. 21, 51395146.
  • Ostroff L. E., Fiala J. C., Allwardt B. and Harris K. M. (2002) Polyribosomes redistribute from dendritic shafts into spines with enlarged synapses during LTP in developing rat hippocampal slices. Neuron 35, 535545.
  • Ou S. H., Wu F., Harrich D., Garcia-Martinez L. F. and Gaynor R. B. (1995) Cloning and characterization of a novel cellular protein, TDP-43, that binds to human immunodeficiency virus type 1 TAR DNA sequence motifs. J. Virol. 69, 35843596.
  • Parker R. and Sheth U. (2007) P bodies and the control of mRNA translation and degradation. Mol. Cell. 25, 635646.
  • Rackham O. and Brown C. M. (2004) Visualization of RNA-protein interactions in living cells: FMRP and IMP1 interact on mRNAs. EMBO J. 23, 33463355.
  • Schratt G. M., Nigh E. A., Chen W. G., Hu L. and Greenberg M. E. (2004) BDNF regulates the translation of a select group of mRNAs by a mammalian target of rapamycin-phosphatidylinositol 3-kinase-dependent pathway during neuronal development. J. Neurosci. 24, 73667377.
  • Seelaar H., Jurgen Schelhaas H., Azmani A. et al. (2007) TDP-43 pathology in familial frontotemporal dementia and motor neuron disease without Progranulin mutations. Brain 130, 13751385.
  • Smart F. M., Edelman G. M. and Vanderklish P. W. (2003) BDNF induces translocation of initiation factor 4E to mRNA granules: evidence for a role of synaptic microfilaments and integrins. Proc. Natl Acad. Sci. USA 100, 1440314408.
  • Steward O. and Worley P. F. (2001) A cellular mechanism for targeting newly synthesized mRNAs to synaptic sites on dendrites. Proc. Natl Acad. Sci. USA 98, 70627068.
  • Sutton M. A. and Schuman E. M. (2006) Dendritic protein synthesis, synaptic plasticity, and memory. Cell 127, 4958.
  • Tan C. F., Eguchi H., Tagawa A., Onodera O., Iwasaki T., Tsujino A., Nishizawa M., Kakita A. and Takahashi H. (2007) TDP-43 immunoreactivity in neuronal inclusions in familial amyotrophic lateral sclerosis with or without SOD1 gene mutation. Acta Neuropathol. 113, 535542.
  • Tang S. J., Meulemans D., Vazquez L., Colaco N. and Schuman E. (2001) A role for a rat homolog of Staufen in the transport of RNA to neuronal dendrites. Neuron 32, 463475.
  • Tiruchinapalli D. M., Oleynikov Y., Kelic S., Shenoy S. M., Hartley A., Stanton P. K., Singer R. H. and Bassell G. J. (2003) Activity-dependent trafficking and dynamic localization of zipcode binding protein 1 and beta-actin mRNA in dendrites and spines of hippocampal neurons. J. Neurosci. 23, 32513261.
  • Ule J., Jensen K. B., Ruggiu M., Mele A. and Darnell R. B. (2003) CLIP identifies Nova-regulated RNA networks in the brain. Science 302, 12121215.
  • Wang I. F., Reddy N. M. and Shen C. K. (2002) Higher order arrangement of the eukaryotic nuclear bodies. Proc. Natl Acad. Sci. USA 99, 1358313588.
  • Wang H. Y., Wang I. F., Bose J. and Shen C. K. (2004) Structural diversity and functional implications of the eukaryotic TDP gene family. Genomics 83, 130139.
  • Wang I. F., Chang H. Y. and Shen C. K. (2006) Actin-based modeling of a transcriptionally competent nuclear substructure induced by transcription inhibition. Exp. Cell Res. 312, 37963807.
  • Wu G. Y., Deisseroth K. and Tsien R. W. (2001) Spaced stimuli stabilize MAPK pathway activation and its effects on dendritic morphology. Nat. Neurosci. 4, 151158.
  • Zalfa F., Giorgi M., Primerano B., Moro A., Di Penta A., Reis S., Oostra B. and Bagni C. (2003) The fragile X syndrome protein FMRP associates with BC1 RNA and regulates the translation of specific mRNAs at synapses. Cell 112, 317327.
  • Zhou Q., Homma K. J. and Poo M. M. (2004) Shrinkage of dendritic spines associated with long-term depression of hippocampal synapses. Neuron 44, 749757.