Address correspondence and reprint requests to Hyong Kyu Kim, Department of Medicine and Microbiology, Medical Research Institute, College of Medicine, Chungbuk National University, Cheongju, 361-763, Korea. E-mail: email@example.com
There is increasing evidence showing that mRNA is transported to the neuronal dendrites in ribonucleoprotein (RNP) complexes or RNA granules, which are aggregates of mRNA, rRNA, ribosomal proteins, and RNA-binding proteins. In these RNP complexes, Staufen, a double-stranded RNA-binding protein, is believed to be a core component that plays a key role in the dendritic mRNA transport. This study investigated the molecular mechanisms of the dendritic mRNA transport using green fluorescent protein-tagged Staufen2 produced employing a Sindbis viral expression system. The kinesin heavy chain was found to be associated with Staufen2. The inhibition of kinesin resulted in a significant decrease in the level of dendritic transport of the Staufen2-containing RNP complexes in neurons under non-stimulating or stimulating conditions. This suggests that the dendritic transport of the Staufen2-containing RNP complexes use kinesin as a motor protein. A mitogen-activated protein kinase inhibitor, PD98059, inhibited the activity-induced increase in the amount of both the Staufen2-containing RNP complexes and Ca2+/calmodulin-dependent protein kinase II α-subunit mRNA in the distal dendrites of cultured hippocampal neurons. Overall, these results suggest that dendritic mRNA transport is mediated via the Staufen2 and kinesin motor proteins and might be modulated by the neuronal activity and mitogen-activated protein kinase pathway.
In mammals, two Staufen genes, Staufen1 (Stau1) and Staufen2 (Stau2) have been isolated (DesGroseillers and Lemieux 1996; Buchner et al. 1999). Each gene is composed of two splice isoforms (63 kDa, Stau163 and 55 kDa, Stau155) for Stau1 (Mallardo et al. 2003) and four splice isoforms (62 kDa, Stau262; 59 kDa, Stau259; 56 kDa, Stau256; and 52 kDa, Stau252) for Stau2 (Duchaîne et al. 2002; Monshausen et al. 2004), respectively. In contrast to Stau1, which has a broad expression pattern, Stau2 has been reported as a brain-specific protein (Duchaîne et al. 2002). The Stau2-containing RNP complexes do not colocalize with the Stau1-containing RNP complexes (Duchaîne et al. 2002), suggesting their different molecular profile or mechanisms in the transport of dendritic mRNA.
In hippocampal neurons, Staufen-containing RNP complexes or RNA granules are transported by microtubule-based movement in dendrites (Köhrmann et al. 1999; Tang et al. 2001), which suggests the involvement of the motor proteins. Consistent with these results, a mammalian Staufen protein was reported to have a microtubule-association domain in the C-terminal region (Wickham et al. 1999; Tang et al. 2001). In Drosophila oocytes, oskar mRNA is transported to the posterior pole through a kinesin motor protein (Brendza et al. 2000). The transport of the RNP complexes, containing Purα, mouse Staufen, the Fragile X mental retardation protein and myosin Va, is associated with the kinesin motor protein in the brain (Ohashi et al. 2002). The inhibition of the kinesin heavy chain (KHC) expression disrupts the transport of testis–brain RBP (translin)-containing RNPs, Ca2+/calmodulin-dependent protein kinase II α-subunit (CAMKIIα), and ligatin mRNAs to the dendrites (Severt et al. 1999). Recently, conventional kinesin superfamily protein 5 (KIF5) was isolated in the RNA granules (Mallardo et al. 2003; Kanai et al. 2004; Villacéet al. 2004). However, the adaptor protein that links the kinesin motor protein with the RNP complexes or RNA granules, or its regulatory mechanisms is not known.
This study examined the molecular mechanisms of dendritic mRNA transport. The results showed that the Staufen2-containing RNP complexes were associated with the KHC and functional inhibition of KHC decreased the amount of the Staufen2-containing RNP complexes in the dendrites, suggesting microtubule-based movement using the kinesin motor protein. Furthermore, the depolarization-induced increase in the Staufen2-containing RNP complexes and CAMKIIα mRNA in the dendrites was impaired by blocking of the mitogen-activated protein kinase (MAP kinase) pathway.
Materials and methods
Dissociated hippocampal neuron cultures were prepared from post-natal 1-day-old rat pups, as described elsewhere (Aakalu et al. 2001). The neurons were plated onto a poly-d-lysine-coated coverslip at a density of 10 000–20 000 cells/cm2 for the low-density culture or onto a poly-d-lysine-coated culture dish for the high-density culture. The cultures were maintained and matured in a growth medium (Neurobasal-A supplemented with B-27 and Glutamax-1; Invitrogen, Carlsbad, CA, USA) for 12 days prior to use.
Sindbis viral expression
The Sindbis virus was used for the exogenous protein expression in the cultured hippocampal neurons because it shows a normal electrophysiological response and long-term potentiation in the hippocampal CA1 region (D’Apuzzo et al. 2001). The following constructs for Sindbis viral expression were prepared according to the method reported by Tang et al. (2002): green fluorescent protein (GFP), GFP-tagged Staufen262 (GFP-Stau), GFP-tagged dominant-negative form of Staufen262-RNA-binding domain (stau-RBD), and myc-tagged Staufen262 (myc-Stau) (Staufen2 cDNA, generous gift from Dr Erin Schuman, Caltech, Pasadena, CA, USA). The Sindbis pseudoviroid was produced according to instructions provided by Invitrogen. For the viral infection, the Sindbis pseudoviroid was added directly to the cultures and incubated in 5% CO2, 37°C for approximately 12 h, depending on the experiments.
Immunoprecipitation and western blotting analysis
The forebrains from 5-week-old male Sprague–Dawley rats were homogenized in an ice-cold lysis buffer [150 mmol/L sodium chloride, 1% IGEPAL® CA-630, and 50 mmol/L Tris Cl (pH 8.0)] in the presence of the protease inhibitors (Roche, Penzberg, Germany). The GFP-Stau or stau-RBD expressing neurons were resuspended in the ice-cold lysis buffer. The protein extracts were separated from the insoluble fractions by centrifugation at 14 000 g and pre-cleared by adding 50 μL of Protein A Sepharose 4 Fast Flow (Amersham Biosciences, Tokyo, Japan). The same amount of the protein extracts was incubated with 3 μg of the anti-KHC monoclonal antibody (clone H2; Chemicon, Temecula, CA, USA) or 3 μg of the mouse IgG (Sigma, St Louis, MO, USA). Subsequently, 50 μL of Protein A Sepharose was added to the extracts. The precipitates were washed three times with 1 mL of a lysis buffer and 1 mL of a wash buffer [50 mmol/L Tris Cl (pH 8.0)]. The pellets of the precipitates were obtained after centrifugation. The pellets were resuspended in 30 μL of a sodium dodecyl sulfate-loading buffer. For the visualization of Stau262 isoform, human embryonic kidney (HEK) 293T cells were transfected with pCMV-Stau262, incubated for 48 h and resuspended in the ice-cold lysis buffer. The proteins were separated on 8% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and analyzed by western blotting enhanced chemiluminescence detection (Amersham Biosciences). The sources of the antibodies used in western blotting are as follows: anti-β-actin antibody (clone AC-74; Sigma) and anti-p44/42 MAP kinase and anti-phospho-p44/42 MAP kinase (Thr202/Tyr204) antibody (Cell Signaling Technology, Beverly, MA, USA). Unless stated otherwise, all reagents were purchased from Sigma-Aldrich (St Louis, MO, USA).
Immunocytochemistry and ethidium bromide staining
For immunostaining, the neurons were fixed with 4%p-formaldehyde and 4% sucrose in ice for 20 min. The membrane was permeabilized by treating the fixed cultures with phosphate-buffered saline (PBS) followed by PBT (1x PBS, 0.1% bovine serum albumin, and 0.1% Triton X-100) for 15 min. The cultures were treated in a pre-block solution (1x PBS, 2% bovine serum albumin, and 0.08% Triton X-100) for 2 h at 25°C. A primary antibody was added to the pre-block solution and incubated overnight at 4°C. After washing with PBT, a secondary antibody was added to the culture and incubated for 2 h at 25°C. This was followed by stringent washing in PBT and PBS. The immunostained cultures were imaged using PBS. For anti-Staufen2 antibody, a glutathione S-transferase (GST)-Staufen2 fusion protein was generated by using the entire Staufen2 open reading frame. Purified recombinant protein was used as the antigen in rabbits. The antiserum from immunized rabbits was affinity purified. The sources of other antibodies are as follows: anti-synaptophysin antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-KHC antibody (clone H2; Chemicon), anti-microtubule-associated protein 2 antibody (Chemicon), anti-c-myc-antibody (clone 9E10; ATCC, Manassas, VA, USA), Alexa Fluor® 594 goat anti-mouse IgG antibody, Alexa Fluor® 594 goat anti-rabbit IgG antibody, and Alexa Fluor® 488 goat anti-rabbit IgG antibody (Molecular Probes, Eugene, OR, USA). After ethidium bromide (EtBr) staining, the image acquisition and analysis were performed using the method reported by Tang et al. (2001). The images of immunocytochemistry and EtBr staining were acquired by confocal microscopy (TCS-SP2 AOBS, Leica Microsystems, Heidelberg, Germany).
Immunoprecipitation and RT-PCR
The myc-Stau expressing neurons were resuspended in a brain ice-cold lysis buffer [10 mmol/L HEPES (pH 7.4), 200 mmol/L NaCl, 30 mmol/L EDTA, and 0.5% Triton X-100] in the presence of the protease inhibitors and 400 U/mL rRNAsin (Promega, Madison, WI, USA). The neuronal extracts were pre-cleared by adding 50 μL of Protein A Sepharose 4 Fast Flow. Twenty percent of the neuronal extracts were used as the control group (CTL), and the remaining 80% of neuronal extracts were immunoprecipitated with 3 μg of the anti-myc antibody or mouse IgG, respectively. The immunoprecipitate was resuspended in diethyl pyrocarbonate-treated water and used for RNA extraction. The first strand cDNA was synthesized using the oligo-dT primer and SuperScriptTM III (Invitrogen) and exposed to 30 cycles of PCR. PCR was performed using Advantage 2 polymerase (BD Biosciences, San Jose, CA, USA) with the sense (5′-TACTTGTTTAGGGTTTTG-3′) and antisense primers (5′-TGTAGCTATTTATTCCAC-3′). The PCR products were then analyzed by 1% agarose gel electrophoresis.
In situ hybridization
The cultured neurons were fixed for 15 min in 4%p-formaldehyde in PBS at 4°C and treated with 0.1 mol/L triethanolamine for 5 min. The neurons were washed with DEPC-treated PBS and subsequently treated with 0.25% acetic anhydride for 10 min. The membrane of the neurons was permeabilized with 0.2 N HCl for 10 min. Pre-hybridization was performed by exposing the neurons to a hybridization solution (50% formamide, 5x SSC, 0.3 mg/mL yeast tRNA, 10 μg/mL heparin, 1x Denhardt’s solution, 0.1% Tween-20, 0.1% CHAPS, and 5 mmol/L EDTA) for 1 h at 25°C. Hybridization was performed using a hybridization solution containing 1–2 μg/mL of digoxygenin (DIG) -labeled antisense or sense CAMKIIα probe overnight at 65°C. The cultures were treated with PBT for 15 min at 25°C after stringent washing using 0.2x SSC at 65°C, incubated in 20% calf serum in PBT at 25°C for 2 h, and further incubated with the alkaline phosphatase-coupled anti-digoxygenin antibody or FITC-conjugated anti-digoxygenin antibody (Roche) overnight at 4°C. After washing with PBT, the signals by the alkaline phosphatase were developed by adding nitroblue tetrazolium chloride/5-bromo-4-chloro-3-indolyl phosphate (Roche) to the cultures at 25°C. The development time of signal was determined empirically in less than 20 h to avoid saturation of the in situ signal. The in situ images were acquired by Nikon inverted microscopy (ECLIPSE TE2000-U, Tokyo, Japan) equipped with a CCD camera (Digital Sight DS-U1, Nikon, Tokyo, Japan) or confocal microscopy.
The digital images were saved on a personal computer in 8-bit grayscale and analyzed using NIH image analysis program (ImageJ program; Bethesda, MD, USA). Neurons with a similar morphology according to phase contrast microscopy and/or fluorescence imaging were selected, and a single main dendrite more than 100 μm in length was scored. In order to remove the signals from soma by simple diffusion, the values of proximal dendritic regions (1–50 μm) were excluded in the data collection. The fluorescent intensity of each dendrite was measured using a 0–255 scale and subtracted from the value of the background image. Each value indicates the mean fluorescent intensity of the indicated dendritic regions in arbitrary units (AU). The relative fluorescent intensity was obtained by normalizing the mean to that of the control group. A Student’s t-test was employed to determine the statistical difference between the groups. All image acquisitions and data analysis were performed in blind experiment.
Cellular knockdown and functional blockade of kinesin
The function of the KHC was inhibited by a set of synthetic oligonucleotides (ODNs) that were specific to the KHC sequences. They were added to the cultures for 3 days before infecting them with the Sindbis virus, initially with a final concentration of 50 μmol/L, then 25 μmol/L for 12 h each (Ferreira et al. 1992; Severt et al. 1999). The following ODNs sequence was used: antisense, 5′-CCGGGTCCGCCATCTTTCTGCCAG-3′ and sense, 5′-TGCCAGAAAGATGGCGGACCCGG-3′. For the control western blotting, the protein extracts from ODN-treated neurons were analyzed by western blotting using the anti-KHC antibody and anti-β-actin antibody. The Sindbis viral infection and expression were analyzed in media containing 7.5 μmol/L of the ODNs in parallel with the inhibitor treatment experiments. For the kinesin inhibitor treatment in the non-stimulating condition, the cultured hippocampal neurons were infected with the Sindbis virus encoding GFP-Stau and incubated for 18 h. The neurons were then treated with either 1 mmol/L 5′-adenylyl imidodiphosphate (AMP-PNP; Sigma) or 5 μmol/L vanadate (Sigma) for 6 h, and fixed immediately. In the stimulating condition, the neurons were infected with the Sindbis virus encoding GFP-Stau for 9 h, and treated with each inhibitor for 3 h prior to KCl stimulation. For KCl stimulation, the growth medium of the culture was exchanged with a HEPES-buffered saline (HBS) containing 60 mmol/L KCl (high-K+ HBS) for 10 min (Murase et al. 2002). Subsequently, high-K+ HBS was immediately exchanged with normal HBS (5.4 mmol/L KCl). The treated neurons were further incubated in 5% CO2, at 37°C for 2 h 50 min, making a total experimental time of 3 h.
Kinase inhibitor treatment
The cultured neurons were infected with the GFP-Stau virus and incubated for 11 h 30 min. The cultured neurons was then pre-treated with a media containing 30 μmol/L PD98059 (Tocris, Avonmouth, UK) or 2 mmol/L EGTA (Sigma) for 30 min. The culture media was exchanged with either high-K+ HBS in the presence or absence of each inhibitor, or normal HBS as the control group, and incubated for 10 min. The neurons were fixed and imaged after the additional incubation period of 2 h 50 min.
Dendritic localizations of Staufen2-containing granules
The introduction and over-expression of GFP-tagged Staufen2 (GFP-Stau) in the neurons resulted in the localization of the Staufen protein to the RNP complexes or RNA granules (Köhrmann et al. 1999). The role of Staufen2 in dendritic mRNA transport was examined by expressing GFP-Stau in the cultured hippocampal neurons using a Sindbis viral expression system (Tang et al. 2001). Thus, we took advantage of the Staufen2-containing granules produced by the Sindbis viral over-expression to determine the molecular mechanisms of dendritic mRNA transport. Initially, whether the Staufen2 antibody can detect the Staufen2-containing granules in the neurons infected with the Sindbis virus encoding GFP-Stau was examined. Immunocytochemistry against Staufen2 showed that Staufen2 is more clustered and highly expressed in GFP-Stau virus-infected neurons than the control uninfected neurons. As shown in Fig. 1a, Staufen2 was detected in the GFP-tagged granules in both the soma and dendrites. The signals for GFP were strongly colocalized with those for Staufen2 (Fig. 1a), suggesting that the GFP signals as a result of the Sindbis viral expression are good indicators for the Staufen2-containing granules. The following experiments were carried out to determine if the over-expressed Staufen2-containing granules contain RNA. In consistent with previous reports of Tang et al. (2001), the GFP signals strongly colocalized with the EtBr signals (Fig. 1a). Furthermore, the GFP signals were detected in the nuclear and nucleolus regions in the early stages of expression, highlighting the role of Staufen2 as a nucleocytoplasmic shuttling protein (Miki et al. 2005).
CAMKIIα mRNA in Staufen2-containing granules
Further analysis was carried out to determine if the over-expressed Staufen2-containing granules contain mRNA. RT-PCR analysis after immunoprecipitation (IP) showed that the over-expressed Staufen2-containing granules contained CAMKIIα mRNA (Fig. 1b). The next study examined whether the Staufen2-containing granules in the neurons infected with the Sindbis virus encoding myc-tagged Staufen (myc-Stau) also contained CAMKIIα mRNA. The level of dendritic CAMKIIα mRNA was examined by in situ hybridization. As shown in Fig. 1a, the immunoreactivity to myc is colocalized with the in situ signals. Overall, these results from two independent experiments suggest that the over-expressed Staufen2-containing granules have similar properties to the endogenous Staufen2-containing RNP complexes or RNA granules.
Association of Staufen2 with kinesin motor protein
A co-IP assay was performed to determine the molecular interaction between Staufen2 and the kinesin motor protein. Initially, the KHC was immunoprecipitated from the forebrain extracts. Subsequently, western blotting using the Staufen2 antibody was carried out to visualize the forebrain extracts and the lysates of HEK 293T cells expressing Stau262 as well as the immunoprecipitated proteins. As shown in Fig. 2a, Stau259 isoform, which has a long C-terminal region, showed an interaction with the KHC. The anti-Staufen2 antibody detected an additional band between 56 and 59 kDa in the lysates of HEK 293T cells expressing Stau262, implying some cross reactivity to other proteins. Another Staufen2 isoform, which has the long C-terminal region, Stau262 was not detected in either the immunoprecipitates or input lane. This is probably due to the relative low expression level of Stau262 isoform compared with the others and consistent with previous reports (Duchaîne et al. 2002; Monshausen et al. 2004).
To further analyze the molecular interaction of Staufen2 with the KHC, the cultured hippocampal neurons were infected with the Sindbis virus encoding GFP-tagged Staufen262 (GFP-Stau, ∼90 kDa) or GFP-tagged RBDs of Staufen262 (stau-RBD, ∼70 kDa), which has RBDs but no C-terminal microtubule-association domain (Fig. 2b). The co-IP assay results also provided an evidence for an interaction between Stau262 and the kinesin motor proteins in the cultured hippocampal neurons. This interaction appears to be mediated via the long C-terminal microtubule-association domain of Staufen2 (Fig. 2b). In addition, the interaction between endogenous Staufen2 and the kinesin motor protein in infected neurons was observed as faint bands or not observed in the co-IP assay using the lysates of cultured neurons. It might be due to the low expression level of Staufen2 in the cultured neurons or the limited sensitivity of co-IP assay. An association from both endogenous and infected neurons in vivo was observed, but in an in vitro binding assays. For example, GST-pull down assay did not detect a direct interaction between Staufen2 and the KHC, suggesting a possibility of indirect association. Overall, these results suggest that Staufen2 isoform, Stau262 or Stau259, which has the long C-terminal region, was associated with the KHC.
Staufen2-containing RNP complexes transport by kinesin motor protein
It was previously reported that dendritic mRNA transport exploits microtubule-based movement (Tang et al. 2001). The antisense oligonucleotides (ODN-A) to the KHC inhibit the function of kinesin in cultured hippocampal neurons efficiently (Ferreira et al. 1992; Severt et al. 1999; Guillaud et al. 2003). Therefore, kinesin was inhibited using ODN-A to determine if a molecular motor protein is involved in the transporting the Staufen2-containing RNP complexes. In this experiment, a treatment with the ODN-A reduced the kinesin expression level to approximately 40% of the control (Fig. 3a). As shown in Fig. 3b, c, and d, the level of the Staufen2-containing RNP complexes in the distal dendrites was decreased significantly after exposure to the ODN-A, but not to the sense oligonucleotides (ODN-S) (51–100 μm; Control: 31.58 ± 1.40, n = 18; ODN-A: 21.56 ± 1.51, n = 23; ODN-S: 28.43 ± 2.70, n = 16; RBD: 16.08 ± 0.94, n = 11; AU). When the background value by simple diffusion (the value of the stau-RBD expression neurons) was subtracted, the treatment with the ODN-A reduced the level of the Staufen2-containing RNP complexes to approximately 35% of the control, showing a consistent result with the previous control experiment (Fig. 3a). In addition, the amount of Staufen2-containing RNP complexes in the distal dendrites was decreased to approximately 54% of the control after 6 h exposure to AMP-PNP, which is a non-hydrolysable analog of ATP (Fig. 3d, AMP-PNP: 24.46 ± 2.39, n = 16; AU) (Hurst et al. 1999). Interestingly, a treatment with vanadate, which inhibits the function of dynein, also decreased the level of the Staufen2-containing RNP complexes to 87.59 ± 5.59 (%, n = 21) of that observed in the control, but this was not statistically significant. Depolarization increases the rate of transport of RNA granules and ribosomes in the dendrites (Knowles et al. 1996; Kim et al. 2005). Therefore, the role of kinesin in the dendritic transport of the Staufen2-containing RNP complexes in the depolarized condition was further examined. Experiments were carried out to determine if the KCl-induced increase in dendritic transport is affected by AMP-PNP, a kinesin inhibitor. The KCl-induced increase of the amount of Staufen2-containing RNP complexes in the distal dendrites was blocked by a treatment with AMP-PNP (Fig. 3e, 51–100 μm; Control: 100.0 ± 9.75%, n = 26; +KCl: 147.0 ± 13.11%, n = 22; AMP-PNP: 92.94 ± 8.90%, n = 29; vanadate: 155.6 ± 17.36%, n = 17).
Colocalization Staufen2-containing granules and kinesin motor protein
Immunostaining of the cultured hippocampal neurons was carried out to determine the localization of kinesin and Staufen2. As expected, staining for kinesin was evident in both the cell body and dendritic regions (Fig. 4). The kinesin signal was colocalized with the Staufen2-containing RNP complexes produced by the Sindbis viral expression. Furthermore, this colocalization was evidently observed in the immunostainings for both microtubule-associated protein 2 and the endogenous Staufen2. However, the immunoreactivity of synaptophysin, which is a pre-synaptic marker, partially overlapped with the Staufen2-containing RNP complexes, suggesting that these RNP complexes are localized near the synaptic regions of dendrites (Fig. 4). Overall, the dendritic transport of the Staufen2-containing RNP complexes employs a microtubule-based movement by using kinesin as the motor protein.
Modulation of Staufen2-containing RNP complexes transport by MAP kinase pathway
The aim of next series of experiments was to determine how the kinase signaling pathways modulate the transport of the Staufen2-containing RNP complexes. The cultured hippocampal neurons were infected with the GFP-Stau virus. The neurons were then pre-treated with either a kinase inhibitor or EGTA, which was followed by KCl stimulation. Image analysis was performed to determine the transport of the Staufen2-containing RNP complexes. As shown in Fig. 5a, a treatment with 30 μmol/L PD98059, a MAP kinase inhibitor, completely blocked the depolarization-induced increase in the dendritic Staufen2-containing RNP complexes (Control: 100.00 ± 9.75%, n = 26; +KCl: 147.13 ± 13.11%, n = 22; PD98059: 92.94 ± 8.90%, n = 29; vanadate: 155.60 ± 17.36%, n = 17). The amount of Staufen2-containing RNP complexes were not affected by a treatment with Rp-cAMP, a protein kinase A inhibitor, or chelerythrine, a protein kinase C inhibitor (data not shown). The MAP kinase inhibitor was further examined to determine if its action is exerted at the level of protein synthesis. It was found that the GFP-Stau level in our experiments was unaffected by the PD98059 treatment (Fig. 5b). In addition, brief KCl stimulation increased the level of extracellular-signal regulated kinase 1/2 (ERK 1/2) phosphorylation (Fig. 5c).
The neuronal activity induced by depolarization increases the levels of a specific type of mRNA in the dendrites (Tongiorgi et al. 1997; Righi et al. 2000; Tiruchinapalli et al. 2003). CAMKIIα mRNA is found in the neuronal dendrites (Burgin et al. 1990; Benson et al. 1992). In consistent with these results, CAMKIIα mRNA was also found in the Staufen2-containing RNP complexes (Fig. 1). Therefore, the level of dendritic CAMKIIα mRNA in a blockade of the MAP kinase pathway was examined by in situ hybridization. The PD98059 treatment also blocked the depolarization-induced increase in CAMKIIα mRNA in the dendrites (Fig. 6, Control: 100.00 ± 8.47%, n = 17; +KCl: 164.10 ± 11.00%, n = 22; EGTA: 96.55 ± 9.00%, n = 22; PD98059: 99.00 ± 5.56%, n = 20). This suggests that the dendritic mRNA transport is modulated by the MAP kinase pathway.
In Drosophila oocytes, conventional kinesin is essential for the posterior localization of oskar mRNA and Staufen (Brendza et al. 2000). Staufen-containing RNP complexes are transported through microtubule-based movement in neurons (Köhrmann et al. 1999; Tang et al. 2001). In agreement with previous studies, components of the cytoskeleton as well as motor proteins including kinesin, dynein, or actin are found in the Staufen-containing RNP complexes (Ohashi et al. 2002; Mallardo et al. 2003; Kanai et al. 2004; Villacéet al. 2004). The kinesin motor protein has been suggested to be a factor in dendritic mRNA transport based largely on the pharmacological intervention of microtubules (Köhrmann et al. 1999; Tang et al. 2001). In this study, the kinesin motor protein was associated with the Staufen2-containing RNP complexes and this interaction was mediated via the long C-terminal microtubule-association domains of Staufen2. The function of the kinesin motor protein was inhibited using ODNs or an inhibitor. Either method decreased the level of the Staufen2-containing RNP complexes in the distal dendrites. Furthermore, the inhibitor treatment completely abolished the increase induced by depolarization. These results provide direct evidence for the involvement of the kinesin motor protein in dendritic mRNA transport.
Several small interfering RNAs of the KHC were tested, but none affected the expression level of the KHC. In this study, antisense approaches were performed to examine the loss of function. Even a treatment with the ODN-A, which is rather outdated or non-specific compared with RNA interference (RNAi), effectively reduced the expression level in the cultured hippocampal neurons. These results are comparable with those of previous studies (Ferreira et al. 1992; Severt et al. 1999; Guillaud et al. 2003).
Conventional kinesin consists of two KHC and two kinesin light chains (Hirokawa et al. 1989). KIF5A was first identified as a brain-specific KHC (Aizawa et al. 1992). KIF5C was also isolated in a mouse brain, and is particularly abundant in motor neurons (Kanai et al. 2000). Recent evidence suggests that KIF5 mediates the transport of the RNP complexes in neurons (Kanai et al. 2004). Kanai et al. isolated RNA granules by a GST-pull down assay using the C-terminal domain of KIF5 as bait. They suggested there was an interaction between Staufen1 and the KHC. In this study, the Staufen2-containing RNP complexes were isolated along with the KHC from the forebrain extracts. Interestingly, Stau259 and/or Stau262-containing RNP complexes were associated with the kinesin motor protein. It will be intriguing to determine whether such RNP complexes contain one or both isoforms. Although there was no direct interaction between Staufen2 and the KHC detected, the results showed that Staufen2 plays a key role in mRNA motile structure and a motor protein. Further study will be needed to identity the linker between Staufen2 and the KHC. The specific isoform of the KHC associated with the Staufen2-containing RNP complexes could not be determined because all KIF5 isoforms are detected by the monoclonal antibody (clone H2) (Pfister et al. 1989). Considering their different expression patterns, it is possible that each isoform plays a different role in different types of neurons. However, additional study will be needed to confirm this.
The main focus of this study was on the kinesin motor protein. However, the role of dynein, which lacks the minus end-directed microtubule motor protein, in dendritic mRNA transport should not be overlooked. Dynein has also been isolated from the RNA granules (Villacéet al. 2004). The back and forth movement of the RNA granules in neurons (Kanai et al. 2004) or oligodendrocytes (Carson et al. 1997) suggests that dynein plays a role in dendritic mRNA transport. In such a context, dynein might be a control protein that adjusts the RNA granules to their locales, rather than increase or decrease the amount of RNA granules.
The MAP kinase pathway is an important mechanism in the synaptic plasticity (Sweatt 2004). The ERK1 and ERK2, which are a member of the MAP kinase group, were detected using immunolabeling methods in the dendrites, and its activation is involved in dendritic morphogenesis (Wu et al. 2001; Goldin and Segal 2003). These results showed that the MAP kinase pathway modulates the increase in the Staufen2-containing RNP complexes induced by the neuronal activity. It is possible that the MAP kinase pathway modulates the association between the kinesin motor protein and the Staufen2-containing RNP complexes or RNA granules through the phosphorylation of Staufen2 or kinesin motor protein. Previous studies have reported that the MAP kinase pathway and kinesin are essential for plant cytokinesis (Nishihama et al. 2002). The phosphorylation of kinesin increases the level of cargo binding and neurite outgrowth in chick neurons (Lee and Hollenbeck 1995). The Sunday driver protein (JIP3/JSAP1), which is a linker protein for the kinesin-vesicular cargos such as organelles or small vesicles, is a scaffold protein of c-jun N-terminal kinase, a member of the MAP kinase group (Kelkar et al. 2000). The other possibilities are that the activation of the MAP kinase pathway can affect motor activation by changing either the speed of transport of the Staufen2-containing RNP complexes or the net translocation by affecting kinesin, which is a plus-end movement motor protein, or dynein, which is minus-end movement motor protein. However, additional study will be needed to confirm this.
These results in this study indicate a new mechanism for dendritic mRNA transport in the cultured hippocampal neurons, which is mediated by Staufen2 and the kinesin motor protein and might be modulated by the MAP kinase pathway. In addition, these results also suggest a regulatory mechanism for dendritic protein synthesis in the long-term phase of the synaptic plasticity and memory.
This work was supported by the Korea Research Foundation Grant funded by the Korean Government (MOERD) (R05-2004-000-10432-0).