Dynamic interaction between P-bodies and transport ribonucleoprotein particles in dendrites of mature hippocampal neurons

Abstract

The dendritic localization of mRNAs and their subsequent translation at stimulated synapses contributes to the experience-dependent remodeling of synapses and thereby to the establishment of long-term memory. Localized mRNAs are transported in a translationally silent manner to distal dendrites in specific ribonucleoprotein particles (RNPs), termed transport RNPs. A recent study suggested that processing bodies (P-bodies), which have recently been identified as sites of RNA degradation and translational control in eukaryotic cells, may participate in the translational control of dendritically localized mRNAs in Drosophila neurons. This study raised the interesting question of whether dendritic transport RNPs are distinct from P-bodies or whether those structures share significant overlap in their molecular composition in mammalian neurons. Here, we show that P-body and transport RNP markers do not colocalize and are not transported together in the same particles in dendrites of mammalian neurons. Detailed time-lapse videomicroscopy analyses reveal, however, that both P-bodies and transport RNPs can interact in a dynamic manner via docking. Docking is a frequent event involving as much as 50% of all dendritic P-bodies. Chemically induced neuronal activity results in a 60% decrease in the number of P-bodies in dendrites, suggesting that P-bodies disassemble after synaptic stimulation. Our data lend support to the exciting hypothesis that dendritically localized mRNAs might be stored in P-bodies and be released and possibly translated when synapses become activated.

Figure 1.

P-body and transport RNP markers do not colocalize in hippocampal neurons. A, Western blot analysis demonstrating specificity of the antibodies used in this study. All antibodies were tested on rat embryonic brain lysates. Note that the anti-Stau2 antibody recognizes the three different Stau2 isoforms (indicated by molecular weights on the right of the lane) expressed in the mammalian brain. B–K, Cultures of mature hippocampal neurons (18 DIV) were labeled with antibodies against the P-body markers DCP1 (D; in green) or Rck (I; in green) and the transport RNP markers Stau2 (E; in red) or Stau1 (J; in red), respectively. B, G, Phase-contrast pictures of the neurons shown in C–E and H–J, respectively, to demonstrate their integrity; insets in B and G show the corresponding DAPI-stained nuclei of those neurons. In both sets of experiments, P-body markers show virtually no colocalization with transport RNP markers in dendrites. C, H, Merged images of Stau2 and DCP1 signals (magnified distal dendrite in F; C) and of Stau1 and Rck signals (magnified distal dendrite in K; H). Please note the presence of docked particles in F and K (arrowheads and magnified insets 1–3). Scale bars, 10 μm.

Figure 2.

Fluorescently labeled P-bodies and transport RNPs move independently in dendrites. A–F, Mature hippocampal neurons (16 DIV) were cotransfected with expression constructs encoding RFP-DCP1 (in red) and ECFP-ZBP1 (A–C; pseudocolored in green) or RFP-DCP1 (in red) and Stau1-EYFP (D–F; pseudocolored in green), respectively. After 14 h of expression, the respective P-body and transport RNP markers were visualized by time-lapse videomicroscopy. B, E, After live imaging, phase-contrast images were acquired to assess the integrity of the neurons. C, Higher magnification of a dendrite (boxed region in A) in which DCP1 particles (in red; arrows and arrowheads) move independently of the transport RNP marker ZBP1 (pseudocolored in green) in a bidirectional manner. The boxed region in D shows that Stau1 (F, arrowheads) moves independently from DCP1 into distal dendrites. See supplemental Videos 1 and 2 (available at www.jneurosci.org as supplemental material). G, Directionality of DCP1, Stau1, and Stau2 particle movements in dendrites of mature hippocampal neurons. Particles were classified as moving unidirectionally if the direction of their movement did not change throughout the examination period. Scale bars, 10 μm.

Figure 3.

P-bodies and transport RNPs are physically distinct particles that temporarily dock in hippocampal neurons. Mature hippocampal neurons (15 DIV) were cotransfected with expression constructs encoding RFP-DCP1 and EGFP-Stau2. A, After 14 h of expression, both P-body and transport RNP markers were visualized by time-lapse videomicroscopy (supplemental Video 4, available at www.jneurosci.org as supplemental material). Scale bar, 10 μm. B, After imaging, a phase-contrast image was acquired to assess the integrity of the neuron. C–E, Higher magnifications of the boxed regions in A. C, D, Independent movement of P-body and transport RNP markers interrupted by transient docking (>5 min). E, Particles docked constantly over 34 min (see also supplemental Videos 5–7, available at www.jneurosci.org as supplemental material). F, Enlargement from a neuron (supplemental Fig. 6, available at www.jneurosci.org as supplemental material) transfected with Stau1-EYFP (green) and RFP-DCP1 (red), showing particles constantly docked for 17 min (see also supplemental Video 8, available at www.jneurosci.org as supplemental material). G, Analysis of transport RNP markers (Stau1-EYFP and EGFP-Stau2) compared with the P-body marker DCP1 in dendrites. Note that, as observed for endogenous transport RNP and P-body markers (Fig. 1), there is virtually no colocalization observed between Staufen proteins and DCP1. Importantly, however, a significant proportion of transport RNPs and P-bodies are found in close apposition to each other (“docking”). H–K, Videomicroscopy of living neurons by confocal microscopy confirms the dynamic nature of docking between transport RNPs and P-bodies. H, I, Cultures of mature hippocampal neurons (14 DIV) coexpressing RFP-DCP1 (red; H) and Stau1-EYFP (green; I). J, Merged fluorescence images. Several images taken in the region indicated by a box in J are shown at higher magnification in K. The sequence shows the behavior of the particles over a period of 25 min. Note that the P-bodies and transport RNPs temporarily dock over periods of ∼8 and 4 min, respectively. Scale bar, 20 μm.

Figure 4.

Chemical stimulation of mature hippocampal neurons disassembles dendritic P-bodies. A–H, Mature hippocampal neurons (18 DIV) were mock treated (A), stimulated with glutamate (B–E) for 2 min and incubated for the indicated time intervals, stimulated with NMDA (F) for 30 min, silenced overnight with TTX, CNQX, d-APV, and strychnine (G), or exposed to oxidative stress with H₂O₂ (H) for 4 h. After treatment, neurons were stained for the P-body marker DCP1. Insets show the boxed regions in A–H. Whereas silencing and oxidative stress (G, H) do not affect dendritic P-bodies, stimulation of neurons with glutamate (B, C) leads initially to a significant reduction in the number of dendritic P-bodies. D, E, After longer periods of time (>30 min) after the glutamate pulse, the number of dendritic P-bodies recovers to basic levels. Scale bars, 10 μm. Bottom (table), Quantification of dendritic P-body numbers in hippocampal neurons after neuronal stimulation with glutamate, NMDA, or BDNF. For the quantification, five (glutamate) and three (NMDA, BDNF) independent experiments (10 neurons each), respectively, were evaluated in a blind manner. See also Figure 5.

Figure 5.

P-bodies localize at the bases of dendritic spines, and BDNF stimulation disassembles dendritic P-bodies. A, B, Mature (14 DIV) hippocampal neurons were cotransfected with expression vectors encoding RFP-DCP1 (B; red) to label P-bodies and citrine (A; merged image) to visualize the morphology of the neuron. Enlargements of part of a dendrite (boxed regions in A, B) show numerous P-bodies located at the bases of dendritic spines (arrowheads). Thirty percent of total DCP1 particles are located at the base of spines (30 neurons quantified; 25 ± 11.9 total DCP1 particles per cell; 7.5 ± 4.3 Dcp1 particles in spines). Mature hippocampal neurons (15 DIV) were either mock treated (C, E, G) or stimulated with 100 ng/ml BDNF (D, F, H) for 4 h. After treatment, neurons were stained for the P-body markers DCP1, Rck, and Ago2, respectively. Insets show the boxed regions in C–H. Interestingly, BDNF stimulation of neurons led to a significant reduction in the number of dendritic P-bodies as detected with either of the three markers. The insets in B–H show the respective phase-contrast images of the neurons. Please note that to ensure comparability between the fluorescent signals from neurons after stimulation with BDNF and mock treatment, respectively, all images were acquired with the same settings and processed in the identical way. Scale bars, 20 μm.