Differential regulation of local mRNA dynamics and translation following long-term potentiation and depression

Abstract

Decades of work have demonstrated that messenger RNAs (mRNAs) are localized and translated within neuronal dendrites and axons to provide proteins for remodeling and maintaining growth cones or synapses. It remains unknown, however, whether specific forms of plasticity differentially regulate the dynamics and translation of individual mRNA species. To address this, we targeted three individual synaptically localized mRNAs, CamkIIa, β-actin, Psd95, and used molecular beacons to track endogenous mRNA movements. We used reporters and CRISPR/Cas9 gene editing to track mRNA translation in cultured neurons. We found alterations in mRNA dynamic properties occurred during two forms of synaptic plasticity, long-term potentiation (cLTP) and depression (mGluR-LTD). Changes in mRNA dynamics following either form of plasticity resulted in an enrichment of mRNA in the vicinity of dendritic spines. Both the reporters and tagging of endogenous proteins revealed the transcript-specific stimulation of protein synthesis following cLTP or mGluR-LTD. As such, the plasticity-induced enrichment of mRNA near synapses could be uncoupled from its translational status. The enrichment of mRNA in the proximity of spines allows for localized signaling pathways to decode plasticity milieus and stimulate a specific translational profile, resulting in a customized remodeling of the synaptic proteome.

Keywords: local protein synthesis; mRNA; mRNA beacons; mRNA localization; neuron protein synthesis.

Fig. 1.

Dendritic mRNAs exist within varied copy number states. (A) Scheme of the molecular beacon design used in this study. A 5-nt stem, with a reporter fluorophore on one side and a quencher dye on the other side, were linked with a 26-nt complementary sequence to Camk2a, β-actin, or Psd95 mRNAs. When not hybridized to a target mRNA, the stem holds the reporter and quencher in proximity, preventing fluorescence. Upon binding to a target mRNA, the stem loop opens, and fluorescence can be detected in a reversible manner. (B) Example images of Gatta quant standard and beacon-labeled neurons shows a heterogenous distribution of particle intensities in live neurons. (Scale bar, 5 μm.) Data shown acquired at 5% laser power. (C) Quantification of the distribution of beacon intensity relative to single fluor labeled ATTO647 standard. Atto647N Camk2a (19 cells, 921 mRNA granules), Atto647N β-actin (21 cells, 1,003 mRNA granules), Atto647N Psd95 (18 cells, 829 mRNA granules). (D) Example smFISH images of dendritically localized β-actin, Camk2a, and Psd95. (Scale bar, 5 μm.) (E) Quantification of the distribution of smFISH intensity. Camk2a (15 cells, 1,005 mRNA granules), β-actin (15 cells, 1,005 mRNA granules), Psd95 (15 cells, 1,005 mRNA granules).

Fig. 2.

Tracking and classifying mRNA dynamics along dendrites in live neurons. (A) Still images from a β-actin Atto647N-labeled dendrite (Movie S6), shown are single frames every minute for 10 min (of 20 total). Individual mRNA puncta are highlighted to illustrate distinct dynamic profiles, mainly stationary (green) or confined with periods of motility (orange and magenta). (Scale bar, 5 μm.) (B) Kymograph from the first 10 min of (Movie S6). Arrow denotes the anterograde direction along the dendrite. (Scale bar, 1 min.) (C) Quantification of mRNA dynamic state: confined, anterograde vs. retrograde for Atto565 Camk2a, Atto565 β-actin, and Atto565 Psd95. n = 59 cells. **P < 0.01, ****P < 0.0001 Holm–Sidak’s multiple comparison test. (D) Quantification of cumulative distance traveled for Atto565 Camk2a, Atto565 β-actin, and Atto565 Psd95. n = 59 cells. (E) Quantification of mRNA velocity for anterograde and retrograde for Atto565 Camk2a (1.013 ± 0.450; −1.083 ± 0.480), Atto565 β-actin (1.054 ± 0.493; −1.110 ± 0.452), and Atto565 Psd95 (0.997 ± 0.426; −1.111 ± 0.479). n = 59 cells.

Fig. 3.

Manipulating ribosome association of mRNAs results in transcript-specific alterations in mRNA dynamics. (A) Schematic representation of the effects of puromycin (Upper) or anisomycin (Lower) on ribosomal association with mRNA. Puromycin results in ribosomal subunit disassembly, whereas anisomycin results in stalling of elongating ribosomes. (B) Quantification of mRNA dynamic state: confined, anterograde vs. retrograde for Atto647N Camk2a, Atto647N β-actin, and Atto647N Psd95 for control vs. puro treated samples. n = 29 cells per condition. *P < 0.05; ***P < 0.001. Paired t test. (C) Quantification of cumulative distance traveled for Atto647N Camk2a, Atto647N β-actin, and Atto647N Psd95 for control vs. puro-treated samples. n = 29 cells per condition. **P < 0.01 Sidak’s multiple comparisons test. (D) Quantification of mRNA dynamic state: confined, anterograde vs. retrograde for Atto647N Camk2a, Atto647N β-actin, and Atto647N Psd95 for control vs. anisomycin-treated samples. n = 15 cells per condition. *P < 0.05; **P < 0.01 paired t test. (E) Quantification of cumulative distance traveled for Atto647N Camk2a, Atto647N β-actin, and Atto647N Psd95 for control vs. anisomycin-treated samples. n = 15 cells per condition. **P < 0.01; ***P < 0.001 Sidak’s multiple comparisons test.

Fig. 4.

Two distinct forms of synaptic plasticity attenuate mRNA dynamics near dendritic spines. (A) Quantification of mRNA dynamic state: confined, anterograde vs. retrograde for Atto647N Camk2a, Atto647N β-actin, and Atto647N Psd95 for control vs. cLTP induced samples. n = 14 cells per condition. *P < 0.05; **P < 0.01; ***P < 0.001. Paired t test. (B) Quantification of cumulative distance traveled for Atto647N Camk2a, Atto647N β-actin, and Atto647N Psd95 for control vs. cLTP induced samples. n = 14 cells per condition. *P < 0.05; **P < 0.01. Sidak’s multiple comparisons test. (C) Quantification of mRNA dynamic state: confined, anterograde vs. retrograde for Atto647N Camk2a, Atto647N β-actin, and Atto647N Psd95 for control vs. mGluR-LTD induced samples. n = 14 cells per condition. *P < 0.05; **P < 0.01; ***P < 0.001. Paired t test. (D) Quantification of cumulative distance traveled for Atto647N Camk2a, Atto647N β-actin, and Atto647N Psd95 for control vs. mGluR-LTD–induced samples. n = 14 cells per condition. ; *P < 0.05; **P < 0.01. Sidak’s multiple comparisons test. (E) Example images of Camk2a RNA FISH signal in neurons volume filled with mCherry under control, +cLTP and +mGluR-LTD conditions. (Scale bar, 2.5 μm.) (F) Quantification of mRNA distance to the nearest spine reveals a slight decrease in the distance for all three mRNAs during plasticity. Control, cLTP, mGluR-LTD ± SD: Camk2a (411.6 ± 300.7; 376.8 ± 292.5; 375.0 ± 275.4), β-actin (474.3 ± 327.1; 413.8 ± 312.3; 391.0 ± 275.8), and Psd95 (537.2 ± 334.3; 403.3 ± 283.7; 454.6 ± 304.7) *P < 0.05; **P < 0.01; ****P < 0.0001; n.s., not significant. Dunn’s multiple comparisons test. n = 15 cells per condition.

Fig. 5.

Real-time visualization of 3′UTR regulated translation during synaptic potentiation and depression. (A) Scheme for the reporters used to assess real-time 3′UTR regulated translational dynamics in live neurons. (B) Scheme of workflow for the treatment and visualization of plasticity regulated protein synthesis. Following induction of cLTP or mGluR-LTD, the indicated pharmacological treatment was washed out (black box) and the samples were then imaged for 2 min every 15 s to acquire a baseline measurement. The cells were then bleached and the fluorescence recovery was monitored every 15 s for 10 min. For controls, either no treatment or treatment with the translational inhibitor anisomycin was used for comparison. (C) Example image of a Camk2a sfGFP reporter under control (top row) +anisomycin (second row), +cLTP (third row), or +mGluR-LTD conditions before bleaching and during the phase of fluorescence recovery. (Scale bar, 5 μm.) (D) Mobile fractions calculated for the translational reporters during control, plasticity induction and anisomycin treatment. *P < 0.05; **P < 0.01; ***P < 0.001. Dunnett’s multiple comparisons test for treated vs. control condition for each construct; n = >14 cells per condition. (E) Recovery rate of fluorescence of dendritic spines to shafts demonstrates a bias for spine fluorescence recovery during plasticity. Kruskal-Wallis test. **P < 0.01; ****P < 0.0001. n = >14 cells per condition.

Fig. 6.

Visualizing endogenous protein translation in real-time during synaptic potentiation and depression. (A) Scheme for the modification of endogenous gene tagging of Camk2a/β-actin/Psd95 with Venus fluorescent protein. T.S.; targeting sequence; arrow indicates start codon. (B) Scheme of workflow for the treatment and visualization of plasticity regulated protein synthesis. Following induction of cLTP or mGluR-LTD, the pharmacological treatment was washed out (black box) and the samples were then imaged for 2 min every 15 s to acquire a baseline measurement. The cells were then bleached and the fluorescence recovery was then monitored every 15 s for 60 min. For controls, either no treatment or treatment with the protein synthesis inhibitor anisomycin was used for comparison. (C) Mobile population during the FRAP recovery for Venus-CAMK2a, Venus–β-ACTIN, and PSD-95–Venus during control and anisomycin treatment. n = 10 cells per condition. Two-tailed paired t test. **P < 0.01; ***P < 0.001; ****P < 0.0001. (D) Mobile population during the FRAP recovery for Venus-CAMK2a, Venus–β-ACTIN, and PSD-95–Venus during control and cLTP. n = 10 cells per condition. Two-tailed paired t test. *P < 0.05; **P < 0.01. (E) Mobile population during the FRAP recovery for Venus-CAMK2a, Venus–β-ACTIN, and PSD–95-Venus during control and mGluR-LTD. n = 9 to 10 cells per condition. Two-tailed paired t test. *P < 0.05. (F) Example images dendritically localized (MAP2, magenta) Puro-PLA signal for PSD-95 (white) under control and stimulated conditions. (Scale bar, 10 μm.) (G) Puro-PLA quantification reveals local protein synthesis underlies the translational responses of CAMK2a, β-ACTIN, and PSD-95 during cLTP and mGluR-LTD. n = 85 cells per condition. Dunnett’s multiple comparisons test. **P < 0.01; ***P < 0.001; ****P < 0.0001.