Cap-dependent translation initiation monitored in living cells

Abstract

mRNA translation is tightly regulated to preserve cellular homeostasis. Despite extensive biochemical, genetic, and structural studies, a detailed understanding of mRNA translation regulation is lacking. Imaging methodologies able to resolve the binding dynamics of translation factors at single-cell and single-mRNA resolution were necessary to fully elucidate regulation of this paramount process. Here live-cell spectroscopy and single-particle tracking were combined to interrogate the binding dynamics of endogenous initiation factors to the 5'cap. The diffusion of initiation factors (IFs) changed markedly upon their association with mRNA. Quantifying their diffusion characteristics revealed the sequence of IFs assembly and disassembly in cell lines and the clustering of translation in neurons. This approach revealed translation regulation at high spatial and temporal resolution that can be applied to the formation of any endogenous complex that results in a measurable shift in diffusion.

Fig. 1. Binding of exogenous Halo-eIF4E to the 5’cap is detected by FCS.

a NIH3T3 cells that express Halo-eIF4E, in which the endogenous counterpart was silenced by shRNA, were treated for 2 h with vehicle (DMSO) or 250 nM torin-1. Averaged autocorrelation curves show temporal diffusion of Halo-eIF4E molecules (ms = milliseconds) in the cytoplasm (black) and in the nucleus (red) in the indicated conditions (N = 10 ± SEM). Halo-eIF4E diffusion is slower in translating cells as compared to cells treated with torin-1. Cytoplasmic autocorrelation best fit with two components (fast – dark yellow dotted curve: τ_fast = 1.14 ± 0.03 ms, and slow – dash-dotted curve: τ_slow = 372.69 ± 10.8 ms). Percentages of slow (D_slow = 0.05 μm²/s) and fast (D_fast = 14.78 μm²/s) moving molecules are indicated. While one-component fits nuclear Halo-eIF4E well (see red curves), a one-component fit (see the blue dashed curve, with dotted fit residual) cannot adequately describe the cytoplasmic Halo-eIF4E autocorrelation curve. b Total cell lysates from the cells described in (a) were analyzed by western blotting with the indicated antibodies. 4E-BPs phosphorylation and CyclinD1 expression were significantly reduced after 2 h torin-1. c, d SNAP_f-eIF4E and SNAP_f-eIF4E W56A (SNAP_f-W56A) were expressed in the cells described above. Total cell lysates were subjected to a cap-pull-down assay. Levels of Halo-eIF4E, SNAP_f-eIF4E, and SNAP_f-eIF4E W56A were detected by western blotting in input (5%) and cap-bound fractions (c) using eIF4E antibody (eIF4E). Averaged autocorrelation curves show temporal diffusion of SNAP_f-eIF4E W56A in the cytoplasm (red) and in the nucleus (black) (N = 10 ± SEM). Only one fast component was detected in both cellular compartments (d). e, f NIH3T3 cells that express both Halo-eIF4E and SNAP_f-4E-BP1 were treated with vehicle (DMSO) or 250 nM torin-1 for 2 h. Total cell lysates were subjected to cap-pull down assay and analyzed by western blotting in the indicated fractions Overexpression of SNAP_f-4E-BP1 is sufficient to increase its binding to eIF4E on the 5’cap (e). Averaged autocorrelation curves show temporal diffusion of Halo-eIF4E in the indicated conditions (N = 10 ± SEM). SNAP_f-4E-BP1 expression is sufficient to displace most of the Halo-eIF4E bound to the 5’cap (f).

Fig. 2. eIF4E is released from the 5’cap upon binding to 4E-BP1.

a Differentiated parental and mESC in which Halo and SNAP_f tags were inserted into the EIF4E and 4EBP1 locus, respectively (Halo-4E^+/+/SNAP-BP1^+/+), were treated with vehicle (DMSO) or 250 nM torin-1 for 1 h and 30 min. Total cell lysates (Input) were subjected to cap-pull-down assay and analyzed by western blotting using the indicated antibodies. The Halo and SNAP_f tags do not affect eIF4E:4E-BP1 binding upon mTOR inhibition. b mESC double knock-in described in (a) were treated with vehicle (DMSO) or 250 nM torin-1. Simultaneous diffusion of _JF585Halo-eIF4E and _JF646SNAP_f-4E-BP1 was analyzed by dual color cross-correlation spectroscopy in the indicated conditions. Cross-correlation was detected, in the cytoplasm, 30 to 1 h 20 min upon mTOR inhibition and with differential diffusion speed. The vehicle showed no correlation over time (gray) (N = 10 ± SEM). c mESC double knock-in described in (a) was treated with vehicle (DMSO) or 250 nM torin-1 for 30 or 90 min. Total cell lysates were analyzed by western blotting using the indicated antibodies. 4E-BP1 and rpS6 were used as loading controls. d–f Individual diffusion of _JF585Halo-eIF4E and _JF646SNAP_f-4E-BP1 was analyzed by FCS in control cells (vehicle) (d) or in torin-1 treated cells (e, f). Averaged autocorrelation curves show two Halo-eIF4E components (fast and slow) and one-fast SNAP_f-4E-BP1 component in the cytoplasm of translating cells. The nuclear diffusion of SNAP_f-4E-BP1 is depicted in blue (d). Upon 30–43 min torin-1 treatment, SNAP_f-4E-BP1 diffusion slows down in the cytoplasm with Halo-eIF4E still moving slower than its nuclear counterpart (e). After 54–90 min torin-1 treatment, both Halo-eIF4E and SNAP_f-4E-BP1 autocorrelations show overall fast diffusion (f) (N = 10 ± SEM).

Fig. 3. eIF4E and eIF4G binding to the mRNA is detected by FCS upon mTOR inhibition.

a Parental mESC and double knock-in Halo-eIF4E^+/+/SNAP-eIF4G^+/+ total cell lysates (input) were subjected to cap-pull-down assay (m⁷GTP-pull-down) and analyzed by western blotting. eIF4G and eIF4E antibodies showed binding of wild-type and tagged proteins to the 5’cap. Tagging of endogenous eIF4E and eIF4G did not affect cap-binding as compared to the parental counterpart. b–e Halo-eF4E^+/+/SNAP-eIF4G^+/+ cells were treated with vehicle (DMSO) or 250 nM torin-1 for 2 h and 5 h respectively. b Simultaneous diffusion of _JF585Halo-eIF4E and _JF646SNAP_f-eIF4G was analyzed by dual color cross-correlation spectroscopy (FCCS) in the indicated conditions. Cross-correlation was detected in the cytoplasm of control cells (left panel), but not in the nucleus (right panel). Minor residual eIF4E:eIF4G was detected in the cytoplasm upon 2 h mTOR inhibition (left panel). c, d Averaged autocorrelation curves representing individual diffusion of _JF585Halo-eIF4E (c) and _JF646SNAP_f-eIF4G (d) in the indicated conditions (N = 10 ± SEM). In control cells, both _JF585Halo-eIF4E (c) and _JF646SNAP_f-eIF4G (d) autocorrelations showed slower diffusion as compared to their nuclear counterparts. No changes were detected in the nuclear counterparts. Cytoplasmic eIF4E molecules diffuse as fast as the nuclear counterpart as early as 2 h upon torin-1 treatment, whereas cytoplasmic eIF4G mirrors the nuclear diffusion only after 5 h. e Total lysates (input) of cells described in (c–e) were subjected to cap-pull-down assay (m⁷GTP pull-down) and analyzed by western blotting with the indicated antibodies. eIF4E:eIF4G dissociation occurred as early as 2 h upon torin-1 treatment.

Fig. 4. eIF4E and eIF4G binding dynamics detected by FCS and SPT.

a NIH3T3 cells that express vector control or Halo-eIF4E and SNAP_f-eIF4G were treated with DMSO (vehicle) or 250 nM torin-1 for 2 h. Total cell lysates (input) were subjected to cap-pull-down assay (m⁷GTP-pull-down) and analyzed by western blotting. eIF4G and eIF4E antibodies detected both endogenous (endog.) and exogenous (exog.) proteins as indicated by the arrows. b Simultaneous diffusion of _JF585Halo-eIF4E and _JF646SNAP_f-eIF4G analyzed by dual-color fluorescent cross-correlation spectroscopy (FCCS) in cell treated with vehicle (left panel) or 250 nM torin-1 for 2 h (torin-1, right panel) in the cytoplasm (gray) and in the nucleus (red) (N = 10 ± SEM). Cross-correlation was detected in the cytoplasm of translating cells, but not in the nucleus, and abolished 2 h upon mTOR inhibition. c, d Individual _JF585Halo-eIF4E (c) and _JF646SNAP_f-eIF4G (d) averaged autocorrelation curves from (b). e Simultaneous single-particle tracking of _JF549Halo-eIF4E and _JF646SNAP_f-eIF4G was simultaneously recorded in the cytoplasm of an NIH3T3 (dotted line outlined the nucleus). Left: Diffusion properties of 4319 trajectories of _JF549Halo-eIF4E and 4001 trajectories of _JF646SNAP_f-eIF4G are displayed via heat maps. Each point is false‐colored according to the mean square displacement calculated over all displacements originating in a circle (r = 80 nm). f Co-movement analysis of _JF549Halo-eIF4E and _JF646SNAP_f-eIF4G. Maximum intensity projections of 5000 frames of _JF549Halo-eIF4E (in magenta) and 5000 frames of _JF646SNAP_f-eIF4G (in green) were simultaneously acquired at 100 Hz. The co-moving _JF549Halo-eIF4E (bold, magenta) and _JF646SNAP_f-eIF4G (bold, green) trajectories are displayed on top. Scale bar: 10 μm. Inset top: co-moving _JF549Halo-eIF4E/_JF646SNAP_f-eIF4G trajectories with their associated diffusion heat maps displayed at higher magnification (scale bar: 1 μm). g (left) The distribution of apparent diffusion coefficients shifts to a slower population when _JF549Halo-eIF4E is comoving with _JF646SNAP_f-eIF4G (in gray). (right) Violin plots of one-step mean square displacements for eIF4E (355,730, in magenta), for eIF4G (462,072, in green), and for co-moving eIF4E/eIF4G pairs (7947, in gray). Box quartile method: Tukey. The median line is shown, whisker method: min and max data. Two-sided t-test with two mean values of two distributions (*** means p  <  0.001, ** means p  <  0.01).

Fig. 5. Single-particle tracking of Halo-eIF4E in primary hippocampal neurons.

Rat hippocampal neurons were activated with TTX withdrawal with or without 250 nM torin-1. a Total cell lysates were analyzed by western blotting with the antibodies indicated next to the corresponding blot (n = 2). b Diffusion heat map of Halo-eIF4E trajectories obtained in inactivated neurons (16 h TTX) at a frame rate of 100 Hz (with a dotted neuronal outline, scale bar: 5 μm. Each point in the image is false‐colored according to the mean square displacement calculated over all displacements originating in a circle (r = 80 nm). Inset: Trajectories and corresponding diffusion heat map (scale bar: 1 μm). c Diffusion map of Halo-eIF4E of activated neurons after TTX withdrawal (scale bar: 5 μm). Inset: Trajectories and corresponding diffusion heat map (scale bar: 1 μm). d Diffusion speed of Halo-eIF4E in neurons treated with 250 nM torin-1 (red colors) (scale bar: 5 μm). e Single particle tracking was recorded at a frame rate of 200 Hz. The cumulative distribution function (CDF) of all single-molecule displacements for Halo-eIF4E is shifted to the left when compared to neurons treated with torin-1, which indicates a shift towards slower diffusion for Halo-eIF4E undergoing cap-dependent translation. CDF was obtained from 1226 and 742 trajectories in 10 individual dendrites of neurons treated with vehicle control or torin-1, respectively. The global two-component fit returns a global apparent D_slow of 0.32 ± 0.006 µm²/s and a global apparent D_fast of 4.10 ± 0.076 µm²/s. While the majority of control Halo-eIF4E exhibit slow diffusion (76 ± 0.6%), this percentage drops to 48 ± 0.6% when treated with torin-1. The data was recorded at 200 Hz, and we report apparent diffusion coefficients for that frame rate. f Left: Medium-intensity projection of 10,000 frames of _JF549Halo-eIF4E recorded simultaneously with sparsely photo-activated _PA-JF646Halo-eIF4E molecules (individual _PA-JF646Halo-eIF4E trajectories are shown in green) at 56 Hz frame rate (scale bar: 10 μm). Right: The associated _PA-JF646Halo-eIF4E diffusion heat map depicts that slow Halo-eIF4E molecules (in blue) linger near what appear to be spines in activated neurons (scale bar: 10 μm). Spines are indicated with an asterisk.

Fig. 6. Simultaneous single-particle tracking of endogenous Halo-eIF4E and SNAP_f-eIF4G in mESC cells that were differentiated into neurons.

a Top left: Diffusion heat map of 18,220 _JF585Halo-eIF4E trajectories from eight 10,000 frames 50 Hz movies obtained from a dense network consisting of more than 100 individual dendritic branches (scale bar: 10 μm). Top center: The corresponding heat map of the simultaneously recorded 26,438 _JF646SNAP_f-eIF4G trajectories. Top right: The distributions of apparent diffusion coefficients of _JF585Halo-eIF4E (in green) and _JF646SNAP_f-eIF4G (in magenta). Two-sided t-test with two mean values of two distributions (*** means p  <  0.001). b Cumulative distribution functions (CDF) were obtained from 26,438 eIF4G and 18,220 eIF4E trajectories. While the majority of Halo-eIF4E exhibit slow diffusion (54.3 ± 1.0%), this percentage increases to 73.2 ± 1.1% for SNAP_f-eIF4G. c Thick dendritic process of a cortical neuron from ARC^P/+;PCP-GFP;v-Glut2Cre mice expressing Halo-eIF4E after TTX withdrawal (dotted outline, scale bar: 10 μm). The median projection of the _JF646Halo-eIF4E channel of 5000 frames was recorded at 100 Hz (in magenta). Overlay of two ARC mRNA trajectories (in green) labeled with PCP-GFP that were simultaneously recorded. 1.6 × 1.6 μm² insets of _JF646Halo-eIF4E and two ARC mRNA trajectories with their associated diffusion heat maps. The mean square displacement (MSD) curve of the two ARC mRNA trajectories depicts corralling of mRNAs and reveals an exploration area of 0.02 μm². d Dendritic branches of cortical neurons from ARC^P/P mice expressing Halo-eIF4E after TTX withdrawal (dotted outline, scale bar: 2 μm). Left: Median projection of the PCP-GFP channel of 6000 frames recorded at 100 Hz outlines a dense region of neuronal processes. Overlaid are the trajectories of four ARC mRNA molecules (in white). Middle: Co-moving Halo-eIF4E molecules (in magenta) and unbound Halo-eIF4E molecules are depicted in gray. Right: The diffusion heat map of co-moving mRNA/ Halo-eIF4E molecules depicts almost static movements of the translation factor interacting with ARC-mRNA. e The distribution of distances of all ARC-mRNA and all eIF4E particles (in gray) consists of a peak of short values (<4 pixels) above a flat baseline. The co-movement algorithm (magenta line) efficiently selects the peak of colocalized trajectories.