A sequential binding mechanism for 5′ splice site recognition and modulation for the human U1 snRNP

Figures

• 

  Fig. 1. A −1A bulge induces dynamic RNA binding to U1 snRNP. A Crystal structure of minimal U1 snRNP (PDB: 4PJO). Arrows indicate relative placement of modifications for single molecule measurements. B RNA oligos containing 5′SS. Shaded region indicates predicted base pairing interactions with the U1 snRNA (shown at the top, above a schematic of the exon (box)/intron (line) junction). C SPR sensorgrams showing the association and dissociation of surface-tethered (top) 11bp and (bottom) 11bp-1A RNAs at various U1 snRNP concentrations (0.02 to 100 nM). D Cartoon schematic of the two-color CoSMoS assay for monitoring U1/RNA interactions. E Fields of view under 633 nm (left, immobilized U1 snRNP molecules) and 532 nm (right, interacting RNA oligos) excitation (scale bar = 20 µm). Inset highlights colocalization (scale bar 1 µm). Fluorescent beads were included as fiducial makers (yellow arrow). Images are rendered by averaging three consecutive images and applying uniform brightness and contrast values. F Fluorescence in arbitrary units (au) across time in seconds (s) showing the binding of 9bp (top) and 9bp-1A (bottom) to surface tethered U1 snRNP (0.33 frames/s, Hz). Black lines indicate…

• 

  Fig. 2. Branaplam stabilizes -1A bulged 5′SS RNA binding to U1 snRNP. A RNA oligos containing -1A bulged 5′SS. Shaded region indicates predicted base pairing interactions with the U1 snRNA (top). B Chemical structure of branaplam. C SPR showing the association and dissociation of 11bp-1A RNA across various concentrations of branaplam (0–5 µM) and at 10 nM U1 snRNP. D Dose response curve showing the fitted dissociation rates (koff) of response units vs branaplam concentration for 11bp-1A (white circles) and 11bp (black circles) RNAs. Data for 11bp-1A is overlaid with the fitted equation to determine EC50 value (1.3 ± 0.1 µM). E Fluorescence in arbitrary units (au) across time in seconds (s) of 1 nM 9bp-1A RNA binding to immobilized U1 in the presence of 100 nM U1-C plus (top) DMSO (0.33 Hz) or (bottom) 10 µM branaplam (0.11 Hz) overlaid with idealizations (black lines). F Dose response curves of 9bp-1A RNA binding to U1 snRNP vs branaplam concentration. The fraction bound at each branaplam concentration (white cirlcle) is overlaid with a fitted equation (solid red line) and the 95% confidence interval of the fitted equation (shaded red region) to estimate an EC50 value (0.45 ± 0.17 µM). G…

• 

  Fig. 3. U1-C is required for branaplam modulation of -1A 5′SS RNA binding. A Fluorescence in arbitrary units (au) across time in seconds (s) of 1 nM 9bp-1A binding in the absence of U1-C with (top) DSMO and (bottom) 10 µM branaplam overlaid with idealizations (black lines). B Violin plots of bound dwell time distributions at 1 nM 9bp-1A RNA with and without U1-C and/or branaplam. DMSO was included in the absence of branaplam. C Change in fluorescence due to branaplam binding to a duplex of 11bp-1A and U1 snRNP in the absence (grey triangles) and presence (white circles) of U1-C by microscale thermophoresis (MST). The change in fluorescence in the presence of U1-C is overlaid with a fitted equation (solid purple) and 95% confidence interval of the fitted equation (shaded purple region) to estimate a KD value (2.69 ± 0.36 µM). D Violin plots showing the bound dwell time distributions across different permutations of U1-C and branaplam concentrations in solution across indicated RNA oligo sequences (SMN2, FOXM1, SF3B3, HTT*). Each violin plot is overlaid with box plot that show the median (horizontal line), interquartile range (IQR, box), and whiskers representing data within…

• 

  Fig. 4. U1-C promotes and stabilizes the binding of a -1A RNA to U1 snRNP. A Dose response curves of 9bp-1A RNA binding to U1 snRNP vs. U1-C concentration. The red line and shading indicate the fit and 95% CI to the fitted equation to estimate an EC50 value (1.3 ± 0.20 nM). B Cumulative probability distribution of unbound dwell times at 1 nM 9bp-1A RNA across U1-C concentrations (0 nM, N = 3106; 0.1 nM, N = 2920; 1 nM, N = 5543; 10 nM, N = 3169; 100 nM, N = 5154). C Unbound time constants (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\tau }_{U}$$\end{document}τU) determined from MLE of a monoexponential distribution for unbound dwell times overlaid with a fit to the fitted equation (EC50 = 1.8 ± 0.5 nM). Plotted \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\tau }_{U}$$\end{document}τU values (circles) are shown as mean ±…

• 

  Fig. 5. The impact of U1-C on 5′SS recognition is sequence dependent. A RNA oligos. Shaded region indicates predicted base pair interactions with the U1 snRNA (top). B Apparent association rates for 9 bp (left) and 9bp-1A RNAs (right) at various concentrations in the absence and presence of saturating U1-C (100 nM). Data are overlaid with linear fits (solid line) to determine kon (9 bp with 100 nM U1-C: kon = 3.9 ± 0.4 × 106 M−1 s−1, R2 = 0.97; 9 bp without U1-C: kon = 1.8 ± 0.5 × 106 M−1 s−1, R2 = 0.95; 9bp-1A with 100 nM U1-C: kon = 3.2 ± 0.6 × 106 M−1 s−1, R2 = 0.98; 9bp-1A without U1-C: kon = 3.2 ± 0.6 × 106 M−1 s−1, R2 = 0.98). Shaded region indicates the 95% confidence interval of the linear regression. C Violin plots showing the distributions of unbound (top) and bound (bottom) dwell times for various RNAs at 0 and 100 nM U1-C. Data for 9bp-1A is presented at 1 nM. N indicates the number of single molecule events included in the violin plot. Each violin plot is overlaid with box plot that show the median (horizontal line), interquartile range (IQR, box), and whiskers representing data within 1.5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym}…

• 

  Fig. 6. A sequential binding mechanism for -1A bulged 5′SS RNA recognition by U1 snRNP and modulation by branaplam. A kinetic model for -1A bulged 5′SS association and dissociation in the presence and absence of U1-C and branaplam. Optimized rate transitions for 9bp-1A are provided in Table 1. Equilibrium arrows in grey indicate transitions that are not supported by our experimental data or mathematical modeling.