NDP52 activates nuclear myosin VI to enhance RNA polymerase II transcription

Figures

• 

  Fig. 1. Myosin VI is distributed throughout the nucleus. a Cartoon depiction of the MVI domains and key features discussed in the text. b Immunofluorescence staining against MVI (magenta) and DNA (cyan) in HeLa cells and isolated nuclei (see Supplementary Fig. 1 for nuclei images). Arrows highlight filamentous structures within the nucleus. Images were acquired at the mid-point of the nucleus. Scale bar 10 μm for whole images and 1 μm for inserts. c Western blot against MVI following HeLa cell fractionation. Tubulin and lamin B are used as cytoplasmic and nuclear markers, respectively. d Representative images of transiently expressed NI- and LI-GFP-MVI in HeLa cells combined with DNA staining (cyan) (Supplementary Fig. 3). Images acquired as in B. e Pull-down of recombinant MVI using a 500 bp DNA substrate. f Plot of band intensities normalised to the input sample (5 μM). Errors bars represent SEM from three independent preparations. (**p < 0.001 by two-tailed t-test between presence and absence of DNA)

• 

  Fig. 2. DNA binding by myosin VI. a, b Fluorescence anisotropy titrations of MVI domains against a 40 bp fluorescein amidite (FAM)-DNA (50 nM). Data fitting was performed as described in Methods (K d ± SEM n = 3 independent experiments). Due to the low binding affinity, Motor1-814 could not be fitted using the model. c Cartoon depicting the secondary structure within the CBD. Binding partner motifs and known lipid binding sites are highlighted along with three predicted clusters of DNA binding. Black lines represent residues predicted to be involved in DNA binding. Sequence alignment shows the conservation of sites B and C. d Structure of CCBD with site B (red), C (blue) and WWY motif (green) (PDB:2KIA54). e, f Fluorescence anisotropy titrations of CBD constructs, as performed in a

• 

  Fig. 3. Regulation of myosin VI backfolding. a FRET titration of NMVITAIL against CBD (1 μM) or CBD(W1221A) in the presence of DNA, or NDP52, at 10 μM. Data fitting was performed as described in Methods. NDP52 and CBD(W1221A) data are plotted as disconnected points due to low signal change. See Supplementary Fig. 4a–c for raw intensity data. b Representative fluorescence spectra of 2 μM GFP-MVITAIL-RFP ± 15 μM NDP52. c FLIM measurements of in vivo MVI conformational dynamics. Above, schematic showing the constructs. Left, GFP intensity and lifetime images. Right, intensity and lifetime images of the biosensor. Lifetime (τ) is represented by a pseudocolor scale. Scale bar is 20 μm. d Pull-down of CBD by His-tagged NMVITAIL both at 10 μM ± 20 μM 40 bp DNA. P and S represent pellet and supernatant fractions. e Fluorescence anisotropy titrations of MVI domains against a 40 bp FAM DNA, as performed in Fig. 2a

• 

  Fig. 4. NDP52 is a nuclear myosin VI binding partner. a Cartoon and structural model depicting NDP52 domains as mentioned in the text. b Immunofluorescence staining against NDP52 (green) and DNA (cyan) in HeLa cells and isolated nuclei, as performed in Fig. 1b. Arrows highlight filamentous structures. Scale bar 10 μm for whole images and 1 μm for inserts. c Western blot against NDP52 following HeLa cell fractionation. Tubulin and lamin B were used as cytoplasmic and nuclear markers, respectively. d FRET titration of MVITAIL constructs against NDP52 (1 μM). Data fitting was performed as described in Methods giving a K d as plotted in e (error bars represent SEM from three independent experiments). f Western blot against NDP52 following isolation from nuclear extracts by MVITAIL constructs. Loading controls for the recombinant proteins is shown in Supplementary Fig. 9. g Immunofluorescence staining against NDP52 (green), MVI (magenta) and DNA (cyan), as performed in Fig. 1b. Insert depicts co-localising foci. Scale bar 10 μm for whole images and 1 μm for inserts. Pearson’s coefficient is shown for HeLa and nuclei (Supplementary Fig. 5b). Error bars represent SEM from 10 images

• 

  Fig. 5. NDP52-dependent dimerisation of myosin VI. a FRET titration of FITC-MVITAIL against 1 μM AF555-MVITAIL ± DNA (20 μM) and NDP52 (20 μM). Data fitting was performed as described in Methods giving a K d as plotted in b. See Supplementary Fig. 7a–c for raw intensity data. b Plot of K d from titrations in a and Supplementary Fig. 7d–g (error bars represent SEM from three independent experiments). c Velocity histogram from sliding filament assay with MVI immobilised alone (blue), through antibody (red) and NDP52 (green). Insert shows first frame (red) and after 60 s (green). d Two routes of NDP52-dependent dimerisation with different stoichiometry. (i) NDP52 unfolds MVI then directly recruits a second molecule. (ii) Each MVI is unfolded by an individual NDP52

• 

  Fig. 6. NDP52 transcription regulation. a Immunofluorescence staining against NDP52 (green), RNAPII (magenta) and DNA (cyan), as performed in Fig. 1b. Insert depicts colocalising foci. Scale bar 10 μm for whole images and 1 μm for inserts. b Immunofluorescence on HeLa nuclei as in (a) with Pearson’s coefficient (error bars represent SEM from 10 images). c Co-immunoprecipitation of NDP52 and RNAPII, using an antibody against NDP52 to immunoprecipitate RNAPII from HeLa cells. Immunoprecipitation does not occur in HeLa cells subjected to siRNA knockdown of NDP52. d Western blot against RNAPII following isolation from nuclear extracts by NDP52 constructs and CoCoA. Loading controls for the recombinant proteins is shown in Supplementary Fig. 9. e Pull-down of CoCoA by His-tagged NMVITAIL both at 5 μM. P and S represent pellet and supernatant fractions. f In vitro transcription by HelaScribe extracts following antibody depletion as described in Methods. Samples were normalised to a non-depleted control reaction (error bars represent SEM **p < 0.001 by two-tailed t-test) from 5 independent experiments (Supplementary Fig. 8a, b)

• 

  Fig. 7. Coupling myosin VI to the RNAPII complex. a In vitro transcription by HelaScribe extracts in the presence of competitor MVI domains at 25 μM, unless stated otherwise. Sites B and C refer to CCBDΔSiteB and CCBDΔSiteC, respectively. TIP refers to the control reaction performed in the presence of 25 μM of the MVI inhibitor (TIP). Samples were normalised to the control sample in Fig. 6f (error bars represent SEM from five independent experiments **p < 0.001 by two-tailed t-test). See Supplementary Fig. 8c for control experiments. b In vitro transcription following antibody depletion and rescue using recombinant MVI (1 μM), NDP52 (5 μM) and F-actin (5 μM), as described in Methods (error bars represent SEM from five independent experiments **p < 0.001 by two-tailed t-test). c Co-immunoprecipitation of MVI and RNAPII, using an antibody against MVI to immunoprecipitate RNAPII from HeLa cells. Immunoprecipitation does not occur in HeLa cells subjected to siRNA knockdown of MVI. d Western blot against RNAPII following isolation from nuclear extracts by MVI constructs. Loading controls for the recombinant proteins is shown in Supplementary Fig. 9. e Working model of MVI in transcription…

• 

  Fig. 8. Myosin VI is linked to oestrogen signalling. a Pull-down of 2 μM CBD by 10 μM GST-ER. P and S represent pellet and supernatant fractions. b Expression of ER gene targets following siRNA knockdown of MVI in MCF-7 cells. Expression is plotted as a percentage of expression in mock cells. ER (ESR1) and β-actin were used to reflect global changes in transcription. c Luciferase reporter assay driven by the ERE promoter in MCF-7 cells. Estradiol led to a 5-fold increase in promoter activity. siRNA knockdown of MVI led to a 3-fold reduction in activity. Ethanol was used as a carrier control for the experiments (error bars represent SEM from three independent experiments **p < 0.001 by two-tailed t-test). d Working model of MVI recruitment by the ER. MVI binds to the ER and then gets activated by NDP52 (or CoCoA), which enables binding to RNAPII through actin. Association with NDP52, or CoCoA, ties MVI with the general transcription co-activators to initiate recruitment of RNAPII