Extended Data Fig. 10: Probing protein accessibility in endogenous mRNP complexes.

a. Schematic of the experiment to probe protein accessibility in mRNP complexes. The nuclear or cytoplasmic extract from K562 cells, tagged homozygously and endogenously with either GFP-3C-THOC5 or GFP-3C-EIF4A3, was incubated with a fluorescently labelled (fluorophore: AF647) 15 kDa anti-GFP nanobody. The extracts were then applied to a sucrose density gradient to separate free proteins from mRNPs, which migrate in heavy (later) sucrose gradient fractions. The gradient fractions were analyzed by SDS–PAGE. Due to high affinity of the anti-GFP nanobody to GFP (~1 pM), the nanobody stays bound to the GFP fusion during gel electrophoresis (see Methods for details). Fluorescence imaging allows quantification of the respective sedimentation profiles for the GFP fusion proteins (GFP-THOC5 or GFP-EI4A3, green channel) and the anti-GFP nanobody-bound fusion proteins (red channel, colored in magenta). When the GFP-tagged protein is accessible in mRNPs, then the anti-GFP nanobody signal closely follows the profile of the GFP-tagged protein. In contrast, when a GFP-tagged protein is inaccessible in mRNPs, the anti-GFP nanobody signal follows the GFP signal in early (light) sucrose gradient fractions that contain free proteins but shows reduced intensity in later (heavy) fractions. b. The anti-GFP nanobody signal closely follows the GFP-THOC5 signal, showing that GFP-THOC5 is accessible in mRNP complexes. Shown is the fluorescence signal from SDS-PAGE gels of GFP-THOC5 nuclear extract incubated with the AF647-labeled anti-GFP nanobody (top) and normalized sedimentation profiles (bottom). Sedimentation plots show mean normalized intensity values determined from three gels (solid lines) and standard deviations (transparent areas). The grey box indicates the peak gradient fractions of purified TREX–mRNPs (see Extended Data Fig. 4). This experiment was done four times. For gel source data, see Supplementary Fig. 9. c. As for panel b, but for GFP-EIF4A3 in nuclear extract. In the high molecular weight fractions of the sucrose density gradient, GFP-3C-EIF4A3 is poorly accessible to the anti-GFP nanobody. This experiment was done four times. For gel source data, see Supplementary Fig. 10. d. As for panel b, but for GFP-EIF4A3 in cytoplasmic extract. In the high molecular weight fractions of the sucrose density gradient, GFP-EIF4A3 remains accessible to the anti-GFP nanobody, in contrast to GFP-EIF4A3 in nuclear extract, which is shown in panel c. This experiment was done twice. For gel source data, see Supplementary Fig. 11. e. Western blot experiment that shows the different depletion efficiencies of THOC1-GFP (ectopically overexpressed; Lenti O/E), GFP-THOC5 (endogenously tagged; endo), or GFP-EIF4A3 (endogenously tagged; endo) from nuclear extract using GFP-Trap resin (containing an anti-GFP nanobody coupled to 90 µm agarose beads) after three rounds of depletion. While THOC1-GFP and GFP-THOC5 are completely depleted in the supernatant, GFP-EIF4A3 is very inefficiently depleted. Anti-PSMA7 blots (a proteasome subunit) serve as loading controls. These experiments were done three times. For gel source data, see Supplementary Fig. 12. f. Cartoon model showing the position and nanobody-accessibility of GFP-tagged THOC5 or EIF4A3 in TREX–mRNPs, based on the accessibility to the anti-GFP nanobody and anti-GFP resin in panels b, c, and e.