Exploring the dynamics and interactions of the N-myc transactivation domain through solution nuclear magnetic resonance spectroscopy

Abstract

Myc proteins are transcription factors crucial for cell proliferation. They have a C-terminal domain that mediates Max and DNA binding, and an N-terminal disordered region culminating in the transactivation domain (TAD). The TAD participates in many protein-protein interactions, notably with kinases that promote stability (Aurora-A) or degradation (ERK1, GSK3) via the ubiquitin-proteasome system. We probed the structure, dynamics and interactions of N-myc TAD using nuclear magnetic resonance (NMR) spectroscopy following its complete backbone assignment. Chemical shift analysis revealed that N-myc has two regions with clear helical propensity: Trp77-Glu86 and Ala122-Glu132. These regions also have more restricted ps-ns motions than the rest of the TAD, and, along with the phosphodegron, have comparatively high transverse (R2) 15N relaxation rates, indicative of slower timescale dynamics and/or chemical exchange. Collectively these features suggest differential propensities for structure and interaction, either internal or with binding partners, across the TAD. Solution studies on the interaction between N-myc and Aurora-A revealed a previously uncharacterised binding site. The specificity and kinetics of sequential phosphorylation of N-myc by ERK1 and GSK3 were characterised using NMR and resulted in no significant structural changes outside the phosphodegron. When the phosphodegron was doubly phosphorylated, N-myc formed a robust interaction with the Fbxw7-Skp1 complex, but mapping the interaction by NMR suggests a more extensive interface. Our study provides foundational insights into N-myc TAD dynamics and a backbone assignment that will underpin future work on the structure, dynamics, interactions and regulatory post-translational modifications of this key oncoprotein.

Keywords: NMR spectroscopy; intrinsically disordered proteins; myc; neuroblastoma; phosphorylation/dephosphorylation; protein–protein interactions.

Figure 1.. N-myc domain structure and TAD sequence.

(A) Domain structure of N-myc showing the positions of the structured C-terminal DNA-binding domain (DBD), myc boxes and the N-terminal transactivation domain (TAD). The TAD is magnified below, showing the positions of myc boxes MB0, MBI and MBII and the Aurora-A interaction helix (AIH). (B) Primary sequence of the N-myc TAD; hydrophobic residues are shown in bold, basic residues in blue, acidic residues in red, polar residues in green. The positions of the myc boxes and AIH are underlined.

Figure 2.. Assigned ¹H–¹⁵N HSQC spectrum of the N-myc TAD.

(A) Overall spectrum (omitting Trp and Arg Hε–Nε correlations). Plot shows the assignment of peaks outside the central region, note some of the weak outlying peaks e.g. Val78, Leu112, Lys51. Gln (ε) and Asn (δ) sidechain NH₂ peaks (¹⁵N ∼113 ppm) are unassigned. (B) A magnified view of the central region showing the lack of dispersion here as well as variation in the intensity of peaks. (C, D) Sidechain Hε–Nε correlations. Four Trp (C) and three Arg (D) sidechain correlations are resolved accounting for Trp51, Trp77, Trp88, Trp117, Arg65, Arg123, and Arg128.

Figure 3.. Structural propensities within N-myc TAD.

Secondary shifts (Δδ) for Cα (A), backbone carbonyl (B) and Cβ (C) nuclei across the sequence of N-myc TAD. Positive runs of Cα/CO Δδ values indicate α-helices, negative runs of Cα/CO Δδ values indicate β-strands and values nearing zero or with variation in sign of consecutive values indicate a lack of structure. The trends are mirrored for Cβ Δδ values. A comparison of Cα Δδ for N-myc TAD with the truncated variants is shown in Supplementary Figure S3A. The four Cβ Δδ values in purple are for the four Cys residues in the sequence; we contend that the anomalous values are due to an error in the reference coil value for Cys [60].

Figure 4.. Relaxation measurements for the N-myc TAD.

¹H–¹⁵N heteronuclear NOEs (A), longitudinal (R₁, B) and transverse relaxation rates (R₂, C). Experiments were carried out at 10°C. Residues missing from the plots include prolines and overlapping/broad peaks for which intensity changes could not be accurately gauged.

Figure 5.. Circular dichroism (CD) spectroscopy of N-myc TAD peptides.

(A) N-myc TAD sequence with the positions of the peptides coloured and myc box positions underlined. (B) Overlaid CD spectra for the three peptides in buffer. (C–E) Comparison of CD spectra for the MBI peptide (C), AIH peptide (D) and MBII+ peptide (E) in buffer, 15% TFE and 30% TFE. Experiments were carried out at 5°C.

Figure 6.. NMR titration experiments for Aurora A kinase domain with N-myc TAD truncates.

(A) ¹H–¹⁵N HSQC spectrum of GB1-N-myc_18–72 (red) and then after addition of increasing amounts of Aurora A kinase domain (lime green, black, sky blue, orange). (B) ¹H–¹⁵N HSQC spectrum of N-myc_64–137 (red) and then after addition of increasing amounts of Aurora A kinase domain (lime green, black, sky blue, orange). (C) Ratio of intensities at a [Aurora A]:[N-myc] molar ratio of 0.2:1 compared with unbound N-myc truncates. The GB1 sequence is renumbered as negative numbers. (D) Changes in chemical shift perturbations as a linear function of [Aurora A]:[N-myc].

Figure 7.. In vitro phosphorylation of N-myc TAD followed by NMR.

(A) ¹H–¹⁵N HSQC spectrum of N-myc TAD (red) and then after addition of ERK1 (blue). (B) ¹H–¹⁵N HSQC spectrum of N-myc TAD S7A, pre-phosphorylated by ERK1 (blue) and then after addition of GSK3 (green).

Figure 8.. Doubly phosphorylated N-myc TAD (pSer62, pThr58) binds to the Fbxw7–Skp1 complex.

Analytical SEC using a Superose12 10/300 GL column was used to assess complex formation between the Fbxw7–Skp1 complex and either singly phosphorylated (pSer62) N-myc TAD (A), or doubly phosphorylated (pSer62, pThr58) N-myc TAD (B). The chromatogram for the complex is shown in blue, while the chromatograms for N-myc TAD alone or Fbxw7–Skp1 alone are shown in red and black, respectively. 0.5 ml fractions were taken at the same positions in both runs and analysed by SDS–PAGE.

Figure 9.. NMR titration of the Fbxw7–Skp1 complex with in situ generated doubly phosphorylated N-myc TAD (pSer62, pThr58).

(A) The central backbone region of the ¹H–¹⁵N HSQC spectrum before (green) and after addition of Fbxw7–Skp1 (pink). (B) Magnified views of the Gly region (left) and Trp Hε–Nε region (right) of the ¹H–¹⁵N HSQC spectrum. Inset is an aliased region (from 84 to 85 ppm) showing Arg Hε–Nε correlations. Peaks that disappear highlight the parts of the sequence that interact most strongly. (C) Ratio of peak intensities for each residue as a function of increasing molar ratio of [Fbxw7–Skp1]:[N-myc].