Adenoviral E1A Exploits Flexibility and Disorder to Target Cellular Proteins

Abstract

Direct interaction between intrinsically disordered proteins (IDPs) is often difficult to characterize hampering the elucidation of their binding mechanism. Particularly challenging is the study of fuzzy complexes, in which the intrinsically disordered proteins or regions retain conformational freedom within the assembly. To date, nuclear magnetic resonance spectroscopy has proven to be one of the most powerful techniques to characterize at the atomic level intrinsically disordered proteins and their interactions, including those cases where the formed complexes are highly dynamic. Here, we present the characterization of the interaction between a viral protein, the Early region 1A protein from Adenovirus (E1A), and a disordered region of the human CREB-binding protein, namely the fourth intrinsically disordered linker CBP-ID4. E1A was widely studied as a prototypical viral oncogene. Its interaction with two folded domains of CBP was mapped, providing hints for understanding some functional aspects of the interaction with this transcriptional coactivator. However, the role of the flexible linker connecting these two globular domains of CBP in this interaction was never explored before.

Keywords: CBP; E1A; IDP; NMR; fuzzy complex.

Figure 1

Structural features of (A) human adenovirus E1A12S and (B) human CBP-ID4. For the latter, also the flanking domains CBP-TAZ2 and CBP-NCBD are displayed for illustrative purposes. The two models were built as described in the experimental section.

Figure 2

Secondary structure propensity of E1A12S (A) and CBP-ID4 (B). Proline residues are highlighted along the protein sequence as black triangles. Positive residues are highlighted as blue squares, negative residues are highlighted as red squares.

Figure 3

Titration of E1A12S with CBP-ID4 followed by ¹H-¹⁵N BEST-TROSY experiments. E1A12S (black) E1A12S:CBP-ID4 at 1:0.5 molar ratio (yellow); E1A12S:CBP-ID4 at 1:1 molar ratio (red); and E1A12S:CBP-ID4 at 1:2 molar ratio (green). All the experiments were acquired at 283 K, using a 22.3 T Bruker Avance III NMR spectrometer equipped with a TCI CryoProbe^TM.

Figure 4

(A) Comparison of the cross-peak intensities in ¹H-¹⁵N BEST-TROSY experiments for E1A12S in the isolated form (black) and E1A12S:CBP-ID4 1:1 complex (red); (B) mapping the chemical shift perturbations of E1A12S signals in the E1A12S:CBP-ID4 1:1 complex using the Garrett equation; (C) comparison of a selected region of ¹H-¹⁵N BEST-TROSY spectrum of E1A12S (black), E1A12S:CBP-ID4 1:0.5 (yellow), E1A12S:CBP-ID4 1:1 (red), and E1A12S:CBP-ID4 1:2 molar ratio (green) with signals assignment.

Figure 5

Nuclear relaxation data for the backbone amide ¹⁵N obtained for E1A12S (black) and the 1:1 complex E1A12S:CBP-ID4 (red). From top to bottom, ¹⁵N longitudinal relaxation rates (R₁, Hz), ¹⁵N transverse relaxation rates (R₂, Hz) and ¹H-¹⁵N NOE values. In the plot, the missing data derive from residues that are not observable or not resolved.

Figure 6

Titration of CBP-ID4 with E1A12S followed by ¹H-¹⁵N FHSQC experiments. CBP-ID4 (black); CBP-ID4:E1A12S at 1:0.5 molar ratio (yellow); CBP-ID4:E1A12S at 1:1 molar ratio (red); and CBP-ID4:E1A12S at 1:2 molar ratio (green). All the experiments were acquired at 283 K, using a 22.3 T Bruker Avance III NMR spectrometer equipped with a TCI CryoProbe^TM.

Figure 7

(A) Comparison of 2D H^α-CON^pro spectra of CBP-ID4 (black) and CBP-ID4:E1A12S 1:1 (red); (B) comparison 2D H^N-CON spectra of CBP-ID4 (black) and CBP-ID4:E1A12S 1:1 (red). The assignment of selected peaks is also reported. (C) Comparison of cross-peak intensities in 2D H^N-CON and 2D H^α-CON^pro experiments for CBP-ID4 in the isolated form (black) and CBP-ID4:E1A12S 1:1 complex (red). All the spectra were acquired at 283 K with a 16.4 T Bruker Avance NEO NMR spectrometer equipped with a ¹³C TXO CryoProbe^TM.

Figure 8

Nuclear relaxation data for backbone amide ¹⁵N obtained for CBP-ID4 (black) and the 1:1 complex CBP-ID4:E1A12S (red). From the top to the bottom, ¹⁵N longitudinal relaxation rates (R₁, Hz), ¹⁵N transverse relaxation rates (R₂, Hz) including proline residues, and ¹H-¹⁵N NOE.

Figure 9

Cartoon describing one of the possible modes of interaction of E1A12S with the CBP-region comprising the TAZ2 domain and ID4. The structure of TAZ2 (grey surface) and of the fragment 53–91 of E1A is the first model of the PDB ID 2KJE entry (solution structure). The fragments 1–52 and 92–243 of E1A as well as the CBP-ID4 fragment were selected among the conformers of Flexible Meccano generated ensembles.