Structural dynamics of the E6AP/UBE3A-E6-p53 enzyme-substrate complex

Abstract

Deregulation of the ubiquitin ligase E6AP is causally linked to the development of human disease, including cervical cancer. In complex with the E6 oncoprotein of human papillomaviruses, E6AP targets the tumor suppressor p53 for degradation, thereby contributing to carcinogenesis. Moreover, E6 acts as a potent activator of E6AP by a yet unknown mechanism. However, structural information explaining how the E6AP-E6-p53 enzyme-substrate complex is assembled, and how E6 stimulates E6AP, is largely missing. Here, we develop and apply different crosslinking mass spectrometry-based approaches to study the E6AP-E6-p53 interplay. We show that binding of E6 induces conformational rearrangements in E6AP, thereby positioning E6 and p53 in the immediate vicinity of the catalytic center of E6AP. Our data provide structural and functional insights into the dynamics of the full-length E6AP-E6-p53 enzyme-substrate complex, demonstrating how E6 can stimulate the ubiquitin ligase activity of E6AP while facilitating ubiquitin transfer from E6AP onto p53.

Fig. 1

Binding of E6 induces a conformational rearrangement of E6AP. Pattern of intralink distribution within E6AP as determined by XL-MS a in the absence of the HPV E6 oncoprotein, b with wild-type (wt) E6, and c with the E6_L50E mutant, which does not bind to E6AP. Intralinks are shown in dark blue and interlinks in dark green; lysine residues are shown in black. The catalytic cysteine residue of E6AP at position 820 is indicated in red. The AZUL domain, HERC2 and E6 binding sites, and the HECT domain are indicated in pastel green, sulfur yellow and sand yellow, respectively. Regions in E6 representing zinc finger motifs are colored in pastel turquoise and the PDZ binding domain is colored in mint turquoise. d Changes in crosslink abundance for each unique crosslinking site with E6AP upon binding of wild-type E6 were used to calibrate and normalize the quantitative XL-MS (q-XL-MS) data. Changes are expressed as fold change (log₂ ratio of abundance of E6AP in the presence of E6 versus abundance of E6AP alone). The horizontal red line indicates the significance threshold (fold change: log2 ratio ≤ ±1.5). Changes in monolinks are shown in blue and significantly changed intralinks within E6AP are shown in green (relative increase upon binding of E6) and red (relative decrease upon binding of E6), respectively. The p values for each quantified link are indicated. e Applying q-XL-MS to recombinantly expressed E6AP in the presence or absence of wild-type E6 identifies numerous high-confidence crosslinks. Only crosslinks that could be reproducibly quantified from the pool of identified high-confidence crosslinks in both samples (with and without E6) (n = 3, each sample analyzed additionally as technical duplicate) and consistently over 3 different biological replicates are shown (violation = 0; p value ≤ 0.01 (two-sided t-test)). Depicted in green are crosslinks that were found to be significantly upregulated upon binding of E6, while downregulated links are shown in red (defined as a log2 change of ≥ ±1.5). Crosslinks with no significant change are depicted in gray while links that could not be reliably quantified are show with a dashed line in light gray. f Quantification of the change in abundance of identified intralinks within E6AP, when incubated with the binding-deficient E6_L50E mutant vs. E6AP alone, reveals no significant up- or downregulated links. For monolinks, see Supplementary Figure 5

Fig. 2

SILAC-XL-MS reveals weak interactions between E6AP N-termini. Identified interlinks between two E6AP protomers (i.e., heavy and light version of E6AP). Shown are the interlinks in the absence (a) and presence of E6 (b)

Fig. 3

E6 contacts the HECT domain of E6AP. a Shown are the identified interlinks between E6AP and E6. Lysine residues and regions of known functions within E6AP and E6 are color-coded as in Fig. 1. b List with identified interlinks between E6AP and E6. Numbering is according to human E6AP isoform 1. Links which can be mapped to existing PDB structures (4XR8; 1C4Z) are highlighted. c Structural model of the binding interface between the HECT domain of E6AP and E6. The localization densities of the HECT domain of E6AP (residues 495–846) and E6 (residues 1–151) are shown in sand yellow and light green, respectively, with a single representative ribbon structure embedded. For clarity, UbcH7, a cognate E2 of E6AP, was also added and is shown in yellow (PDB: 1C4Z). Detected interlinks between E6 and E6AP that map to PDB structures 4XR8 and 1C4Z, respectively, are highlighted in orange. The catalytic cysteine on position C820 in the HECT domain of E6AP is shown as a red ball. The model is shown from the top (left) and from a side view (right)

Fig. 4

Distinctive binding sites within the E6AP–E6–p53 complex. a Shown are the identified interlinks between E6AP, E6, and p53. Lysine residues and regions of known functions within E6AP and E6 are color-coded as in Fig. 1. Functional domains in p53—transactivation domain, DNA binding domain, oligomerization domain and regulatory region—are shown in signal white, gray white, telegrey, and white aluminum, respectively. Crosslinks between sequence identical peptides that are bijective within the particular protein sequence are shown as red loops. b List with identified interlinks between p53, E6AP, and E6. Numbering is according to human E6AP isoform 1. Links which can be mapped to existing PDB structures (4XR8; 1C4Z) are highlighted. c Structural model of the binding interface between the HECT domain of E6AP, E6, and p53. The localization densities of the HECT domain of E6AP (residues 495–846), E6 (residues 1–151), and p53 (residues 94–292) are shown in sand yellow, light green, and telegrey, respectively, with a single representative ribbon structure embedded. For clarity, UbcH7, a cognate E2 of E6AP was also added and is shown in yellow (PDB: 1C4Z). Detected interlinks between E6, p53, and E6AP that map to PDB structures 4XR8 and 1C4Z, respectively, are highlighted in orange. The catalytic cysteine at position C820 in the HECT domain of E6AP is shown as a red ball. Residue K292 of p53 (most C-terminal lysine residue in the model) is shown as a gray ball. The model is shown from a side view (left), using the exact same view as in Fig. 3 right panel, and rotated around its axis (right)

Fig. 5

Model of E6-mediated ubiquitylation of p53. Model of E6-mediated ubiquitylation of p53 in the E6AP–E6–p53 enzyme–substrate complex. Upon binding, E6 induces a conformational rearrangement in E6AP that brings the AZUL domain closer to the HECT domain. This leads to the concomitant positioning of E6 and the substrate p53 into the direct vicinity of the catalytic center of E6AP within the enzyme–substrate complex. There, E6 now functions as an adaptor facilitating the direct transfer of ubiquitin from the HECT domain of E6AP to p53