Severe acute respiratory syndrome coronavirus papain-like protease: structure of a viral deubiquitinating enzyme

Abstract

Replication of severe acute respiratory syndrome (SARS) coronavirus (SARS-CoV) requires proteolytic processing of the replicase polyprotein by two viral cysteine proteases, a chymotrypsin-like protease (3CLpro) and a papain-like protease (PLpro). These proteases are important targets for development of antiviral drugs that would inhibit viral replication and reduce mortality associated with outbreaks of SARS-CoV. In this work, we describe the 1.85-A crystal structure of the catalytic core of SARS-CoV PLpro and show that the overall architecture adopts a fold closely resembling that of known deubiquitinating enzymes. Key features, however, distinguish PLpro from characterized deubiquitinating enzymes, including an intact zinc-binding motif, an unobstructed catalytically competent active site, and the presence of an intriguing, ubiquitin-like N-terminal domain. To gain insight into the active-site recognition of the C-terminal tail of ubiquitin and the related LXGG motif, we propose a model of PLpro in complex with ubiquitin-aldehyde that reveals well defined sites within the catalytic cleft that help to account for strict substrate-recognition motifs.

Fig. 1.

Organization of the SARS-CoV genome. The location of the different nsps in ORF1a/1ab and the ORFs for structural and accessory proteins are marked. PLpro and 3CLpro cleavage sites are indicated by red and black vertical lines, respectively. (Inset) Arrangement of different functional subdomains of nsp3. The location of PLpro is highlighted in red. The N- and C-terminal cleavage sites that define the boundaries of nsp3 are indicated by ↓.

Fig. 2.

Domain organization and structural motifs of SARS-CoV PLpro. (A) Locations of the Ubl (pink), thumb (green), palm (yellow), and fingers (pale blue) domains are indicated by colored boxes. α-Helices (orange) and β-sheets (blue) are numbered and depicted as ribbons. The zinc atom (red) is shown in space-fill representation, and zinc-coordinating cysteines and catalytic-triad residues are shown as ball-and-stick representations. (B) Structural superposition of residues 1–71 of ubiquitin (yellow) with residues 4–62 of the Ubl domain of SARS-CoV PLpro (violet). α1 and β1–3 of PLpro are labeled. The N and C termini of the aligned proteins are indicated. The rms deviation of 50 aligned residues is 2.12 Å at 16% sequence identity. (C) Stereoview of the electron density of the tetrahedrally coordinated zinc atom. A 2F_o − F_c map is contoured at 1.8σ (blue), and an F_o − F_c omit map of the zinc atom is contoured at 8σ (red).

Fig. 3.

Comparison of SARS-CoV PLpro with the cellular DUBs USP14 (PDB ID code 2AYN) and HAUSP (PDB ID code 1NB8). One hundred eighty-two residues of each protein, as chosen by the Web-based server SSM, were structurally aligned and superimposed. A ribbon diagram shows PLpro in blue, USP14 in red, and HAUSP in yellow. Catalytic triad residues are shown by ball-and-stick representations.

Fig. 4.

The SARS-CoV PLpro and USP14 active sites. (A) SARS-CoV PLpro catalytic triad residues, C112, H273, and D287, and other important active-site residues. Distances between residues are indicated in angstroms. The hydrogen bond between D109 and W97 is indicated by an arrow. (B) Comparison of USP14 and SARS-CoV PLpro BL1 and BL2 loop regions. Corresponding regions of unbound USP14 (red), Ubal-complexed USP14 (yellow), and PLpro (blue) are shown superimposed. The BL1 and BL2 loop regions are indicated. The BL1 loop region of PLpro is colored in green. The catalytic triad residues are shown by a ball-and-stick representation.

Fig. 5.

Comparison of the ubiquitin-binding surfaces of HAUSP, USP14, and PLpro based on modeling studies. For modeling ubiquitin into the SARS-CoV PLpro active site, the structures of the Ubal-bound forms of HAUSP (1nbf) and USP14 (2ayo) were superimposed onto the PLpro structure and analyzed. Contacts at the C-terminal tail of ubiquitin and two interacting surfaces of the PLpro palm domain were manually edited and minimized by using cns. The ubiquitin molecule is shown as a ribbon diagram.

Fig. 6.

Hypothetical model of the interaction of ubiquitin with the PLpro active site based on the structures of HAUSP and USP14 complexed with Ubal. (A) Modeled interactions between the C-terminal tail of Ubal (pink backbone) and the PLpro (green backbone). PLpro residues are labeled in black, and the ubiquitin side chains are labeled P1–P5 in blue. Proposed hydrogen bonds are indicated by dashed lines. The BL2 loop is shown in magenta. (B) A surface representation of the PLpro active-site tunnel is shown complexed with modeled Ubal. The Ubal is shown by a ball-and-stick representation. The P1–P5 positions of Ubal are labeled.