Structure of the Reston ebolavirus VP30 C-terminal domain

The crystal structure of the Reston ebolavirus VP30 C-terminal domain shows a rotated interface in comparison to the previous structure of the Zaire ebolavirus VP30 C-terminal domain.


Introduction
Ebolaviruses can cause hemorrhagic fever in humans, with a fatality rate as high as 90% (Burke et al., 1978). Four of the five ebolaviruses, including Ebola virus (EBOV; formerly Zaire ebolavirus), are found in Africa. However, Reston ebolavirus (RESTV) is uniquely Asian in origin. RESTV has been identified in bats and primates (Miranda et al., 1999;Rollin et al., 1999;Taniguchi et al., 2011), as well as swine (Barrette et al., 2009;Sayama et al., 2012). However, in humans RESTV appears to be nonpathogenic and transmits poorly for reasons that are not fully understood (Miranda et al., 1999).
Four proteins are essential for viral transcription: nucleoprotein (NP), VP30, VP35 and the RNA-dependent RNA polymerase (L) (Mü hlberger et al., 1999). VP30 appears unique to the filoviruses and is essential for rescuing recombinant ebolavirus (Enterlein et al., 2006;Theriault et al., 2004). VP30 allows the polymerase to read beyond a cis-RNA element in the NP mRNA 5 0 untranslated region (Weik et al., 2002) and to re-initiate transcription at gene junctions (Martínez et al., 2008). Furthermore, VP30 phosphorylation modulates the composition and function of the RNA synthesis machinery (Biedenkopf et al., 2013;Martinez et al., 2011). Both VP30 N-and C-terminal domains have been associated with oligomerization and nucleocapsid-association functions (Hartlieb et al., 2003(Hartlieb et al., , 2007. A basic cluster in the C-terminal domain (CTD) contributes to the association of VP30 with NP and is essential for transcription activation (Hartlieb et al., 2007). Previous crystallographic studies demonstrated that the EBOV VP30 CTD forms a dimer by donation of the C-terminal helix 7 to the neighboring monomer (Hartlieb et al., 2007). To date, the EBOV VP30 CTD is the only filovirus VP30 structure available. Here, we present the crystal structure of the VP30 CTD from Reston ebolavirus.
Cells were resuspended in lysis buffer [25 mM Tris-HCl pH 8.0, 200 mM NaCl, 50 mM arginine, 1 mM tris(2-carboxyethyl)phosphine (TCEP), 10 mM imidazole, 0.5% glycerol, 125 U Benzonase, 0.71 mg ml À1 lysozyme and one EDTA-free complete protease inhibitor tablet (Roche)] and lysed via sonication. The lysate was clarified by centrifugation and filtration. The protein was eluted from HiTrap Ni 2+ Chelating columns (GE Healthcare) with an imidazole step gradient. Pooled fractions were incubated for 18 h with Ulp-specific protease 1 (Ulp-1), intended to cleave the Smt domain from VP30. However, SDS-PAGE analysis showed that the His-Smt tag was unable to be removed (Fig. 1a). The protein was dialyzed against 25 mM Tris pH 8.0, 200 mM NaCl, 1 mM TCEP, 1% glycerol for 12 h and was concentrated to 18 mg ml À1 for crystallization. Macromolecule-production information is given in Table 1.

Crystallization
Crystals grew by vapor diffusion in sitting-drop trays in 10% PEG 6000, 100 mM HEPES pH 7.0 in 2-3 weeks using 0.4 ml protein solution and 0.4 ml reservoir solution at 16 C. The rod-shaped crystals were cryoprotected with 20% ethylene glycol prior to flash-cooling in liquid nitrogen. Details are given in Table 2.

Data collection and processing
Data were collected on beamline 5.0.1 at the Advanced Light Source with the detector set at a distance of 300 mm, with 0.5 oscillations and 5 s exposures for a total of 300 frames The data were reduced using XDS (Kabsch, 2010) and XSCALE. Data-collection and processing statistics are given in Table 3.

Structure solution, refinement and analysis
The structure was determined by molecular replacement using Phaser (McCoy et al., 2007) with the structure of EBOV VP30 CTD in monomeric form (Hartlieb et al., 2007; PDB entry 2i8b) as a search model. The structure was refined with REFMAC (Murshudov et al., 2011) and model building was performed with Coot (Emsley & Cowtan, 2004). The final structure was validated with MolProbity (Chen et al., 2010). Protein interfaces were analyzed using PISA (Krissinel & Henrick, 2007) and the shape-correlation statistic S C (Lawrence & Colman, 1993).

Results and discussion
In the crystal structure, there are two copies of the RESTV VP30 CTD in the asymmetric unit (Fig. 1b). The VP30 domains retain their Smt fusion domains from purification (Fig. 1a). One of the Smt domains helps facilitate crystal-packing interactions, whereas the second Smt domain is disordered. The two copies of RESTV VP30 in the asymmetric unit are nearly identical in structure and superimpose with an r.m.s.d. of 0.76 Å .
The fold of RESTV VP30 CTD is helical and closely resembles that of EBOV VP30, with an r.m.s.d. of 1.15-2.14 Å separating the C backbones of the EBOV and RESTV structures. Like EBOV VP30 (Hartlieb et al., 2007), RESTV VP30 forms a dimer by packing of its extended C-terminal helix 7 into a conserved hydrophobic face on the    neighboring monomer (Fig. 1c). Dimerization in both EBOV and RESTV VP30 is also facilitated by similar hydrophobic interactions between the globular domains and a conserved hydrogen bond between the side chain of Lys180 in helix 2 and the main chain of Cys251 in the linker between helices 6 and 7. Superposition of the RESTV and EBOV VP30 CTD structures shows that the RESTV VP30 CTD assembly is significantly rotated about the dimer interface compared with EBOV (Fig. 2a). The domain rotation lowers the buried surface area on each monomer (RESTV, 1630 Å 2 ; EBOV, 1900 Å 2 ) but maintains a similar surface complementarity (RESTV, S C = 0.69; EBOV, S C = 0.68), suggesting that the RESTV VP30 CTD dimer is a relevant conformation (Lawrence & Colman, 1993), and conservation of residues within the dimer interface suggests this conformation could also exist for EBOV.
It is difficult to discern the impact of the crystal-packing interactions on the EBOV or RESTV VP30 CTD conformations, as neither conformation is compatible with the packing of the other ebolavirus species. Additionally, the Smt domain in the RESTV VP30 CTD structure contributes to the crystal packing as well as making contacts with the opposite chain in the asymmetric unit (buries 545 Å 2 ). However, there is no apparent structural reason for the Smt domain to constrain the VP30 conformation and the Smt domain and VP30 CTD are not expected to form a stable complex in solution (Krissinel & Henrick, 2007).
The observed rotation in the RESTV VP30 dimer buries Arg179 and Lys180 (Fig. 2b), which are instead solvent-exposed in the EBOV structure. These residues are important in VP30 function and their mutation is detrimental to VP30 oligomerization, NP binding and transcription initiation (Hartlieb et al., 2007). The occlusion of Arg179 and Lys180 in the RESTV dimer interface suggests that mutations at these positions could disrupt VP30 dimerization.
EBOV and RESTV VP30 have 84% sequence identity within the CTD and conserve both the overall structure and hydrophobic interfaces (Fig. 1c). While we are unable to rule out the influence of the Smt on the conformation of the RESTV dimer, the observation of two dimer conformations in the available structures suggests that there may be inherent differences between the African and Asian viral species. Alternatively, the VP30 CTD could adopt multiple conformations in solution: structural changes in VP30 may reflect its role in modulation of RNA synthesis or another role in the ebolavirus lifecycle.