Figure 2.

Conservation of the ebolavirus GP N-trimer and design of peptide N-trimer mimics. (A) Schematic of the primary structure of ebolavirus GP2, indicating the fusion loop (brown), N-trimer (blue), C-peptide (yellow), and transmembrane domain (TM, purple), all shown approximately to scale. The sequences of the N-trimer region (residues 558–596) and the C-peptide region (residues 597–633) (Zaire ebolavirus species, representative Mayinga strain isolated in Zaire in 1976⁶⁰) contained in the peptides described in this study are indicated. In the C-peptide, cysteine 609, which is proposed to disulfide bond with GP1,⁷⁶ is mutated to alanine in our constructs (red). Below the Zaire N-trimer is an alignment of the sequences from the 4 additional ebolavirus species plus Marburgvirus and Cuevavirus filoviruses. Genbank accession codes are indicated (right). Conserved changes (score of 0 or higher in BLOSUM62 matrix⁷⁷) are highlighted in gray, nonconserved in cyan. Notably, 3/5 and 5/5 ebolavirus species are 100% identical in the N39 and N21 regions, respectively. The 2014 epidemic is caused by the Zaire ebolavirus species and is 100% identical in this region.³ *Reston and likely Cuevavirus (Lloviu virus) are not pathogenic to humans. (B) Schematics and sequences of the N-trimer mimics and their corresponding binding site mutants [eboIZN21 and eboIZN21(D2); eboIZN39IQ and eboIZN39IQ(D3)]. The designed coiled coils, IZ_m and IQ, are shown in gray, while the ebolavirus N-trimer is shown in blue. The a and d positions of the coiled-coil heptad repeats are indicated by a larger bolded font, including a stutter at the N-terminal end of the ebolavirus N-trimer as seen in the crystal structures,^27,28 where the coil is underwound, leading to an atypical 3-4-4-3 pattern (instead of the standard 3-4, or a-g, periodicity of a heptad repeat). The alanine residues along the C-peptide binding groove that are mutated to aspartate in the binding site mutants are shown (orange).