Alignment of HDV amino acid sequences from African (dFr-910, dFr-47, dFr-73, dFr-48, dFr-45, and dFr-644,), Italian (A20, genotype I), Japanese (-S, genotype IIA), Peruvian (-1, genotype III), and Taiwanese (TW-2b, genotype IIB) isolates. Dots indicate identities to dFr-910, whereas asterisks indicate positions conserved across all aligned sequences. Bold type indicates polymorphisms. Motifs corresponding to functional properties are boxed (NLS, nuclear localization signal; ARM, arginine-rich motifs; PKR, protein kinase R). The carboxy-terminal part of LHD constitutes a highly variable proline-rich domain corresponding to the packaging signal except for three conserved residues, a tryptophan, a cysteine, and a glutamine, corresponding, respectively, to the UAG-to-UGG editing site, the farnesylation box CXXX, and the C-terminal residue. Letters on the bottom line indicate when all (capital letters) or 50 to 99% (lowercase letters) of the 10 aligned sequences yielded identical secondary-structure consensus predictions (H and h, helix; C and c, coiled; E and e, beta-sheet).