The 3699 nt genome of olive latent virus 1 (OLV-1), described years ag
o from Southern Italy as a putative sobemovirus, was completely sequen
ced. OLV-1 genomic RNA was not polyadenylated and had a structure virt
ually identical to that of species of the Necrovirus rather than the S
obemovirus genus. Five open reading frames (ORFs) were identified, of
which the 5'-proximal encoded a 23 K protein and ended with an amber c
odon whose readthrough could yield a putative 82 K product. This polyp
eptide had extensive sequence similarity with polymerases of serotypes
A and D of tobacco necrosis necrovirus (TNV-A and TNV-D) and species
of the family Tombusviridae and related genera (Dianthovirus and Machl
omovirus). Two small ORFs followed, which encoded polypeptides of 8K a
nd 6K, respectively. The 6K product had extensive homology with the co
mparable 6 K protein of TNV-A and was also related to the 11 K protein
of shallot latent carlavirus, one of the ''triple block'' polypeptide
s involved in cell-to-cell virus movement. The 3'-proximal ORF was in
the same position as the coat protein (CP) cistron of necroviruses and
encoded a 30K product related to CP of both TNV-A and -D. Computer-as
sisted comparative analysis of structural and non-structural proteins
of OLV-1, TNV-A and TNV-D disclosed an overall distant relationship be
tween OLV-1 and TNV-D. OLV-1 genome appeared homologous to that of TNV
-A, but differences from TNV-A were the absence of the small ORF downs
tream of the CP cistron and in the low degree of sequence identity in
CP (39% aa identity). OLV-1 is serologically distantly related to TNV-
A and even more distantly related to TNV-D. We propose that OLV-1 is a
necrovirus species in its own right.