This is a formal proposal for the introduction of a new adenovirus genus in order to allocate several bovine, ovine, and bird adenoviruses of peculiar character which are presently outstanding in their respective (Mast- and Aviadenovirus) genera. First of all, it should be emphasized, that the establishment of a new (third) genus was proposed to the Study Group as early as 1969, but was refused until the availability of more convincing proofs. The findings presented below are not the solely or primary basis of the proposal, on the contrary, they are intended to support and justify the old observations. In the mid '60s, several "strange" bovine adenovirus (BAV) strains were isolated first by Bartha, Aldasy and Csontos in Hungary (BAV types 4, 5, & 8), later in other countries (BAV types 6 & 7). These BAVs were unusual, because they 1. required primary cultures of bovine testicle cells (and could not be grown on kidney epithelial cells, or cell lines as BAV types 1, 2 and 3); 2. exhibited different (elevated) heat resistance compared to the earlier existing BAVs; 3. provoked in cell cultures nuclear inclusion bodies of noticeably different morphology; and 4. did lack the common complement fixing antigen found in all other members of the Mastadenovirus genus. Bartha proposed the separation of these viruses into a separate group (they are now called subgroup 2 BAVs, versus subgroup 1 BAVs which are similar to most HAVs.). Later, the establishment of a new genus was also proposed, but the Study Group stated in its Second Report that further proofs are needed. We have started the comparison of the DNA of BAVs in 1983. Initially, restriction enzyme analysis was performed, and the method seemed to be feasible for type identification. It was noticed, however, that 1. subgroup 2 BAVs had app. 20% smaller genome size (28-30 kb); 2. SmaI enzyme had very few (subgroup 1) or no sites in BAVs, while EcoRI had many recognition sites on the genome of subgroup 2 BAVs. Later, by Southern blot hybridization studies, no homology was found between subgroup 1 and subgroup 2 BAVs. Subgroup 2 BAVs did not hybridize to any other members of the Mastadenovirus genus either. We have started the DNA sequencing of BAV-4 in 1992, and noticed that it had a very high (>60%) AT content. Recently, partial or complete DNA sequences of several animal adenoviruses became available. The analysis of these data shows, that all members of the subgroup 2 BAVs are genetically very closely related to each other. The quantitatively measurable genetic distance between these viruses and other mastadenoviruses is at least as large as that between mastadenoviruses and aviadenoviruses.
For the illustration of our findings, we attach the graphical results of the distance matrix analysis performed on several different genes. (The length of the branches indicates the phylogenetic distance between the different viruses.) The bootstrap values are given for 100 data sets, and demonstrate the high statistical significance of the found tree topology. It is especially notable, that calculating with 100 different samplings of the total gene the members of the proposed new genus were found to cluster together on all the 100 evolutionary trees. Parsimony analysis showed similar results. The DNA sequence of the complete genome of a "strange" ovine adenovirus isolate (called OAV287, but by serology and RE analysis being obviously related to BAV-7) was determined by Both and his group in Australia.
Originally, they intended to engineer this virus for foreign gene delivery/expression, but no E3 analogue region could be found at the conventional location (between the pVIII and fibre genes). By careful analysis of the complete sequence, compared to group C HAVs, several differences could be revealed in the genome organization of OAV287. The most remarkable finding was the difficulty in identification of the early regions. A real homologue of the E1A region could not be found, instead, a long ORF coding for a new structural protein was found on the complementary strand. Regions E1B and E4 could be hypothesized only based on very slight homology of several ORFs, but if these homologies are true, then the E3 analog region (hypothesized solely on its deletable nature) is located on the far right end of the genome, after the E4 region, and is transcribed leftward, like E4. Additional remarkable difference is the lack of the genes of protein V and protein IX. We have intentionally examined the respective regions of several other subgroup 2 BAVs (BAV-4 and BAV-6), and found each time very high-level sequence homology and identical genome organization with OAV287. DNA sequence homology detectable by Southern hybridization between EDS (egg drop syndrome) virus (= duck adenovirus 1) and BAVs was reported by Zakharchuk et al., in 1993. Although the authors have not concluded it, from the published autoradigraph it is obvious, that extensive DNA homology exist mainly between EDS virus and subgroup 2 BAVs. We have analyzed the DNA sequence (determined by Meehan et al.) of the protease gene of the EDS virus and found it to be most closely related to the protease gene of BAV-4, BAV-7 and OAV287. It has to be mentioned, that EDS virus was recorded as atypical aviadenovirus in the earlier reports of the Study Group, and described as possible "candidate for a separate genus". We have cloned and are now sequencing different parts of this virus, in order to determine its genomic organization. Compared to aviadenoviruses, an AT-rich coding strategy, accompanied by compressed genome (e.g. overlapping hexon and protease genes) could also be observed. Recently, the DNA sequence of the complete genome of the CELO (chicken embryo lethal orphan) virus (= fowl adenovirus 1) was also determined by Cotten and al., 1996. The genomic organization of this virus is also different from that of the well-known human, or typical mammalian adenoviruses. Upon the above described criteria, the adenoviruses recognized to date might clearly be classified into three well-defined groups. Since we believe, that the biased usage of AT rich codons (although for yet unknown reason) is a characteristic feature of the OAV287, EDS, and subgroup 2 bovine adenoviruses, we propose the name ATadenovirus for the new genus. Members of the new genus: BAV-4 BAV-5 BAV-6 BAV-7 BAV-8 OAV287 EDS virus Candidate member: HEV (turkey haemorrhagic enteritis virus) (mainly because of its high AT content)
