Genomic Characteristics And Biological Properties Of Duck Enteritis Virus And Its Attenuated Strain

Duck viral enteritis, also known as duck plague, is an acute, contagious and lethal disease of ducks, geese and swans. The disease causes high mortality in domestic and wild waterfowl, resulting in significant economic losses. The causative agent of this disease is duck enteritis virus (DEV). DEV taxonomically belongs to the species Anatid herpesvirus1, in the genus Mardivirus, subfamily Alphaherpesvirinae, family Herpesviridae. In this paper, to explain the molecular basis of attenuation of DEV via in vitro serial passages, the virulent reference strain of DEV (DEV CSC) in China was serially passaged in chicken embryo fibroblast (CEF), then the genomic characteristics and biological properties were compared between the attenuated mutant and its parental virus. A DEV rtcombinant strain with a deletion in both gl and gE genes was constructed to identify the role of DEV gl and gE genes in determining viral virulence and their affect on virus growth in the tissue cell. The aim of this paper is to advance our understanding the molecular basis for the pathogenesis and develop a gene-deleted vaccine to control duck plague.DEV CSC genome was sequenced on a pyrosequencing platform, the Genome Sequencer20(GS20) system (454Life Sciences). The DEV CSC genome was162,131bp long and contained78predicted open reading frames (ORFs), which encoded76putative proteins. The DEV CSC genome sequence was compared with the published genome sequences of DEV. There was99.99%homology of genome sequence between CSC and CHv isolated in Sichuan in2000. Twenty-one nucleotide substitutions (15non-synonymous and6synonymous) and52nucleotide deletions/insertions were detected. Except one3-bp insertion in UL41, all other deletions/insertions were in non-coding regions. Most of the non-synonymous substitutions (n=10) were in the5’end of genome. There was98.92%gene sequence identity between DEV CSC and DEV2085isolated in German in2005, mainly because of1,170-bp deletion at5’end of the unique long (UL) in DEV2085. Compared with DEV CSC genome, DEV K p63, which was attenuated by serial passages in chicken embryo, was4,040bp shorter in length mainly because of3,513-bp and528-bp deletions at the5’and3’ends of UL, respectively. At the nucleotide level,63of the76ORFs in the DEV CSC genome were100%identical to the ORFs in the DEV K p63genome. Two ORFs (UL56and US10) had frame-shift mutations in the C-terminal regions, while UL2had a176-aa deletion in DEV K p63.To gain a better understanding of the genetic changes needed for attenuation, DEV CSC was attenuated by serial passages in CEF, then the complete genome sequence of the80th passage was determined and compared with its parental virus. Moreover, the in vitro and in vivo properties were described. Cytopathogenicity was not discernible during the first4serial passages. However, at the5th passage DEV p5, minute and discernible plaques were observed after72hours postinoculation. Cytopathogenicity increased with virus passage. The average plaques size of DEV p25was significantly smaller than DEV p20(p<0.01), however, there was not significantly different from DEV p25to DEV p80. The complete genome sequence of DEV p80was determined and compared with its parental virus. DEV p80genome is160,328bp in length. An1,801-bp deletion from145,818to147,618nt was identified in the genome of DEV p80, which affected two genes encoding gl and gE. Moreover, there were12base substitutions, which led to6ami no acid conversions existed in open reading frames LORF4, UL51, UL9, UL7, UL4and US3. DEV p80and its parental virus were inoculated intramuscularly to ducks to compare their pahogenicity with respect to induction of clininal sign, rectal temperature, viremia, virus shedding, gross and histopathologic lesion, immunohistochemical and real time PCR detection. None of the ducks inoculated with DEV p80showed either obvious elevation of rectal temperature or any clinical sign.There was no obviously gross and histopathologic lesion in disgestive tract and lymphoid organs, and the copies of DEV DNA were lower than ducks inoculated with DEV CSC. Moreover, DEV p80protected ducks from lethal DEV challenge as soon as5days postinoculation.In other herpesviruses, gI and gE genes have benn reported to be important virulence factors. To identify the role of DEV gI and gE genes in determining viral virulence, a DEV recombinant DEV-△US78-GFP with a deletion in both gI and gE genes was constructed. The recombinant was assayed in vitro and in vivo. The average sizes of DEV-AUS78-GFP plaques were significantly smaller than its parent(p<0.001). On the other hand, the growth curves were different. The virus titer in both cell and supernatant were less80-fold than its parent, and the time to reach a peak was late24hours and36hours, respectively. Four weeks ducks inoculated with DEV-AUS78-GFP did not show clininal sign and gross lesion. These results suggest intact DEV gI and gE genes are important factors in efficient cell-cell viral spread in vitro and for expression of DEV virulence.Taken together, we finished the complete genome sequence of DEV CSC. To our knowledge, It was first reported that there was an1,801-bp deletion in both gI and gE genes in the mutant attenuated by serial passages in CEF. The attenuated mutant was avirulent in natural host ducks and protected ducks from lethal DEV challenge. We constructed a DEV recombinant with a deletion in both gI and gE genes, and comfirmed DEV gI and gE genes are important factors in efficient cell-cell viral spread in vitro and for expression of DEV virulence as shown for other herpesviruses. The present study is contributed to our understanding the molecular basis for the pathogenesis and development a gene-deleted vaccine to control duck plague.