Isolation and genome characterization of a novel duck tembusu virus with a 74 nucleotide insertion in the 3 non translated region

During investigations into the outbreak of duck Tembusu virus (DTMUV) infection in 2011 in China, a DTMUV strain (DTMUVAH2011) was isolated from the affected ducks. The length of the genome of the DTMUVAH2011 strain was found to be 11,064 nucleotides and to possess 10,278 nucleotides of one open reading frame (ORF), flanked by 94 nucleotides of the 5′ nontranslated region (NTR) and 692 nucleotides of the 3′ NTR. In comparison with five fully sequenced TMUV genomes, the genome of DTMUVAH2011 had a 74 nucleotide insertion in the 3′ NTR. Comparison of the DTMUVAH2011 fully deduced amino acid sequences with those of other Tembusu virus strains reported recently in China showed they had a highly conserved polyprotein precursor, sharing 98.9% amino acid identities, at least. The overall divergences of amino acid substitutions were randomly distributed among viral proteins except for the protein NS4B, the protein NS4B was unchanged. Knowledge of the biological characters of DTMUV and the potential role of the insertion in the 3′ NTR in RNA replication will be useful for further studies of the mechanisms of virus replication and pathogenesis

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=cavp20

Avian Pathology

ISSN: 0307-9457 (Print) 1465-3338 (Online) Journal homepage: https://www.tandfonline.com/loi/cavp20

Isolation and genome characterization of a novel duck Tembusu virus with a 74 nucleotide insertion

in the 3′ non-translated region

Jingman Wang, Weiting Liu, Gang Meng, Kangning Zhao, Jinyan Gu, Puyan Chen & Ruibing Cao

Chen & Ruibing Cao (2015) Isolation and genome characterization of a novel duck Tembusu virus with a 74 nucleotide insertion in the 3′ non-translated region, Avian Pathology, 44:2, 92-102, DOI: 10.1080/03079457.2015.1006167

To link to this article: https://doi.org/10.1080/03079457.2015.1006167

Published online: 05 Mar 2015

Submit your article to this journal

Article views: 1002

View related articles

View Crossmark data

Citing articles: 3 View citing articles

ORIGINAL ARTICLE

Isolation and genome characterization of a novel duck

Tembusu virus with a 74 nucleotide insertion in the

Jingman Wang, Weiting Liu, Gang Meng, Kangning Zhao, Jinyan Gu, Puyan Chen, and

Key Laboratory of Animal Diseases Diagnostic and Immunology, Ministry of Agriculture, College of Veterinary Medicine, Nanjing Agricultural University, Nanjing, China

During investigations into the outbreak of duck Tembusu virus (DTMUV) infection in 2011 in China, a DTMUV strain (AH2011) was isolated from the affected ducks The length of the genome of the DTMUV-AH2011 strain was found to be 11,064 nucleotides and to possess 10,278 nucleotides of one open reading frame

comparison with five fully sequenced TMUV genomes, the genome of DTMUV-AH2011 had a 74 nucleotide

other Tembusu virus strains reported recently in China showed they had a highly conserved polyprotein precursor, sharing 98.9% amino acid identities, at least The overall divergences of amino acid substitutions were randomly distributed among viral proteins except for the protein NS4B, the protein NS4B was unchanged Knowledge of the

useful for further studies of the mechanisms of virus replication and pathogenesis

Introduction

Since April 2010, a novel infectious agent emerged in

China and caused extensive epidemics in egg-laying ducks

The affected ducks (including Pekin duck, Cherry Valley

Pekin duck, Shaoxing duck, Jinyun duck, Longyan duck,

Jinding duck and Khaki-Campbell duck) showed severe

drops in egg production, declines in feed uptake, paralysis,

severe oophoritis and regression and mortality (Su et al.,

2011; Tang et al., 2012; Yun et al., 2012a) Hyperaemia,

haemorrhage, degeneration, distortion and lymphocyte

infiltration in ovaries and interstitial inflammation in the

portal areas of livers were the main pathological changes

observed consistently in almost all diseased ducks Based

on these changes and biological characters, the disease was

named duck haemorrhagic oophoritis (DHO), and the

emerging agent, a novel virus was shown to belong to a

new genotype of Tembusu virus (TMUV) and was

desig-nated duck Tembusu virus (DTMUV) (Cao et al., 2011b;

Yun et al.,2012b)

TMUV is a mosquito-borne flavivirus It is a member of

the Ntaya virus serogroup of the genus Flavivirus, family

Flaviviridae (Lindenbach & Rice,2003) TMUV was first

described in 1957 in Kuala Lumpur, Malaysia A

chick-origin TMUV strain, chick-originally designated as Sitiawan

virus, was reported later causing encephalitis and retarded

growth in broiler chicks (Kono et al., 2000) Clinical

investigations demonstrated the DTMUV could also affect

geese A report about TMUV-SDHS isolated from house sparrows showed a close relationship with the YY5 strain, and this latter strain was highly pathogenic to ducks This finding may play an important role in transmitting the virus among different species (Tang et al.,2012) In view of the epidemic characteristics of the DTMUV, as well as the hazard and losses that it caused in the duck industry, more attention to the virus is needed

Like other flaviviruses, DTMUVs are spherical, enveloped viruses of approximately 40–60 nm in diameter Flavivirus genome RNA has a type I cap structure at its 5′ terminus but lacks a polyadenylated tail at its 3′ terminus (Dong et al.,

2007) The genome consists of single-stranded, positive-sense RNA of approximately 11,000 nt with a long open reading frame (ORF) encoding one large putative polypro-tein The putative polyprotein is co- and post-translationally cleaved by viral and cellular proteases into three structural proteins [capsid (C), membrane (M) and envelope (E)] and seven nonstructural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B and NS5) that function in virus replication, proteolysis and virus maturation (Lindenbach & Rice,2003) The ORF is flanked by 5′ and 3′ non-translated regions (NTRs) Both NTRs form specific secondary stem-loop structures and functional domains that are essential for virus replication and assembly (Charlier et al., 2002; Markoff,

2003; Gritsun & Gould,2006)

We examined the genomic traits of DTMUV to provide the basic information for improving disease surveillance,

*To whom correspondence should be addressed: E-mail: crb@njau.edu.cn

(Received 20 March 2014; accepted 7 September 2014)

Vol 44, No 2, 92 –102, http://dx.doi.org/10.1080/03079457.2015.1006167

phylogenetic classification and development of sensitive

and specific detection methods based on conserved

sequences Taken together, our results showed that the

overall genome organization of DTMUV-AH2011 was

similar to that of known flaviviruses, but it also had its

unique characteristic

Materials and Methods

Clinical and pathological investigations We were notified by the duck

farmers of an unknown egg-drop disease Clinical examination and the

production records were checked for several affected flocks in Anhui

province in China since February 2011 To exclude avian influenza virus

infection, serum samples were collected from the affected flocks, and

antibody titres against H5 and H9 subtype avian influenza A virus were

tested with hemagglutination inhibition assays according to the standard

method Ducks in the affected flocks showing morbidity or death within 24

hours were taken to our laboratory for necropsy.

Virus isolation Specimens of spleen and ovary tissues with apparent

clinical signs from the affected flocks were collected and processed for virus

isolation Briefly, the tissues were homogenized in sterile

phosphate-buffered saline (pH 7.2) to give a 20% suspension (w/v) (Cao et al.,

2011b ; Yun et al., 2012a ) The samples were first tested by reverse

transcription polymerase chain reaction (RT-PCR) using primers ED3F/

ED3R (DTMUV-ED3F: ACTCCATGGATCAGGGTTTGAAGCTGAAA,

DTMUV-ED3R: CGGACTCGAGTGTGCTCCCACTTCTATG)

Subse-quently, positive samples were used for virus isolation The filtered

homogenates was inoculated into 9-day-old specific pathogen-free (SPF)

embryonated chicken eggs (0.2 mL/embryo) via the yolk sac The embryos

were carefully monitored When embryos died between 24 and 120 hours

post-inoculation, the allantoic fluids were harvested and stored at −20°C

until they were used for another round of inoculations or for molecular

analyses The isolated virus was cultured for at least three passages for

amplification and sequencing At the eighth passage, virus titre of the

allantoic fluid was evaluated by the 50% embyro lethal dose (ELD50) on

9-day-old SPF embryonated chicken eggs.

Virus titre of the cell adapted DTMUV For cell culture passage, the virus was cultured in baby hamster kidney (BHK-21) cell line BHK-21 cells were grown in Dulbecco ’s modified Eagle’s medium (Invitrogen, Grand Island, NY, USA) supplemented with 10% heat inactivated foetal calf serum, 100 U penicillin/mL and 100 µg streptomycin/mL at 37°C under 5%

CO 2 atmosphere Cell monolayers were infected with allantoic fluids from the fourth passage of the SPF embryonated chicken eggs (0.1 mL of a 1:10 dilution in medium), and the cells were observed daily to check for development of cytopathic effects (CPE) The cells were freeze –thawed three times after 72 h post-inoculation, cellular debris was removed by low-speed centrifugation and the supernatant fluid was re-inoculated onto fresh BHK-21 cells until the typical CPE of DTMUV appeared At the fifth passage, virus titre of the cell culture supernatants was determined by plaque assay as described by others (Okuno et al., 1984 ; Yun et al., 2009 ) Cells were then fixed with formaldehyde and stained with crystal violet to visualize the plaques The infected cells were examined by electron microscopy (data not shown).

RNA extraction and RT-PCR The viral RNA was extracted from 200 μL allantoic fluids from the fourth passage of the SPF embryonated chicken eggs by using the MiniBEST viral RNA Extraction kit (TaKaRa Biotech-nology, Dalian, China) according to the manufacturer ’s protocol The RNA was eluted in a final volume of 50 μL of elution buffer and stored at −80°C until further use RNA was used as the template for first-strand cDNA synthesis using PrimeScript® reverse transcriptase (TaKaRa Biotechnology, Dalian, China) with random hexanucleotide primer according to the manufacturer ’s instructions The cDNA was subsequently stored at −20°C.

In order to amplify the completed genome, 12 pairs of primers ( Table 1 ) were designed according to the BYDV and the YY5 strain genome sequences (GenBank access number: JF312912 and JF270480, respect-ively) PCR was carried out with 2~3 μL cDNA template Products of amplification were evaluated by electrophoresis in a 1% agarose gel DNA bands visualized by GoldView TM Nucleic acid staining and UV transillu-mination were excised from the agarose gel and purified using the Takara Agarose Gel DNA Purification kit according to manufacturer ’s protocol (TaKaRa Biotechnology, Dalian, China).

Sequencing Purified DNA products were ligated into the pMD18-T Easy vector according to manufacturer ’s instructions (TaKaRa Biotechnology),

Table 1 Primers used for synthesis and amplification of cDNA from DTMUV genome.

Primers name Polarity Sequence (direction 5′–3′) Gene positiona Product length (bp)

a Region in which the primer was designed and the number of the first nucleotide position in YY5 genome.

and ligated plasmids were introduced into DH5 α competent Escherichia

coli through chemical transformation Transformed cells were spread onto

LB/ampicillin plates Individual colonies were used to inoculate 5 mL LB/

ampicillin broth After overnight incubation at 37°C with shaking, plasmids

were extracted from 1 to 2 mL bacterio-fluids using the Takara MiniBEST

Plasmid Purification kit (TaKaRa Biotechnology) To verify the presence of

DNA inserts, extracted plasmids were digested with EcoRI and Sal I, and

examined by agarose gel electrophoresis.

The entire genome sequence of DTMUV-AH2011 was generated from 12

overlapping fragments Overlapping consensus sequences were assembled

using the SeqMan programs to generate contiguous full-genome sequences.

Analysis of nucleotide and deduced amino acid sequence identities were

performed by LASERGENE analysis software (DNASTAR Inc., Madison,

WI, USA) This software was also used to predict the entire ORF for the

polyprotein of TMUV as well as to determine the 5 ′ and 3′ NTR Deduced

amino acid sequences were analyzed and potential cleavage sites were

predicated by using the software program SignalP 3.0.

RNA secondary structure analysis The RNA secondary structural model

of the viral 3 ′ NTR of the DTMUV genomic RNA was constructed by the

mfold program, and the energy minimization program of Mfold website

server ( http://mfold.rna.albany.edu/?q=mfold/RNA-Folding-Form ) under

default folding conditions (37°C, 1 M NaCl, no divalent ions and no limit

on distance between paired bases) was used for constructing secondary

structural model (Zuker, 2003 ) To predict the RNA secondary structures

more accurately, the genomic RNA of two DTMUV strains

(DTMUV-AH2011 and YY5) was used The portable, windows-type RNAviz version

2.0 software ( http://rnaviz.sourceforge.net/ ) was used to draw or modify the

RNA secondary structure for the visualization or publication quality of

RNA secondary structure (De Rijk et al., 2003 ).

Multiple alignments and sequence analyses The complete genome of the

DTMUV-AH2011 strain (GenBank access number: KJ958533) was

ana-lyzed for searching ORF and translating to amino acid sequence using

Lasergene software package (DNASTAR) Percent sequences similarity/

divergence was calculated using the MegAlign program found in the

Lasergene analysis software (Thompson et al., 1997 ; Tamura et al., 2011 ).

The nucleotides and deduced amino acid sequences of the individual viral

proteins of the DTMUV-AH2011 strain were compared with those of 12

flaviviruses in Ntaya virus serogroup, including some TMUV strains

reported recently in China and some Bagaza virus strains ( Table 2 ).

Phylogenetic analysis To better understand the phylogenetic relationship

between DTMUV-AH2011 and other flaviviruses, full-length genome

sequences of other previously published flavivirus strains of varying

serogroup isolated from different locations and sources, and across a

number of years, were downloaded from GenBank, including sequences

from Murray Valley encephalitis virus (MVEV), West Nile virus (WNV),

Japanese encephalitis virus (JEV), Kunjin virus (KUNV), St Louis

encephalitis virus (SLEV), Dengue virus (DENV), Yellow fever virus

(YFV), Powassan virus (POWV), Langat virus (LANV), Louping ill virus

(LIV), Tick-borne encephalitis virus (TBEV), Culex flavivirus ( Table 3 ) and

some other Tembusu-related strains reported recently in China Neighbour-joining phylogenetic trees based on nucleotide sequences were constructed using Clustal program The robustness of phylogenetic construction was evaluated by bootstrapping using 1000 replicates In Ntaya virus serogroup,

to study the phylogeny of the TMUV E and NS5 protein, the nucleotides sequences of the E protein and NS5 protein of TMUV and other flaviviruses were aligned and then analyzed by Neighbour-joining method with 1000 bootstrap replicates in the Clustal W program to produce phylogenetic trees (Larkin et al., 2007 ).

Results Clinical and pathological investigations Since February

2011, many ducks developed the signs of DHO and died in several duck farms in Anhui province in China The disease spread fast around the duck-producing regions and resulted

in serious economic losses for the duck farms The affected ducks manifested clinical signs of heavy egg drop, retarded growth, high fever and loss of appetite As the disease progressed, some ducks exhibited an uncoordinated gait, and were reluctant or unable to walk, or even died At necropsy, severe ovarian haemorrhage, oophoritis and regression were consistently observed, and ruptured ovarian follicles were also found Enlarged spleens were occasion-ally found (Figure 1A, Band C) All clinical signs and the pathologic changes of the affected ducks were similar to the clinical cases described by Su et al (2011)

Virus isolation and virus titre of the cell adapted DTMUV To investigate whether the disease was caused

by DTMUV, specimens of spleen and ovary tissues from the affected flocks were collected, viral RNA was extracted from homogenates of the tissues, and tested by RT-PCR using primers specific for DTMUV All the suspect clinical specimens were positive when primers specific for DTMUV (ED3F/ED3R) were used, and nucleotide sequencing con-firmed that the infectious agent was the DTMUV RT-PCR with the primers ED3F/ED3R resulted in the amplification

of a 354 bp band, and some of the results are shown in

Figure 2

In order to isolate the DTMUV, one of the viral positive filtered homogenates was inoculated into 9-day-old SPF embryonated chicken eggs (0.2 mL/embryo) via the yolk sac All of the inoculated embryos died during 2–6 days post-inoculation with severely cutaneous haemorrhages The allantoic fluids did not have haemagglutinating activity The isolated virus was designated as DTMUV-AH2011

At the eighth passage, the 50% embyro lethal dose (ELD50)

Table 2 Detailed background information of 9 TMUV strains reported recently in China and three Bagaza virus strains used in this study.

of the allantoic fluid was 10−3.84/0.2mL In addition, most embryos died at the third day post-inoculation

For cell culture identification, the allantoic fluids of the fourth passage of the SPF embryonated chicken eggs were collected and serially passaged in BHK-21 monolayer cells

as described in Materials and Methods At the fifth passage, the virus produced a marked CPE in the BHK-21 cells at 72 hours after infection, focal CPE appeared with the cells rounding up and floating free from the surface of the flask (Figure 3A, B and C) The supernatant fluid had an infectivity titre of 2.8 × 105PFU/mL (Figure 4)

Full-length genome and deduced amino acid sequence analysis To further investigate the molecular characters of the DTMUV-AH2011 strain, the complete genome was amplified with several pairs of primers listed in Table 1

(Figure 5) On assembling these sequences, the length of the

Figure 1 Gross lesions of clinical samples from the infected

ducks (A) Mild haemorrhage of ovarian follicles and slightly

enlarged spleens with some necrotic foci in the early stage of

infection; (B) hyperaemia, severe haemorrhage and regression of

ovarian follicles; (C) enlarged spleens with large necrotic foci.

Table 3 Background information of 20 selected strains of flaviviruses used in this study.

Figure 2 Detection of the viral RNA from clinically infected

duck samples The theca folliculus and spleens tissues of the

infected duck collected from different regions was assessed using

RT-PCR with DTMUV E gene-speci fic primers (ED3F/ED3R) –:

negative control; 1, 2, 3, 4, 5, 6 and 7: samples from several duck

farms in Anhui province, these PCR products were sequenced and

their identities were con firmed; M: DNA marker DL 2000.

Figure 3 CPE of DTMUV-AH2011 in BHK-21 cells (A) Normal BHK-21 cell control; (B) Phase contrast photomicrographs of BHK-21 cells 60 h post-infection with DTMUV-AH2011, mild CPE occurred with the cells rounding up; (C) Phase contrast photo-micrographs of BHK-21 cells 72 h post-infection with DTMUV-AH2011, extensive CPE occurred with the cells rounding up and floating free from the surface of the flask The viral stock after five passages was used as the inoculum.

genome of the DTMUV-AH2011 strain was found to be 11,064 nucleotides and to possess 10,278 nucleotides of one ORF, flanked by 94 nucleotides of the 5′ NTR and 692 nucleotides of the 3′ NTR The ORF encoded 3425 amino acids (aa) including three structural and seven non-structural proteins

The phylogenetic analysis of the flavivirus genome showed that the isolated strain, DTMUV-AH2011, was genetically more similar to the Ntaya virus serogroup, especially some TMUV strains reported recently in China The nucleotide sequences of the DTMUV-AH2011 strain were compared with those of 12 flaviviruses in Ntaya virus serogroup, including some TMUV strains reported recently

in China and some Bagaza virus strains In comparison with five fully sequenced DTMUV genomes currently available,

we found that they had a high nucleotide sequence identity, all sharing approximately 97.7% nucleotide identities In the case of 3′ NTR, the DTMUV-AH2011 strain had a 74 nucleotide insertion in the 3′ NTR This insertion contained 69-nt duplicated nucleotides and was located at the 5′-proximal part of the 3′ NTR, which was located 69-nt downstream of the translation stop codon in the forward direction (Figure 6) To exclude whether the insertion was

an artefact, viral RNA was extracted from the homogenates

of corresponding tissues, and the 3′ NTR was amplified by RT-PCR and the amplicon was sequenced directly As the results described above, the insertion was also detected At

Figure 4 Virus titre and representative focus (or plaque)

morphology of the supernatant fluid of the fifth passage on

BHK-21 cells 1, 2, 3, 4 and 5: cell monolayers were infected with the

10-fold dilution from 1 × 10−1to 1 × 10−5; M: BHK-21 cell mono ‐

layers were mock infected (uninfected control) All wells of 6-well

plate were infected with 1ml inoculated fluid.

Figure 5 PCR products of 12 fragments of DTMUV-AH2011 1,

2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12: 12 fragments displayed in order of

primer; M: DNA marker DL 15000.

Figure 6 Organization of the genomic of the DTMUV (above), and (below) an expanded view of the multiple sequence alignment of the 3 ′ NTR of six available fully sequenced DTMUV strains including DTMUV-AH2011 sequenced in this study For sequence alignment, dots indicate the conserved nucleotide sequences in DTMUV strains, and hyphens indicate the missing nucleotide sequences The 74 nucleotide insertion of our isolated strain in this region is indicated by a box.

the same time, the 3′-distal part of the 3′ NTR exhibited

relatively high sequence identity among different strains of

the flaviviruses However, through RNA secondary

struc-ture prediction, we found this inserted sequence formed two

small stem-loop structures (Figure 7) The two small

stem-loop structures principally modified the structure of the

5′-proximal part of the 3′ NTR The newly inserted

sequence does not significantly interfere the RNA

second-ary structure of the 3′ NTR of the DTMUV-AH2011 strain,

especially, the 3′-distal part of the 3′ NTR

The nucleotides of the individual viral proteins of the

DTMUV-AH2011 strain were also compared with those

of other 12 flaviviruses All of the results are shown in

Table 4 Briefly, in comparison with nine TMUV strains,

the nucleotides of the individual viral proteins revealed

high levels of similarity (all nucleotide sequence identities

>96.0%) Nucleotide changes were scattered throughout

all the individual viral genes and no obvious difference

occurred between the viral proteins When compared with

some Bagaza virus strains in the Ntaya group, the NS3 gene

of DTMUV-AH2011 shared 74.9%, 75.0% and 75.1%

nucleotide identity with that of Spain-H2010, DakAr-B209

and 96363, respectively It was more conserved among the

viral protein encoding regions In contrast, the NS2A gene

had a relatively higher rate of nucleotide change In

addition, the sequence similarities were very high between

DTMUV-AH2011 and the other TMUV strains reported

recently in China

At the level of amino acids, the individual viral proteins

of the DTMUV-AH2011 strain were also compared with

those of the other 12 flaviviruses As shown inTable 5, they

had highly conserved polyprotein precursor and had high

amino acid sequence identity between DTMUV-AH2011

strain and nine TMUV strains reported recently in China

(amino acid identity ≥98.9%); however, the amino acid

sequence identities between DTMUV-AH2011 strain and

three Bagaza virus strains were relatively lower, ranging

from 80.7% (DakAr-B209) to 81.2% (Spain-H2010) for the

corresponding amino acid sequences In comparison with

the BYD-1 (the first reported duck-origin TMUV strain),

the ORF of DTMUV-AH2011 had a total of 30 amino acid

substitutions The overall divergence at the level of amino

acid substitution was 0.9%, these changes were randomly

distributed among the viral proteins except the protein

NS4B The protein NS4B was most conservative among all

of the currently available TMUV strains reported recently in

China, there was no amino acid substitution detected

However, compared with three Bagaza virus strains in the

Ntaya group, the protein NS4B was not most conservative

among the viral proteins, the most conservative protein was

Figure 7 Predicted secondary structure of 3′ NTR of TMUV.

The folding pattern was generated by the mfold program as descri ‐

bed in Materials and Methods (A) DTMUV-AH2011 strain;

(B) Tembusu YY5 strain.

protein NS3 (the amino acid identities among them being approximately 86.0%), this was consistent with the corre-sponding nucleotides sequence homology Besides, between DTMUV-AH2011 and BYD-1, proteins NS4A, NS3 and NS2A were largely conservative, with only 0.3%, 0.4% and 0.4% amino acid substitution, respectively In contrast, capsid and NS1 proteins had 4.2% and 1.4% amino acid divergence, respectively, with a relatively higher rate of amino acid change among the structural proteins and nonstructural proteins, respectively Overall, according to our multiple alignments and sequence analyses, E protein was relatively conservative among all the structural pro-teins, and protein NS4B (between DTMUV-AH2011 and other TMUV strains) or protein NS3 (between DTMUV-AH2011 and Bagaza virus strains) was relatively conservat-ive among all the nonstructural proteins However, capsid proteins were relatively variable among all the structural proteins, and NS1 protein (between DTMUV-AH2011 and other TMUV strains) or protein NS2A (between DTMUV-AH2011 and Bagaza virus strains) was relatively variable among all the nonstructural proteins

Multiple alignments and phylogenetic analysis To estab-lish the phylogenetic relationship between DTMUV-AH2011 and other flaviviruses, a phylogenetic tree was constructed using the complete genome sequences of DTMUV-AH2011 and 28 reference strains of flavivirus The DTMUV-AH2011 strain, which was isolated in 2011, together with other Tembusu strains reported recently in China were grouped into a branch divergent on the phylogenetic trees (Figure 8) All strains of Tembusu-related viruses were strongly clustered together and closely related to Bagaza virus Much evidence has suggested that the flavivirus E protein is the major determinant of viral neurovirulence and neuroinvasiveness (McMinn,1997; Cao

et al., 2011a) To study the phylogenetic relationship of Ntaya virus group, a phylogenetic tree was constructed using the reported structural gene (partial E gene) and non-structural gene (partial NS5 gene) of some viruses The E gene phylogenetic tree, produced by the NJ method and comprising a total of 28 virus strains, is shown inFigure 9A For comparison with the E gene phylogeny, phylogenetic analysis using cognate partial NS5 gene sequence informa-tion was carried out The NS5 phylogenetic tree produced here (Figure 9B) was virtually consistent with those produced in previous studies using the NS5 gene These analyses showed that the phylogenetic tree based on the partial E gene and partial NS5 gene corresponded well with the tree based on the full-length genome, with a minor difference

Discussion Since April 2010, a serious disease of acute infectious DHO

in layer ducks had quickly spread in Southeast China (Cao et al.,2011b; Su et al.,2011; Li et al.,2012) It had significantly affected the duck industry and resulted in serious economic losses in Southeast China The same clinical sign of egg drop syndrome was also found in several duck farms in Anhui province in China since February 2011 In this study, we attempted to isolate and identify the causative pathogen(s) responsible for this acute infectious disease The causative pathogen was isolated from the diseased ducks and designated as DTMUV-AH2011,

a newly emerged duck-origin TMUV (Yun et al., 2012a; Tang et al.,2013)

The gene sequence analysis showed that the

DTMUV-AH2011 shared high similarities to all recently reported

strains of TMUV except for the 3′ NTR The 3′ NTR was

found to be 692 nucleotides, containing a 74 nucleotide

insertion, which was a specific tandem repeat sequence in

DTMUV-AH2011 genome compared with all TMUV

strains for which the 3′-terminal sequences were available

Interestingly, this insertion was also detected in five other

samples collected from the same flocks of the same area

whereas this insertion was not detected in virus strains

detected in other regions in November 2012 suggesting that

the DTMUV/AH/2011 strain was specific for that region

Both NTRs of flavivirus form specific secondary

stem-loop structures and functional domains that are essential for

virus replication and assembly (Charlier et al., 2002;

Markoff,2003; Gritsun & Gould,2006) Previous research

showed that the presence of the 3′-proximal domains of the

JEV 3′ NTR was sufficient to maintain the replication

competency of the genomic RNA, whereas the presence of

the remaining 5′-proximal domains displayed an

enhancer-like function in promoting RNA replication efficiency

(Alvarez et al., 2005; Yun et al., 2009) Through RNA

secondary structure prediction, the inserted sequences

formed two small stem-loop structures, and the structure

of the proximal part of the 3′ NTR was modified The

newly inserted sequences did not significantly change the

RNA secondary structure of the 3′ NTR of the

DTMUV-AH2011 strain, especially in the 3′-distal part of the 3′

NTR, where most of the important elements in viral

translation, replication and assembly are concentrated

Mutational analysis of the 3′ NTR using recombinant

TBEV cDNA clones and a Kunjin virus replicon indicated

that the deletions in the 5′-proximal domains of 3′ NTR

may not affect the growth properties in vitro or virulence in mice (Khromykh & Westaway, 1997; Mandl et al., 1998) Therefore, only a relatively short region of the flavivirus 3′ NTR was required for genomic RNA replication, and nucleotide insertion or deletion at the 5′-proximal part of

3′ NTR would not cause a lethal mutant

There is a report that foot-and-mouth disease virus (FMDV) strains often differ in the presence or absence of certain deletions in the 5′ NTR, suggesting that recombina-tion events generating inserrecombina-tions or delerecombina-tions in the noncoding regions appear to be very frequent during evolution of FMDV (Escarmís et al., 1995) Under the natural positive selection pressure, the viral replicase may

be able to exercise the genome plasticity by deleting or adding genomic RNA stretches when structural and func-tional constraints allow (Hahn et al.,1987) The biological significance of the presence of the insertion in the 3′ NTR

in the DTMUV-AH2011 strain is unknown The occurrence

of insertions in DTMUV may be similar to that of FMDV However, the identification of the major determinants of viral replication within this region would require further investigation Whether the direct repeat sequences regulated DTMUV-AH2011 genome replication in connection with other cis-acting replication elements present within the genome is worth exploring

In the present study, the individual viral proteins of the DTMUV-AH2011 strain were also compared with other flaviviruses at the level of amino acid As described above, the protein NS4B was the most conservative among the viral proteins, there were no amino acid substitutions between them In particular, there was evidence to suggest that the central region of NS4B plays a role in the virulence phenotype of flaviviruses (Hanley et al., 2003;

Figure 8 Phylogenetic analysis of flaviviruses based on the full-length genome Phylogenetic analyses were performed by use of the MEGALIGN program (DNASTAR) and by using the Clustal alignment algorithm TMUV-AH2011 is surrounded by a box Bootstrap probabilities of each node were calculated using 1000 replicates Scale bars indicate the number of nucleotide substitutions per site Horizontal branch lengths are proportional to genetic distance while vertical branch lengths have no signi ficance Trees were constructed using the neighbour-joining method in Clustal W (Thompson et al., 1997 ).

McArthur et al., 2003; Wicker et al., 2006) Previous

studies have described mutations in the NS4B protein of

attenuated or passage-adapted mosquito-borne flaviviruses,

such as WNV suggesting that this protein plays an

important role in replication and pathogenesis

(Puig-Basagoiti et al., 2007; Wicker et al., 2012) However,

flavivirus NS4B was also found to inhibit the

interferon-signalling cascade at the level of nuclear STAT

phosphor-ylation, and it also had been implicated in modulation of

the stress-induced unfolded protein response

(Munoz-Jordan et al., 2003; Ambrose & Mackenzie, 2011) This

might be responsible for extensive epidemics of DTMUV

in egg-laying and breeder ducks since April 2010 in China,

with all strains being highly pathogenic to affected ducks

The function of the various non-structural proteins of

flavivirus in virus replication was more and more clear NS3

and NS5 form components of the viral replicase responsible

for replicating RNA Their sequences demonstrated a high degree of conservation among different flaviviruses (Hahn

et al.,1987) Whether amino acid changes of NS3 and NS5 protein among different DTMUV strains lead to function alteration needs further investigation E protein is the major structural protein of flavivirus, contains virus neutralizing epitopes, and plays an important role in virus receptor-binding, entry and fusion (Kuhn et al.,2002; Modis et al.,

2004) High identity of E protein between TMUV-AH2011 and other DTMUV strains may be due to low immune pressure at the early epidemic stage of the pathogen

In conclusion, the DTMUV-AH2011 is highly likely responsible for the acute infectious disease that occurred

in Anhui in China in 2011 The detailed sequence analysis between DTMUV-AH2011 and strains reported recently in China showed that these new DTMUV strains were highly conserved except that DTMUV-AH2011 has 74 nucleotides

Nucleotide Substitutions (x100) Bootstrap Trials = 1000, seed = 111

0 27.1

5 10

15 20

25

TMUV-FS.seq TMUV-JM.seq 84.7

TMUV-JS804.seq 56.1

TMUV-BYD-1.seq 27.3

TMUV-ZJ 407.seq TMUV-ZJ GH-2.seq 96.6

TMUV-YY5.seq 60.9

TMUV-SD.seq 39.1

NA

TMUV-ZJ-6.seq 33.5

DTMUV-AH2011.seq 99.7

DTMUV-JS2010.seq 100.0

HQ833331(Tembusu virus-Fengxian).seq 100.0

AF013409(Tembusu virus-THCAr).seq 100.0

AF013408(Tembusu virus-MM1775).seq 100.0

AB110487(Tembusu virus-4256 00).seq AB110489(Tembusu virus-Sitiawan).seq 74.6

AB110485(Tembusu virus-1665 96).seq 46.4

AB110486(Tembusu virus-3186 98).seq 99.8

87.3

AF013392(Ntaya virus).seq 100.0

AF013363(Bagaza virus-DakAr B209).seq AF013377(Israel turkey meningoencephali 100.0

51.4

JF737835(Marisma mosquito virus-HU566 0 100.0

JF707852(Mediterranean Culex-HU3679_06)

B

Nucleotide Substitutions (x100) Bootstrap Trials = 1000, seed = 111

0 109.5

20 40

60 80

100

AF372419(Edge Hill virus).seq AF372421(Saboya virus).seq 100.0

AF372413(Aroa virus).seq 99.6

AF372406(Alfuy virus).seq 28.5

AF372405(West Nile virus).seq 72.5

AF372409(Rocio virus).seq AF372414(Ilheus virus).seq 94.6

100.0

AF372407(Bagaza virus).seq AF372415(Israel turkey meningoencephali 100.0

99.4

AF372416(Ntaya virus).seq 100.0

AB110491(Tembusu virus-3186-98).seq AB110490(Tembusu virus-1665 96).seq 98.8

AB110493(Tembusu virus-Sitiawan).seq 85.6

92.6

AB110494(Tembusu virus-MM1775).seq 100.0

AB110495(Tembusu virus-ThCar105 92).seq 100.0

DTMUV-JS2010.seq 96.0

TMUV-ZJ 407.seq TMUV-ZJ GH-2.seq 99.3

TMUV-YY5.seq 92.7

TMUV-FS.seq TMUV-JM.seq 99.5

36.8

TMUV-SD.seq TMUV-JS804.seq 22.6

TMUV-ZJ-6.seq NA

20.6

DTMUV-AH2011.seq 20.6

TMUV-BYD-1.seq 26.9

JQ669731(Tembusu virus-HN1).seq 43.3

100.0

HE565481(JEV-HN08-Mos.YY-1).seq

A

Figure 9 Phylogenetic relationships of the flaviviruses in Ntaya virus serogroup The multiple sequence alignments were obtained by CLUSTAL alignment algorithm and trees were constructed by the neighbour-joining method The length of the branches represents the distance between sequence pairs Bootstrap support values, given as a percentage of 1000 replicates, are shown DTMUV-AH2011 and DTMUV-JS2010 are surrounded by black boxes The latter was also isolated by us (A) The phylogenetic relationships of the partial E gene

in Ntaya virus serogroup; (B) The phylogenetic relationships of the partial NS5 gene in Ntaya virus serogroup The GenBank accession number for the TMUV-JS2010 is under application.