- Email: [email protected]
Genetic reassortment of infectious bursal disease virus in nature
BBRC Biochemical and Biophysical Research Communications 350 (2006) 277–287 www.elsevier.com/locate/ybbrc
Genetic reassortment of infectious bursal disease virus in nature Yongwei Wei, Jianrong Li 1, Jiangtao Zheng, Hong Xu, Long Li, Lian Yu
*
Institute of Preventive Veterinary Medicine, College of Animal Science, Zhejiang University, Hangzhou 310029, PR China Received 14 August 2006 Available online 20 September 2006
Abstract Infectious bursal disease virus (IBDV), a double-stranded RNA virus, is a member of the Birnaviridae family. Four pathotypes of IBDV, attenuated, virulent, antigenic variant, and very virulent (vvIBDV), have been identified. We isolated and characterized the genomic reassortant IBDV strain ZJ2000 from severe field outbreaks in commercial flocks. Full-length genomic sequence analysis showed that ZJ2000 is a natural genetic reassortant virus with segments A and B derived from attenuated and very virulent strains of IBDV, respectively. ZJ2000 exhibited delayed replication kinetics as compared to attenuated strains. However, ZJ2000 was pathogenic to specific pathogen free (SPF) chickens and chicken embryos. Similar to a standard virulent IBDV strain, ZJ2000 caused 26.7% mortality, 100% morbidity, and severe bursal lesions at both gross and histopathological levels. Taken together, our data provide direct evidence for genetic reassortment of IBDV in nature, which may play an important role in the evolution, virulence, and host range of IBDV. Our data also suggest that VP2 is not the sole determinant of IBDV virulence, and that the RNA-dependent RNA polymerase protein, VP1, may play an important role in IBDV virulence. The discovery of reassortant viruses in nature suggests an additional risk of using live IBDV vaccines, which could act as genetic donors for genome reassortment. 2006 Elsevier Inc. All rights reserved. Keywords: IBDV; Reassortant virus in nature; Evolution; Virulence
Infectious bursal disease virus (IBDV) is the causative agent of a highly contagious immunosuppressive disease in young chickens. IBDV is a member of the family Birnaviridae and contains a double-stranded RNA genome consisting of two segments (A and B) within a non-enveloped icosahedral shell that is 60 nm in diameter [1]. Segment A (3.3 kb) contains two partially overlapping open reading frames (ORFs). The smaller ORF1 encodes the non-structural viral protein VP5 (17 kDa), which is not essential for viral replication or infection, but may play a role in pathogenicity [2–4]. The larger ORF2 encodes a 110 kDa polyprotein, which is cleaved by autoproteolysis to produce precursor pVP2 (48 kDa), VP4 (28 kDa), and VP3 (32 kDa) [5,6]. VP2 and VP3 are the major structural *
Corresponding author. Fax: +86 571 86971894. E-mail address: [email protected] (L. Yu). 1 Present address: Department of Microbiology and Molecular Genetics, Harvard Medical School, Boston, MA, USA. 0006-291X/$ - see front matter 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.09.040
proteins of the virion, and VP4 is the viral protease [6]. VP2 is the major host-protective immunogen responsible for inducing neutralizing antibodies and is related to antigenic variation, virulence, and apoptosis [7,8]. Genome segment B contains one ORF, encoding VP1 (97 kDa), which is the RNA-dependent RNA polymerase (RdRp) responsible for viral genome replication and RNA synthesis [9,10]. There are two different serotypes of IBDV (I and II). Serotype II virus is avirulent to turkeys and chickens [11]. Serotype I virus, which is pathogenic to chickens, can be further divided into attenuated, classical virulent, antigenic variant, and vvIBDV strains, based on pathotype. Classical virulent strains cause bursal inflammation and severe lymphoid necrosis in infected chickens, resulting in immunodeficiency and 20–30% mortality. Antigenic variant strains cause rapid atrophy of the bursa without inflammation, hemorrhage, or mortality [12,13]. Very virulent strains of IBDV, which first emerged in Europe in the early 1990s, can cause 60–100% mortality in chickens [14]. Usually,
278
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
pathogenic bursal-derived field strains cannot be directly adapted to cells without extensive passage in cell culture. Genome reassortment has been demonstrated for many animal RNA viruses such as influenza virus [15], Jatobal virus [16], Rift Valley fever virus [17], and Cucumber mosaic virus [18], as well as plant viruses such as Tobravirus [19], Cucumovirus [20], and Rice grassy stunt virus [21]. These studies indicated that genetic reassortment plays important roles in the evolution and emergence of viruses with altered disease potential or host range. It has been suggested that gene recombination and/or genome reassortment may occur in IBDV [14,22]. However, direct evidence was lacking, until recently (2006), Le Nouen et al. [23] described a reassortant strain which has a vvIBDV-like segment A but a classical IBDV-like segment B. In this report, we have identified a Chinese IBDV isolate, ZJ2000, which is a natural genomic reassortant virus whose genomic segment A is derived from an attenuated strain, and whose segment B is derived from a very virulent strain. This reassortant virus showed a virulent phenotype both in SPF chickens and chicken embryos.
onto 4 ml of a 40% sucrose cushion and ultracentrifuged at 22000g for 3 h at 4 C. The pelleted virus particles were resuspended in TEN-buffer (10 mM Tris/HCl, 100 mM NaCl, and 1 mM EDTA, pH 8.0). Virus identification Agar gel precipitation (AGP), virus-serum neutralization (VN), indirect immunofluorescence (IF), and electronic microscopy were used to further confirm that the isolate was an IBDV strain. The AGP assay was used to detect IBDV antigen in bursa tissue homogeneous suspension from sick chickens. VN was carried out by varying the serum levels with constant virus concentration on primary SPF CEFs as described previously [25]. Virus-infected CEF cells were subjected to IF using an antiIBDV polyclonal antibody and a FITC-conjugated anti-mouse IgG. Finally, the purified virus particles were negatively stained with 2% (w/v) uranyl acetate and observed by electronic microscopy. Virus growth curves in cells
Materials and methods
Viral growth was determined by tissue culture infectious dose50 (TCID50) on CEF cells. CEF cells were infected with each virus at MOI = 1. After 1 h adsorption, the inoculum was removed, cells were washed with MEM, fresh MEM (supplemented with 2% FBS) was added, and infected cells were incubated at 37 C. Aliquots of the cell culture fluid were removed at the indicated intervals and viral titer determined by TCID50. The cells were examined microscopically to record the CPE. TCID50 was calculated according to the Reed and Muench method. Attenuated strains JD1 and HZ2 served as control in the experiment.
SPF eggs and SPF chickens
Extraction of viral dsRNA
Specific pathogen free (SPF) eggs and chickens were purchased from the Experimental SPF Chicken Farm, Shandong Institute of Poultry Science, Shandong, PR China. Chickens were hatched and raised in a disease containment building, and then moved into isolators prior to inoculation. Flexible plastic isolators were supplied with filtered intake and exhaust air.
The original bursal tissue with clinical signs of IBD, as well as the virus particles purified from cell culture, were subjected to RNA extraction. Viral samples were digested by proteinase K (0.5 mg/ml) for 2 h at 50 C in the presence of 0.5% sodium dodecyl sulfate (SDS). The mixture was extracted twice with phenol/chloroform/isoamylalcohol (25:24:1) two times and once with phenol/chloroform (25:24). The viral dsRNA was precipitated with ethanol in the presence of sodium acetate (0.3 M final concentration, pH 5.6) and dissolved in RNase free water. Finally, the dsRNAs dried and resuspended in 30 ll RNase free water.
Virus strains In early 2000, severe field outbreaks were observed in commercial flocks in Zhejiang province, PR China. High mortalities of over 50% were recorded in the flocks. Pathological examination of sick chickens showed severe gross lesions in the bursa of Fabricius and spleen, which may be caused by IBDV. Therefore, the virus sample was named as ZJ2000. BC6/ 85, a standard classic virulent IBDV strain, was used as positive control and was purchased from China National Institute for Supervision of Veterinary Pharmaceuticals, Beijing, PR China. The attenuated IBDV strains, JD1 and HZ2, were used as a non-virulent control [24]. Virus isolation and purification Bursas were harvested from chickens with suspected IBDV infection from a field outbreak in Zhejiang Province, China. Bursas with clinical signs of IBDV were collected aseptically into RPMI 1640 medium (Invitrogen Life Technologies, Carlsbad, CA). Each bursa was minced using scissors and passed through a steel mesh to obtain a homogeneous suspension. After centrifugation at 1000g for 20 min at 4 C, the supernatant was collected. The supernatant was inoculated into primary chicken embryo fibroblast (CEF) cells which were maintained in Dulbecco’s modified essential medium (DMEM) supplemented with 10% fetal bovine serum (FBS) at 37 C in a humidified 5% CO2 incubator. Seventytwo hours post-inoculation, the supernatant was harvested and used for infection of fresh CEFs. Virus was plaque purified from the supernatant, and large-scale purification of virus was performed as described previously [24]. Briefly, ZJ2000 was grown in CEFs and harvested after 72 h postinfection. After freezing and thawing three times, the infected CEF cells were centrifuged at 5000g for 10 min. The clear supernatant was layered
Generation of full-length genomic segments A and B To obtain full-length genomic segments A and B of the ZJ2000 isolate, reverse transcription was performed with a primer complementary to the 3 0 end of the coding strand, and the M-MULV reverse transcriptase system (TakaRa Bio Inc., Otsu, Japan). Briefly, viral dsRNAs mixed with negative sense primer were incubated at 70 C for 5 min and cooled on ice for 2 min. Ten microliters of RT reaction buffer mixture containing 4 ll of 5· reaction buffer, 2 ll of 10 mM dNTP mix, 20 U of RNase inhibitor, and 40 U M-MULV reverse transcriptase was added, and the reaction mixture was incubated at 50 C for 1 h. Finally, the reaction mixture was heat inactivated at 70 C for 10 min and chilled on ice. Using the firststrand cDNA as a template, the cDNA fragment was amplified by high fidelity one shot LA-PCR (TakaRa Bio Inc.). Primers A1, 5 0 -CGGAATT CGGATACGATCGGTCTGACCCCGG-3 0 (located at 1–23) and A2, 5 0 -TAGGTACCGGGGACCCGCGAACGGATC-3 0 (located at 3242– 3261) were used to amplify the segment A. The amplifications for segment A were performed as follows: 94 C for 2 min; 30 cycles of 94 C for 15 s, 61 C for 30 s, 68 C for 3 min; and final elongation at 72 C for 10 min. Primers B1, 5 0 -GGATACGATGGGTCTGACCCT-3 0 (located at 1–22) and B2, 5 0 -GGGGCCCCCGCAGGCGAA-3 0 (located at 2808–2827) were used to amplify segment B. Amplification for segment B was performed using the following program: 94 C for 3 min; 30 cycles of 94 C for 15 s, 58 C for 30 s, 72 C for 3 min; and final elongation at 72 C for 10 min. The PCR products were gel purified from 1% agarose gel and ligated into pMD18-T vector. The ligation products were transformed into Escherichia coli DH5a cells and grown on LB agar plates containing
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287 ampicillin, X-gal, and IPTG. To obtain precise sequences, three independent positive clones were sequenced at Shanghai Shenggong Company (Shanghai, China). Genomewide sequence analysis of IBDV strains Sequence analysis was performed with the aid of DNAstar 5.0 software (DNASTAR, Inc., Madison, WI). Multiple sequence alignments were performed by DNAstar and phylogenetic tree analysis was performed by Jotun Hein method in MEGA3.1. IBDV strains used in the analysis were obtained from GenBank and their pathological phenotype is summarized in Fig. 2. Pathogenicity to SPF chicken embryos Ten-day-old SPF-embryonated eggs were inoculated with 0.4 ml of the bursa suspension via the chorioallantoic membrane (CAM) route and were incubated at 37 C for one week. The number of dead embryos was recorded every 6 h. Egg lethal dose50 (ELD50) of the virus isolate was determined using embryonated eggs by the Reed and Muench method [26]. Pathogenicity to SPF chickens Experiment 1. The pathogenicity studies were performed after identification of the reassortant ZJ2000 using plaque purified virus. Five-week-old specific pathogen free (SPF) chickens were randomly divided into two groups with 10 chickens per group. Chickens in the first group were inoculated with 105 ELD50 of strain ZJ2000. Chickens in the second group served as normal controls. All of the chickens were euthanized at seven days post-infection. Bursas and spleens were removed and weighed, and the mean organ/body weight (BW) ratio was determined according to Ismail and Saif [27] using the following formula: (organ weight in grams · 1000)/BW in grams. All of the bursal samples were fixed in 10% formaldehyde and embedded with wax. Thin sections were prepared for Hematoxylin–Eosin (HE) staining and histopathological examination of bursal lesions. Thin sections of bursa were also subjected to RNA extraction and RT-PCR as described above. Segments A and B were sequenced again to confirm that the disease was caused by ZJ2000. Experiment 2. Five-week-old SPF chickens were randomly divided into four groups with 15 chickens per group. Chickens in group 1 were inoculated with a standard virulent IBDV, BC6/85, as a positive control. Chickens in group 2 were inoculated with ZJ2000. Chickens in group 3 were inoculated with an attenuated IBDV strain, JD1, as a non-virulent control. Chickens in group 4 were inoculated with PBS and served as normal controls. Chickens in groups 1–3 were inoculated at a dosage of 2 · 104 ELD50. All of the chickens were euthanized at seven days postinfection. Other procedures were identical to those used in Experiment 1.
279
gen. To further identify the virus, the supernatant was inoculated into CEF cells. Interestingly, CPE was observed at 60 h post-inoculation of the first passage (Fig. 1a), indicating that this strain can directly adapt to cell culture. IBDV positive antiserum specifically neutralized the isolate as determined by a virus-serum neutralization assay in CEF cells. To further confirm that the isolate was IBDV, virus-infected CEF cells were subjected to an indirect immunofluorescence (IF) assay with a polyclonal anti-IBDV antibody. As shown in Fig. 1c, immunofluorescence signal was detected in virus-infected cells, but not in uninfected cells (Fig. 1d). After two passages in CEF cells, the virus was plaque purified and a large-scale purification of virus was prepared. Negative stain electron microscopy (EM) examination of the purified virus revealed non-enveloped icosahedral particles of approximately 58–60 nm in diameter (Fig. 1e). No other viral particles were found in the EM. Taken together, these results confirm that the isolate ZJ2000 is an IBDV strain. Cloning and sequencing of the full-length genomic segments A and B
The mean organ/body weight ratios of the challenged groups were compared with those of the control groups for statistical significance using analysis of variance followed by Fisher’s least significance difference.
The dsRNA was isolated from plaque purified virus particles or original sick bursa tissue by Proteinase K treatment in the presence of SDS. Two specific bands (segment A and segment B) of 3.3 kb and 2.9 kb were visualized in 1.0% agarose gels (data not shown). Using a primer complementary to the 3 0 -terminus of the genomic segments, the first-strand cDNA was synthesized using an M-MULV reverse transcriptase system. The full-length genomic cDNA of ZJ2000 was amplified by a long and optimal PCR, and cloned into the PMD18-T vector. Three independent clones of segment A and segment B from purified virus or bursa tissue were sequenced. No sequence differences were observed between the clones, and the sequence data obtained from plaque purified virus were completely identical to those from original bursa tissue containing the ZJ2000 isolate. The full-length segments A and B of ZJ2000 were 3259 and 2827 bp, respectively, and the sequences were deposited in the GenBank database (Accession Nos. AF321056 and DQ166818).
Results
Segment A of virulent ZJ2000 is highly homologous to those of attenuated strains
Statistical analysis
Isolation and identification of ZJ2000 Bursas with clinical signs of IBD from commercial flocks were isolated, homogenized in PBS, and the supernatant was collected after low-speed centrifugation. A strong immunoprecipitation line was formed between the homogeneous supernatant and standard IBDV positive antiserum as judged by the AGP assay (data not shown), indicating that the supernatant may contain IBDV anti-
Segment A of ZJ2000 contains 3259 bp and is highly homologous to that of attenuated strains. The 5 0 -NCR and 3 0 -NCR of ZJ2000 segment A contain 96 and 50 nucleotides, respectively, which are almost identical to those of attenuated strains JD1, HZ2, CEF94, P2, and CT. In the 3 0 -NCR, a nucleotide (C) deletion at position 3235 was found in the ZJ2000 strain. Within segment A, two partially overlapping ORFs encode VP5 and polyprotein VP2/4/ 3. The amino acid sequence of VP5 of ZJ2000 is completely
280
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
Fig. 1. Isolation and identification of IBDV strain ZJ2000. Phase contrast microscopy showing CPE in ZJ2000-infected cells at 60 h post-inoculation in the first passage (a) and uninfected CEF cells (b). Immunofluorescence detection of ZJ2000-infected CEF cells (c) and uninfected CEF cells (d) with an IBDV antibody. Negative stain electron microscopy of ZJ2000 virus particles (e).
Fig. 2. Unique amino acid substitutions in polyprotein and VP1 protein of vvIBDV and attenuated strains. V, virulent; VV, very virulent; AT, attenuated.
identical to that of attenuated strains HZ2, JD1, NB, CEF94, CT, and P2. However, VP5 of ZJ2000 is less related to vvIBDV strains, which have an N-terminal extension of four amino acids (MLSL) and two unique amino acid changes (R49G and W137R). The polyprotein of ZJ2000 is also highly related to the attenuated strains HZ2, NB, JD1, CEF94, P2, and CT, with amino acid identity between 97.7 and 99.0%. The polyprotein of ZJ2000 shares less homology with those of vvIBDV and variant strains, with amino acid identity between 89 and 95%. Fifteen amino acids, 222P, 242V, 253H, 256V, 279N, 284T, 299N, 330R, 451I, 541V, 680C, 715P, 751H, 980A, and 1005T, which are conserved in all attenuated strains, are also found in ZJ2000 (Fig. 2 and Supplementary Fig. 7a). In addition, six amino acids (positions 30G, 200N, 280S, 511R, 540R, and 952D) are unique to ZJ2000 (Supplementary Fig. 7a).
Segment B of virulent ZJ2000 is highly homologous to those of very virulent strains Notably, segment B of ZJ2000 contains 2827 bp and is highly homologous to that of very virulent strains. The 5 0 -NCR and 3 0 -NCR of ZJ2000 segment B contain 96 and 50 nucleotides, respectively, and are completely identical to vvIBDV strains D6948, HK46, UK611, and BD399, with the exception of one unique nucleotide substitution at position 83 (T to C) in the 5 0 -NCR of ZJ2000. As in most IBDV strains, VP1 of ZJ2000 contains 879 amino acids. VP1 of ZJ2000 showed the highest sequence homology (99.4%) to vvIBDV strain D6948 and was also highly homologous (98–99.3%) to other known vvIBDV strains including UK661, HK46, BD399, GZ96, OKYM, and Harbin-1. However, VP1 of ZJ2000 is less related to that of classic virulent and attenuated strains, with identities
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
281
between 96.9 and 97.6%. As shown in Fig. 2 and Supplementary Fig. 7b, eight amino acids (4V, 61I, 145T, 287A, 508K, 511S, 646S, and 687P), which are conserved in all vvIBDV strains, are also conserved in ZJ2000, strongly suggesting that VP1 of ZJ2000 is derived from a vvIBDV. There are five amino acids (positions 14M, 57S, 74D, 437P, and 584H) which are unique in ZJ2000 (Supplementary Fig. 7b). Phylogenetic analysis further confirms that ZJ2000 is a reassortant virus As shown in Fig. 3, the phylogenetic trees of 15 IBDV strains are split into two major branches: one encompassing attenuated and classic virulent strains (HZ2, NB, JD1, CEF94, D78, CEF94, CT, P2, IM, and GLS) and the other containing vvIBDV strains (D6948, UK661, HK46, OKYM, and BD399). Strikingly, the polyprotein of ZJ2000 formed a cluster with all attenuated and classical virulent strains (Fig. 3a). However, VP1 of ZJ2000 formed a cluster with all very virulent strains (Fig. 3b). These results further support the conclusion that segment A of ZJ2000 is derived from an attenuated strain and segment B is derived from a very virulent strain.
Fig. 4. Replication kinetics of ZJ2000 in CEF cells. CEF cells were infected with each virus at MOI = 1. Viral growth was determined by TCID50 on CEF cells as described in Materials and methods. Values are the average of three independent experiments.
24 h post-infection, ZJ2000 had a titer approximately 3 log lower than JD1 and HZ2 strains (Fig. 4). Pathogenicity of ZJ2000 in SPF chicken embryos
Replication kinetics of ZJ2000 in CEF cells To examine the efficiency of viral replication in vitro, CEF cells were infected with each virus and the virus titers were analyzed by TCID50. Fig. 3 shows the single step growth curve of each virus (expressed as log10 TCID50/ ml) in CEF cells at different time intervals post-infection. Clearly, ZJ2000 exhibited delayed replication kinetics as compared to attenuated JD1 and HZ2 strains (Fig. 4). At
The supernatant suspension containing the virus isolate ZJ2000 was inoculated into 10-day-old SPF chicken embryos at a dosage of 0.4 ml per embryo. The normal control group was injected with 0.4 ml PBS solution. In the first passage, the embryo mortality was 100% (10/10) at one week post-inoculation. The allantoic fluids of the dead embryos were harvested for further passage in embryos. In the following five
Fig. 3. Phylogenetic tree analysis of amino acid sequences of polyprotein (a) and VP1 protein (b). Phylogenetic trees were generated by the Jotun Hein method using MEGA3.1 software.
282
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
in the ZJ2000 group were significantly higher than in normal chickens (P < 0.05). Experiment 2: It has been established that mortality and morbidity caused by virulent IBDV is highly dependent on the challenge dose [8,27,29]. To determine whether the high mortality seen in Experiment 1 was due to an excessive dose of virus, a similar experiment was performed by infecting chickens with the lower dose of 2 · 104 ELD50. As the critical controls, chickens were also infected with the standard classic virulent strain BC6/85 (group 1) or the attenuated strain JD1 (group 3). The recommended challenge dose for this standard virulent strain is 2 · 104 ELD50 per chickens. At this dose, chickens in group 1 (BC6/85) and group 2 (ZJ2000) showed typical clinical signs of IBD as previously described. By day 5 post-infection, 4 out of 15 of the BC6/85 infected chickens were dead and 4 out of 15 of the ZJ2000 infected chickens were dead (Table 2). The clinical symptoms and mortality caused by ZJ2000 at 2 · 104 ELD50 were less severe than at the higher dose of 105 ELD50. All the bursae infected with BC6/85 and ZJ2000 exhibited gross lesions including inflammation, swelling, and edema (Fig. 5B, panels a and b). The bursa/ body and spleen/body ratio in both groups were significantly higher than normal chickens (P < 0.05). As a nonvirulent control, chickens in group 3, which received the attenuated IBDV strain JD1, did not show any clinical symptoms or gross lesions caused by IBDV (Fig. 5B, panel c). All of the chickens in this group survived, and the bursa/body and spleen/body ratios were not significantly different from normal chickens (P < 0.05). All of the bursal samples from Experiments 1 and 2 were fixed in 10% formaldehyde and embedded with wax. Thin sections were prepared for Hematoxylin–Eosin (HE) staining and histopathological examination. All of the bursae showed severe necrosis, atrophy, B lymphocyte depletion in follicles, aggregation of heterophils and macrophages, and loss of the outline of follicular architecture such that
Table 1 Pathogenicity of ZJ2000 in SPF chicken embryos Passagesa
ZJ2000 Control
P1
P2
P3
P4
P5
P6
10/10 0/10
10/10 0/10
10/10 0/10
8/10 0/10
10/10 0/10
9/10 0/10
a Ten-day-old SPF-embryonated eggs were inoculated with 0.4 ml of the suspension via the chorioallantoic membrane (CAM) route and were incubated at 37 C for a week. Number of dead chicken embryos at seven days post-inoculation during the six passages.
passages, 80–100% embryo mortality was recorded with a dose of 0.4 ml of the infected allantoic fluids (Table 1). The gross lesions of the infected embryos were hemorrhage on the legs, swelling, dwarfing, and curling. All of the chicken embryos in the normal control group survived. Pathogenicity of ZJ2000 in SPF chickens Experiment 1: The reassortant ZJ2000 provided a good model to study the pathogenicity of IBDV. SPF chickens were infected with ZJ2000 at a dosage of 105 ELD50 using plaque purified virus. Chickens inoculated with ZJ2000 showed typical clinical signs of IBD at day 2 post-challenge, including watery diarrhea, dehydration, depression, and death. By day 5 post-infection, 7 out of 10 chickens were dead (Table 2). The dead chickens were severely dehydrated and showed muscular hemorrhages at necropsy. At the gross pathological level, all the bursae from chickens infected with ZJ2000 exhibited significant inflammation, swelling, and edema, and were surrounded by thickened gelatinous serosa and exudates (Fig. 5A, panel a). On the bursa mucosal surface, the plicae were gray-white and edematous with occasional hemorrhage or bleeding. The spleen also showed significant inflammation and edema (Fig. 5A, panel c). The bursa/body and spleen/body ratios
Table 2 Pathogenicity of ZJ2000 in SPF chickens Exp.
Exp. 1 Exp. 2
a
Group
1 2 1 2 3 4
Virusa
ZJ2000 PBS BC6/85 ZJ2000 JD1 PBS
Dosage 5
1 · 10 / 2 · 104 2 · 104 2 · 104 /
Bursa lesions scoreb 0
1
2
3
0 10 0 0 5 15
0 0 0 0 10 0
0 0 0 4 0 0
10 0 15 11 0 0
B/BW ratioc
S/BW ratiod
Deade
Morbidityf
ADg
6.14* 3.45 6.83* 6.91* 5.22 4.91
3.01* 1.87 3.29* 2.95* 2.34 2.10
7/10 0/10 4/15 4/15 0/15 0/15
10/10 0/10 15/15 15/15 0/15 0/15
10/10 0/10 15/15 15/15 15/15 0/15
Chickens were inoculated with virus at the age of five weeks. Chickens in the control group were injected with PBS. The details of bursal lesion scores were described in our previous report [25,28]. Score 0, no damage; score 1, mild damage; score 2, moderate damage; score 3, severe damage. c B/BW, (average bursa in grams · 1000)/total body weight in grams. Values show the averages for 10 or 15 chickens. *Values are significantly different from the normal control (P < 0.05). d S/BW, (average spleen in grams · 1000)/total body weight in grams. Values show the averages for 10 or 15 chickens. e Number of dead chickens at 5 days post-inoculation. f Number of chickens with clinical symptom of IBD at 5 days post-inoculation. g AD, antigen detection. RT-PCR was performed to detect IBDV antigen from the bursa. b
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
283
Fig. 5. (A) Bursa and spleen samples from chickens inoculated with ZJ2000 at a dosage of 105 ELD50 using plaque purified virus. Severe gross lesions were observed in bursa from chickens inoculated with ZJ2000 (a), but not from normal chickens (b). Spleens from chickens inoculated with ZJ2000 (c) were significantly swollen as compared to those from normal chickens (d). (B) Bursa samples from chickens inoculated with BC6/85 (a), ZJ2000 (b), JD1 (c), and PBS (d) at a dosage 2 · 104 ELD50. Gross lesions were observed in bursa from chickens inoculated with BC6/85 (a) and ZJ2000 (b), but not from chickens inoculated with JD1 (c) and PBS (d). (C) Histopathological examination of bursa samples. (a–c) Bursal samples from chickens inoculated with ZJ2000; (d), normal bursa sample.
it is replaced by proliferated connective tissue and fibroplasias (Fig. 5C, panels a–c). None of the chickens in the normal control showed any sign of disease or bursa lesion. To confirm that the mortality, morbidity, and bursa lesions observed in these experiments were indeed caused by ZJ2000, total RNA was extracted from each bursa sample and genomic segments A and B were amplified by RTPCR as described above. Subsequent sequencing showed that all segment A sequences were classic IBDV-like and all segment B sequences were vvIBDV-like.
Discussion ZJ2000 is a novel natural reassortant virus Double-stranded RNA viruses have a high rate of mutation, recombination, and reassortment. Sequence analysis of limited region of the IBDV genome has suggested that genome recombination and/or reassortment may occur among different pathotypes of IBDV, but direct evidence is still lacking [14,22,30]. Most recently (2006), Le Nouen
284
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
et al. [23] described a reassortant strain which had a vvIBDV-like segment A but a classical IBDV-like segment B. However, their study was based solely on the ORF sequences of segments A and B, and did not distinguish between single gene recombination and genome reassortment. Based on the full-length genomic sequence (including NCRs), we have identified a natural reassortant IBDV strain (ZJ2000) which has a vvIBDV-like segment B but attenuated IBDV-like segment A. ZJ2000 is different from Le Nouen’s isolate, which represents a reciprocal reassortant strain. ZJ2000 is a natural genome reassortant isolate, but not a reassortant virus between attenuated and vvIBDV during the subsequent passage in tissue culture. First of all, it is well documented that vvIBDV (bursalderived field strains) cannot directly adapt to tissue culture [7,31–33]. Theoretically, vvIBDV cannot provide its genome segments as a reassortment donor during the cell culture because it was not able to replicate in cell culture. Second, ZJ2000 was able to directly adapt to cell culture even at the first passage in CEF cells. Subsequently, we plaque purified the virus at the first passage. This should limit the possible reassortment in cell culture to the minimum level. Finally, the sequences of segments A and B were determined from the original sick bursa tissue as well as multiple plaque purified viruses. All segment A sequences were identical and attenuated IBDV-like, and all segment B sequences were vvIBDV-like. This further confirmed that ZJ2000 is a natural reassortant IBDV strain, and not a mixed population of vvIBDV and attenuated IBDV strains, and also ruled out genome reassortment during passage in cell culture. Therefore, we conclude that ZJ2000 is a novel natural reassortant IBDV strain, which provides a useful tool to study the replication and pathogenicity of IBDV. VP2 is not the sole determinant of IBDV virulence It has been shown that VP2 is the major determinant of virulence, cell tropism, and pathogenic phenotype of IBDV. Amino acid substitution at positions 253 (Q to H), 279 (D to N), and 284 (A to T) of the VP2 protein was necessary and sufficient to adapt antigenic variant strains (E/Del and GLS) and vvIBDV strains (HK46 and UK661) to cell culture [7,31–33]. However, there is some evidence in the literature that VP2 is not the sole determinant of virulence. For example, replacement of the VP2 gene of an attenuated strain (CEF94) by a VP2 from a vvIBDV strain (D6948) resulted in a mosaic IBDV which caused increased damage to the bursa, but induced neither morbidity nor mortality in young chickens [34]. VP2 of ZJ2000 contains these characteristic amino acids (253H, 279N, and 284T). Consistent with this property, ZJ2000 was able to directly adapt to CEF cells. Extensive CPE was observed until 60 h post-inoculation at the first passage (Fig. 1a). The virulence of IBDV does not correlate with its replication in cell culture. For example, vvIBDV replicates
very fast in bursa, but cannot adapt to cell culture directly. In contrast, attenuated IBDV strain grows to very high titer in cell culture, but does not replicate as efficiently as vvIBDV in bursa. ZJ2000 was much more virulent than JD1 and HZ2, however, virus growth curve showed that ZJ2000 has delayed growth kinetics in cell culture as compared to these attenuated strains (Fig. 4), indicating that vvIBDV-derived segment B has a negative effect on viral replication in cell culture. It has also shown that VP1 of virulent IBDV carries the determinant for cell-specific viral replication [7]. Thus, VP1 may accelerate virus replication in vivo, but diminish the replication in vitro. However, the virulence of ZJ2000 in chickens indicates that VP2 is not the sole virulence determinant. VP1 protein plays an important role in IBDV virulence An increasing amount of evidence indicates that the RNA-dependent RNA polymerase (RdRp) protein contributes to virulence. Indeed, many viruses such as Measles virus [35], rinderpest virus [36], human parainfluenza virus type 2 [37], influenza virus [38], and vesicular stomatitis virus [39] have been attenuated by modification of their RdRp. IBDV VP1 protein is the RdRp responsible for viral genome replication and RNA synthesis. It is also linked to both ends of the dsRNA genome, as well as to the internal surface of the capsid. The formation of VP1–VP3 complexes plays a critical role in IBDV replication and the morphogenesis of IBDV particles. Thus, VP1 may contribute to virulence by altering its polymerase function and by affecting morphogenesis of IBDV particles. Direct evidence of VP1 involvement in viral replication and virulence was shown by Liu and Vakharia [40]. Using a reverse genetics system, they generated a reassortant virus rGLSBDB containing segment B from GLSBD (bursa-derived) and segment A from GLSTC (tissue culture-adapted). This virus exhibited delayed replication kinetics in CEF cells. However, it propagated efficiently in the BF and caused severe bursal lesions, whereas the tissue culture-adapted GLSTC virus replicated less efficiently and induced only mild bursal lesions. By constructing genomic reassortant viruses from attenuated CEF94 and vvIBDV D6948, Boot et al. [41] further confirmed that the enhanced virulence of vvIBDV is partly determined by its B segment. In this study, we showed that natural reassortant ZJ2000 is virulent although segment A is highly homologous to that of attenuated strains, providing another evidence that VP1 contributes to the virulence. Comparison of the VP1 sequences of classical and vvIBDV strains identified eight conserved amino acid differences (Fig. 2 and Supplementary Fig. 7b). Identification of specific amino acids in the VP1 protein that contribute to virulence remains an interesting problem. Multiple factors may contribute to the virulence of ZJ2000 In this study, we showed that ZJ2000 was pathogenic to SPF chickens and embryos. At a dosage of 105 ELD50 per
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
chicken, ZJ2000 caused 70% (7/10) mortality. At a lower dosage (2 · 104 ELD50), ZJ2000 cause 26.7% mortality (4/15), which was similar to the mortality caused by the standard virulent strain BC6/85. At both dosages, both ZJ2000 and BC6/85 caused severe bursa damage. This suggests that ZJ2000 is of comparable virulence to the standard classic virulent strain. Actually, ZJ2000 has also been used as a challenge virus to evaluate the efficacy of vaccine [28]. Multiple factors may be involved in the enhanced virulence of ZJ2000. First, the vvIBDV-derived VP1 protein may modulate the virulence of ZJ2000. The vvIBDV-derived VP1 protein of ZJ2000 has six additional unique amino acid substitutions, which may also contribute to the virulence of ZJ2000. Second, ZJ2000 has unique characteristics in segment A. For example, in the second smaller hydrophilic domain of VP2, ZJ2000 has an amino acid substitution at 280, which may influence the antigenicity and increase the virulence. In addition, there are two unique amino acid substitutions at positions 511 (L to R) and 540 (G to R) which are close to the VP2–VP4 cleavage site and may affect virulence through the processing of the polyprotein by the viral cleavage protease. Finally, other amino acid changes unique for ZJ2000 in VP3 and VP4 may also be involved in the increased virulence. For example, it has been shown that exchange of the C-terminus of VP3 from vvIBDV resulted in attenuation of the virus [42]. Possible model of genomic reassortment of IBDV in nature Fig. 6 shows a possible mechanism of genomic reassortment of IBDV in nature. It has been documented that live
285
vaccine has poor efficacy against vvIBDV [14,27,43]. Therefore, the outbreak of vvIBDV in live vaccine vaccinated flocks could lead to the co-infection of the two viruses, vvIBDV, and the attenuated live vaccine strain, in a single chicken. Replication of the two viral strains in the same target organ, the bursa, may lead to exchange of the double-stranded genomic RNA segments to generate new reassortant viruses. The discovery of a reassortant virus in nature suggests an additional risk of using live IBDV vaccines. Live vaccine could not only lead to virulence reversion due to continued passage in flocks, but could also act as a genetic donor for genome reassortment. Therefore, design of a safer and more effective vaccine is necessary to control IBDV. The rate of genomic reassortment event in nature To date, thousands of IBDV strains have been isolated worldwide in the GeneBank. However, almost all known strains were not reassortant virus, indicating the reassortment rate is low in nature. Also, it has not been documented that researchers can isolate a reassortant virus from vvIBDV and attenuated IBDV co-infected chickens in laboratory, raising the possibility that the occurring of reassortment is not simply due to co-infection of two types of IBDV. Many factors may contribute to the reassortment process. These factors may include: vaccine pressure (using different kind of vaccine such as inactivated vaccine and different attenuated live vaccines); environment (such as temperature, light, humidity, and breeding density) and immune system of chickens (such as response to vaccine
Fig. 6. Model of genomic reassortment of IBDV in nature.
286
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
and environment). Also, reassortment event may require virus-specific factors such as specific amino acid substitutions in viral proteins. Most recently, it was shown that non-structural proteins (NSP2) of Rotavirus, a doublestranded RNA virus, plays an important role in reassortment restriction between different Rotavirus groups [44]. Finally, reassortment may take a long time. Virus population may require transferring from chickens to chickens and generation to generation. In conclusion, we have identified a natural genetic reassortant IBDV, whose genomic segment A is derived from an attenuated strain, and whose segment B is derived from a very virulent strain. This reassortant virus and the classic virulent IBDV have an equivalent virulence in SPF chickens. Our results also provide further evidence that the VP1 protein plays an important role in determining the virulence of IBDV. Acknowledgments We thank Weijie Chen for help in collecting the bursa tissues from IBDV field outbreaks in Zhejiang Province. We thank Dr. Yaowei Huang for help in animal challenge experiment. We thank Professor Wei Jin for their kind help in histopathological examination. We thank Melina Agosto and Marco Morelli for critical reading of the manuscript. This research was supported by the grants from the National High Technology and Research and Development Program of China (No. 2004BA757C) and the Key Project of Zhejiang Province (No. 2003C22002). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bbrc. 2006. 09.040. References [1] H. Mu¨ller, C. Scholtissek, H. Becht, The genome of infectious bursal disease virus consists of two segments of double-stranded RNA, J. Virol. 31 (1979) 584–589. [2] E. Lombardo, A. Maraver, I. Espinosa, A. Fernandez-Arias, J.F. Rodriguez, VP5, the nonstructural polypeptide of infectious bursal disease virus, accumulates within the host plasma membrane and induces cell lysis, Virology 277 (2000) 345–357. [3] E. Mundt, J. Beyer, H. Mu¨ller, Identification of a novel viral protein in infectious bursal disease virus-infected cells, J. Gen. Virol. 76 (1995) 437–443. [4] K. Yao, V.N. Vakharia, Induction of apoptosis in vitro by the 17kDa nonstructural protein of infectious bursal disease virus: possible role in viral pathogenesis, Virology 285 (2001) 50–58. [5] A.A. Azad, S.A. Barrett, K.J. Fahey, The characterization and molecular cloning of the double-stranded RNA genome of an Australian strain of infectious bursal disease virus, Virology 143 (1985) 35–44. [6] A.B. Sanchez, J.F. Rodriguez, Proteolytic processing in infectious bursal disease virus: identification of the polyprotein cleavage sites by site-directed mutagenesis, Virology 262 (1999) 190–199.
[7] M. Brandt, K. Yao, M. Liu, R.A. Heckert, V.N. Vakharia, Molecular determinants of virulence, cell tropism, and pathogenic phenotype of Infectious Bursal Disease Virus, J. Virol. 75 (2001) 11974–11982. [8] K.J. Fahey, K. Erny, J. Crooks, A conformational immunogen on VP-2 of infectious bursal disease virus that induces virus-neutralizing antibodies that passively protect chickens, J. Gen. Virol. 70 (1989) 1473–1481. [9] U. Spies, H. Mu¨ller, H. Becht, Properties of RNA polymerase activity associated with infectious bursal disease virus and characterization of its reaction products, Virus Res. 8 (1987) 127–140. [10] U.I. von Einem, A.E. Gorbalenya, H. Schirrmeier, S.E. Behrens, T. Letzel, E. Mundt, VP1 of infectious bursal disease virus is an RNAdependent RNA polymerase, J. Gen. Virol. 85 (2004) 2221–2229. [11] N. Ismail, Y.M. Saif, P.D. Moorhead, Lack of pathogenicity of five serotype 2 infectious bursal disease viruses in chickens, Avian Dis. 32 (1988) 757–759. [12] D.J. Jackwood, R.J. Jackwood, Infectious bursal disease viruses: molecular differentiation of antigenic subtypes among serotype 1 viruses, Avian Dis. 38 (1994) 531–537. [13] D.B. Snyder, D.P. Lana, P.K. Savage, F.S. Yancey, S.A. Mengel, W.W. Marquardt, Differentiation of infectious bursal disease viruses directly from infected tissues with neutralizing monoclonal antibodies: evidence of a major antigenic shift in recent field isolates, Avian Dis. 32 (1988) 535–539. [14] M.D. Brown, M.A. Skinner, Coding sequences of both genome segments of a European ‘very virulent’ infectious bursal disease virus, Virus Res. 40 (1996) 1–15. [15] A. Wallensten, V.J. Munster, J. Elmberg, A.D. Osterhaus, R.A. Fouchier, B. Olsen, Multiple gene segment reassortment between Eurasian and American lineages of influenza A virus (H6N2) in Guillemot (Uria aalge), Arch. Virol. 150 (2005) 1685–1692. [16] M.F. Saeed, H. Wang, M. Suderman, D.W. Beasley, A. Travassos da Rosa, L. Li, R.E. Shope, R.B. Tesh, A.D. Barrett, Jatobal virus is a reassortant containing the small RNA of Oropouche virus, Virus Res. 77 (2001) 25–30. [17] M.D. Bowen, S.G. Trappier, A.J. Sanchez, R.F. Meyer, C.S. Goldsmith, S.R. Zaki, L.M. Dunster, C.J. Peters, T.G. Ksiazek, S.T. Nichol, RVF Task Force, A Reassortant Bunyavirus isolated from acute hemorrhagic fever cases in Kenya and Somalia, Virology 291 (2001) 185–190. [18] M.J. Roossinck, Evolutionary history of Cucumber mosaic virus deduced by phylogenetic analyses, J. Virol. 76 (2002) 3382–3387. [19] D.J. Robinson, W.D.O. Hamilton, B.D. Harrison, D.C. Baulcombe, Two anomalous tobravirus isolates: evidence for RNA recombination in nature, J. Gen. Virol. 68 (1987) 2551–2561. [20] P.S. White, F.J. Morales, M.J. Roossinck, Interspecific reassortment in the evolution of a cucumovirus, Virology 207 (1995) 334–337. [21] G.J. Miranda, O. Azzam, Y. Shirako, Comparison of nucleotide sequences between northern and southern Philippine isolates of rice grassy stunt virus indicates occurrence of natural genetic reassortment, Virology 266 (2000) 26–32. [22] L.L. Kong, A.R. Omar, M. Hair-Bejo, I. Aini, H.F. Seow, Sequence analysis of both genome segments of two very virulent infectious bursal disease virus field isolates with distinct pathogenicity, Arch. Virol. 149 (2004) 425–434. [23] C. Le Nouen, G. Rivallan, D. Toquin, P. Darlu, Y. Morin, V. Beven, C. de Boisseson, C. Cazaban, S. Comte, Y. Gardin, N. Eterradossi, Very virulent infectious bursal disease virus: reduced pathogenicity in a rare natural segment-B-reassorted isolate, J. Gen. Virol. 87 (2006) 209–216. [24] L. Yu, J.R. Li, Y.W. Huang, J. Dikki, R. Deng, Molecular characteristics of full-length genomic segment A of three infectious bursal disease viruses in China: two attenuated strains and one virulent field strain, Avian Dis. 45 (2001) 862–874. [25] J. Li, Y. Huang, X. Liang, M. Lu, L. Li, L. Yu, R. Deng, Plasmid DNA encoding antigens of infectious bursal disease viruses induce protective immune responses in chickens: factors influencing efficacy, Virus Res. 98 (2003) 63–74.
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287 [26] L. Reed, H. Muench, Simple method of determining fifty percent endpoints, Am. J. Hyg. 27 (1938) 493–497. [27] N. Ismail, Y.M. Saif, Immunogenicity of infectious bursal disease viruses in chickens, Avian Dis. 35 (1991) 460–469. [28] J. Li, X. Liang, Y. Huang, S. Meng, R. Xie, R. Deng, L. Yu, Enhancement of the immunogenicity of DNA vaccine against infectious bursal disease virus by co-delivery with plasmid encoding chicken interleukin 2, Virology 329 (2004) 89–100. [29] G.A. Abdel-Alim, Y.M. Saif, Immunogenicity and antigenicity of very virulent strains of infectious bursal disease viruses, Avian Dis. 45 (2001) 92–101. [30] T. Yamaguchi, M. Ogawa, M. Miyoshi, Y. Inoshima, H. Fukushi, K. Hirai, Sequence and phylogenetic analyses of highly virulent infectious bursal disease virus, Arch. Virol. 142 (1997) 1441–1458. [31] B.L. Lim, Y. Cao, T. Yu, C.W. Mo, Adaptation of very virulent infectious bursal disease virus to chicken embryonic fibroblasts by site-directed mutagenesis of residues 279 and 284 of viral coat protein VP2, J. Virol. 73 (1999) 2854–2862. [32] E. Mundt, Tissue culture infectivity of different strains of infectious bursal disease virus is determined by distinct amino acids in VP2, J. Gen. Virol. 80 (1999) 2067–2076. [33] A.A. van Loon, N. de Haas, I. Zeyda, E. Mundt, Alteration of amino acids in VP2 of very virulent infectious bursal disease virus results in tissue culture adaptation and attenuation in chickens, J. Gen. Virol. 83 (2002) 121–129. [34] H.J. Boot, A.A. ter Huurne, A.J. Hoekman, B.P. Peeters, A.L. Gielkens, Rescue of very virulent and mosaic infectious bursal disease virus from cloned cDNA: VP2 is not the sole determinant of the very virulent phenotype, J. Virol. 74 (2000) 6701–6711. [35] M. Takeda, A. Kato, F. Kobune, H. Sakata, Y. Li, T. Shioda, Y. Sakai, M. Asakawa, Y. Nagai, Measles virus attenuation associated with transcriptional impediment and a few amino acid changes in the polymerase and accessory proteins, J. Virol. 72 (1998) 8690–8696.
287
[36] D.D. Brown, B.K. Rima, I.V. Allen, M.D. Baron, A.C. Banyard, T. Barrett, W.P. Duprex, Rational attenuation of a morbillivirus by modulating the activity of the RNA-dependent RNA polymerase, J. Virol. 79 (2005) 14330–14338. [37] S.M. Nolan, S.R. Surman, E. Amaro-Carambot, P.L. Collin, B.R. Murphy, M.H. Skiadopoulos, Live-attenuated intranasal parainfluenza virus type 2 vaccine candidates developed by reverse genetics containing L polymerase protein mutations imported from heterologous paramyxoviruses, Vaccine 23 (2005) 4765–4774. [38] M. Quinlivan, D. Zamarin, A. Garcia-Sastre, A. Cullinane, T. Chambers, P. Palese, Attenuation of equine influenza viruses through truncations of the NS1 protein, J. Virol. 79 (2005) 8431–8439. [39] J. Li, E.C. Fontaine-Rodriguez, S.P. Whelan, Amino acid residues within conserved domain VI of the vesicular stomatitis virus large polymerase protein essential for mRNA cap methyltransferase activity, J. Virol. 79 (2005) 13373–13384. [40] M. Liu, V.N. Vakharia, VP1 protein of infectious bursal disease virus modulates the virulence in vivo, Virology 330 (2004) 62–73. [41] H.J. Boot, A.J. Hoekman, A.L. Gielkens, The enhanced virulence of very virulent infectious bursal disease virus is partly determined by its B-segment, Arch. Virol. 150 (2005) 137–144. [42] H.J. Boot, A.A. ter Huurne, A.J. Hoekman, J.M. Pol, A.L. Gielkens, B.P. Peeters, Exchange of the C-terminal part of VP3 from very virulent infectious bursal disease virus results in an attenuated virus with a unique antigenic structure, J. Virol. 76 (2002) 10346–10355. [43] G.A. Abdel-Alim, Y.M. Saif, Immunogenicity and antigenicity of very virulent strains of infectious bursal disease viruses, Avian Dis. 45 (2001) 92–101. [44] Z.F. Taraporewala, X. Jiang, R. Vasquez-Del Carpio, H. Jayaram, B.V. Prasad, J.T. Patton, Structure-function analysis of Rotavirus NSP2 octamer by using a novel complementation system, J. Virol. 80 (2006) 7984–7994.
Genetic reassortment of infectious bursal disease virus in nature Yongwei Wei, Jianrong Li 1, Jiangtao Zheng, Hong Xu, Long Li, Lian Yu
*
Institute of Preventive Veterinary Medicine, College of Animal Science, Zhejiang University, Hangzhou 310029, PR China Received 14 August 2006 Available online 20 September 2006
Abstract Infectious bursal disease virus (IBDV), a double-stranded RNA virus, is a member of the Birnaviridae family. Four pathotypes of IBDV, attenuated, virulent, antigenic variant, and very virulent (vvIBDV), have been identified. We isolated and characterized the genomic reassortant IBDV strain ZJ2000 from severe field outbreaks in commercial flocks. Full-length genomic sequence analysis showed that ZJ2000 is a natural genetic reassortant virus with segments A and B derived from attenuated and very virulent strains of IBDV, respectively. ZJ2000 exhibited delayed replication kinetics as compared to attenuated strains. However, ZJ2000 was pathogenic to specific pathogen free (SPF) chickens and chicken embryos. Similar to a standard virulent IBDV strain, ZJ2000 caused 26.7% mortality, 100% morbidity, and severe bursal lesions at both gross and histopathological levels. Taken together, our data provide direct evidence for genetic reassortment of IBDV in nature, which may play an important role in the evolution, virulence, and host range of IBDV. Our data also suggest that VP2 is not the sole determinant of IBDV virulence, and that the RNA-dependent RNA polymerase protein, VP1, may play an important role in IBDV virulence. The discovery of reassortant viruses in nature suggests an additional risk of using live IBDV vaccines, which could act as genetic donors for genome reassortment. 2006 Elsevier Inc. All rights reserved. Keywords: IBDV; Reassortant virus in nature; Evolution; Virulence
Infectious bursal disease virus (IBDV) is the causative agent of a highly contagious immunosuppressive disease in young chickens. IBDV is a member of the family Birnaviridae and contains a double-stranded RNA genome consisting of two segments (A and B) within a non-enveloped icosahedral shell that is 60 nm in diameter [1]. Segment A (3.3 kb) contains two partially overlapping open reading frames (ORFs). The smaller ORF1 encodes the non-structural viral protein VP5 (17 kDa), which is not essential for viral replication or infection, but may play a role in pathogenicity [2–4]. The larger ORF2 encodes a 110 kDa polyprotein, which is cleaved by autoproteolysis to produce precursor pVP2 (48 kDa), VP4 (28 kDa), and VP3 (32 kDa) [5,6]. VP2 and VP3 are the major structural *
Corresponding author. Fax: +86 571 86971894. E-mail address: [email protected] (L. Yu). 1 Present address: Department of Microbiology and Molecular Genetics, Harvard Medical School, Boston, MA, USA. 0006-291X/$ - see front matter 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.09.040
proteins of the virion, and VP4 is the viral protease [6]. VP2 is the major host-protective immunogen responsible for inducing neutralizing antibodies and is related to antigenic variation, virulence, and apoptosis [7,8]. Genome segment B contains one ORF, encoding VP1 (97 kDa), which is the RNA-dependent RNA polymerase (RdRp) responsible for viral genome replication and RNA synthesis [9,10]. There are two different serotypes of IBDV (I and II). Serotype II virus is avirulent to turkeys and chickens [11]. Serotype I virus, which is pathogenic to chickens, can be further divided into attenuated, classical virulent, antigenic variant, and vvIBDV strains, based on pathotype. Classical virulent strains cause bursal inflammation and severe lymphoid necrosis in infected chickens, resulting in immunodeficiency and 20–30% mortality. Antigenic variant strains cause rapid atrophy of the bursa without inflammation, hemorrhage, or mortality [12,13]. Very virulent strains of IBDV, which first emerged in Europe in the early 1990s, can cause 60–100% mortality in chickens [14]. Usually,
278
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
pathogenic bursal-derived field strains cannot be directly adapted to cells without extensive passage in cell culture. Genome reassortment has been demonstrated for many animal RNA viruses such as influenza virus [15], Jatobal virus [16], Rift Valley fever virus [17], and Cucumber mosaic virus [18], as well as plant viruses such as Tobravirus [19], Cucumovirus [20], and Rice grassy stunt virus [21]. These studies indicated that genetic reassortment plays important roles in the evolution and emergence of viruses with altered disease potential or host range. It has been suggested that gene recombination and/or genome reassortment may occur in IBDV [14,22]. However, direct evidence was lacking, until recently (2006), Le Nouen et al. [23] described a reassortant strain which has a vvIBDV-like segment A but a classical IBDV-like segment B. In this report, we have identified a Chinese IBDV isolate, ZJ2000, which is a natural genomic reassortant virus whose genomic segment A is derived from an attenuated strain, and whose segment B is derived from a very virulent strain. This reassortant virus showed a virulent phenotype both in SPF chickens and chicken embryos.
onto 4 ml of a 40% sucrose cushion and ultracentrifuged at 22000g for 3 h at 4 C. The pelleted virus particles were resuspended in TEN-buffer (10 mM Tris/HCl, 100 mM NaCl, and 1 mM EDTA, pH 8.0). Virus identification Agar gel precipitation (AGP), virus-serum neutralization (VN), indirect immunofluorescence (IF), and electronic microscopy were used to further confirm that the isolate was an IBDV strain. The AGP assay was used to detect IBDV antigen in bursa tissue homogeneous suspension from sick chickens. VN was carried out by varying the serum levels with constant virus concentration on primary SPF CEFs as described previously [25]. Virus-infected CEF cells were subjected to IF using an antiIBDV polyclonal antibody and a FITC-conjugated anti-mouse IgG. Finally, the purified virus particles were negatively stained with 2% (w/v) uranyl acetate and observed by electronic microscopy. Virus growth curves in cells
Materials and methods
Viral growth was determined by tissue culture infectious dose50 (TCID50) on CEF cells. CEF cells were infected with each virus at MOI = 1. After 1 h adsorption, the inoculum was removed, cells were washed with MEM, fresh MEM (supplemented with 2% FBS) was added, and infected cells were incubated at 37 C. Aliquots of the cell culture fluid were removed at the indicated intervals and viral titer determined by TCID50. The cells were examined microscopically to record the CPE. TCID50 was calculated according to the Reed and Muench method. Attenuated strains JD1 and HZ2 served as control in the experiment.
SPF eggs and SPF chickens
Extraction of viral dsRNA
Specific pathogen free (SPF) eggs and chickens were purchased from the Experimental SPF Chicken Farm, Shandong Institute of Poultry Science, Shandong, PR China. Chickens were hatched and raised in a disease containment building, and then moved into isolators prior to inoculation. Flexible plastic isolators were supplied with filtered intake and exhaust air.
The original bursal tissue with clinical signs of IBD, as well as the virus particles purified from cell culture, were subjected to RNA extraction. Viral samples were digested by proteinase K (0.5 mg/ml) for 2 h at 50 C in the presence of 0.5% sodium dodecyl sulfate (SDS). The mixture was extracted twice with phenol/chloroform/isoamylalcohol (25:24:1) two times and once with phenol/chloroform (25:24). The viral dsRNA was precipitated with ethanol in the presence of sodium acetate (0.3 M final concentration, pH 5.6) and dissolved in RNase free water. Finally, the dsRNAs dried and resuspended in 30 ll RNase free water.
Virus strains In early 2000, severe field outbreaks were observed in commercial flocks in Zhejiang province, PR China. High mortalities of over 50% were recorded in the flocks. Pathological examination of sick chickens showed severe gross lesions in the bursa of Fabricius and spleen, which may be caused by IBDV. Therefore, the virus sample was named as ZJ2000. BC6/ 85, a standard classic virulent IBDV strain, was used as positive control and was purchased from China National Institute for Supervision of Veterinary Pharmaceuticals, Beijing, PR China. The attenuated IBDV strains, JD1 and HZ2, were used as a non-virulent control [24]. Virus isolation and purification Bursas were harvested from chickens with suspected IBDV infection from a field outbreak in Zhejiang Province, China. Bursas with clinical signs of IBDV were collected aseptically into RPMI 1640 medium (Invitrogen Life Technologies, Carlsbad, CA). Each bursa was minced using scissors and passed through a steel mesh to obtain a homogeneous suspension. After centrifugation at 1000g for 20 min at 4 C, the supernatant was collected. The supernatant was inoculated into primary chicken embryo fibroblast (CEF) cells which were maintained in Dulbecco’s modified essential medium (DMEM) supplemented with 10% fetal bovine serum (FBS) at 37 C in a humidified 5% CO2 incubator. Seventytwo hours post-inoculation, the supernatant was harvested and used for infection of fresh CEFs. Virus was plaque purified from the supernatant, and large-scale purification of virus was performed as described previously [24]. Briefly, ZJ2000 was grown in CEFs and harvested after 72 h postinfection. After freezing and thawing three times, the infected CEF cells were centrifuged at 5000g for 10 min. The clear supernatant was layered
Generation of full-length genomic segments A and B To obtain full-length genomic segments A and B of the ZJ2000 isolate, reverse transcription was performed with a primer complementary to the 3 0 end of the coding strand, and the M-MULV reverse transcriptase system (TakaRa Bio Inc., Otsu, Japan). Briefly, viral dsRNAs mixed with negative sense primer were incubated at 70 C for 5 min and cooled on ice for 2 min. Ten microliters of RT reaction buffer mixture containing 4 ll of 5· reaction buffer, 2 ll of 10 mM dNTP mix, 20 U of RNase inhibitor, and 40 U M-MULV reverse transcriptase was added, and the reaction mixture was incubated at 50 C for 1 h. Finally, the reaction mixture was heat inactivated at 70 C for 10 min and chilled on ice. Using the firststrand cDNA as a template, the cDNA fragment was amplified by high fidelity one shot LA-PCR (TakaRa Bio Inc.). Primers A1, 5 0 -CGGAATT CGGATACGATCGGTCTGACCCCGG-3 0 (located at 1–23) and A2, 5 0 -TAGGTACCGGGGACCCGCGAACGGATC-3 0 (located at 3242– 3261) were used to amplify the segment A. The amplifications for segment A were performed as follows: 94 C for 2 min; 30 cycles of 94 C for 15 s, 61 C for 30 s, 68 C for 3 min; and final elongation at 72 C for 10 min. Primers B1, 5 0 -GGATACGATGGGTCTGACCCT-3 0 (located at 1–22) and B2, 5 0 -GGGGCCCCCGCAGGCGAA-3 0 (located at 2808–2827) were used to amplify segment B. Amplification for segment B was performed using the following program: 94 C for 3 min; 30 cycles of 94 C for 15 s, 58 C for 30 s, 72 C for 3 min; and final elongation at 72 C for 10 min. The PCR products were gel purified from 1% agarose gel and ligated into pMD18-T vector. The ligation products were transformed into Escherichia coli DH5a cells and grown on LB agar plates containing
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287 ampicillin, X-gal, and IPTG. To obtain precise sequences, three independent positive clones were sequenced at Shanghai Shenggong Company (Shanghai, China). Genomewide sequence analysis of IBDV strains Sequence analysis was performed with the aid of DNAstar 5.0 software (DNASTAR, Inc., Madison, WI). Multiple sequence alignments were performed by DNAstar and phylogenetic tree analysis was performed by Jotun Hein method in MEGA3.1. IBDV strains used in the analysis were obtained from GenBank and their pathological phenotype is summarized in Fig. 2. Pathogenicity to SPF chicken embryos Ten-day-old SPF-embryonated eggs were inoculated with 0.4 ml of the bursa suspension via the chorioallantoic membrane (CAM) route and were incubated at 37 C for one week. The number of dead embryos was recorded every 6 h. Egg lethal dose50 (ELD50) of the virus isolate was determined using embryonated eggs by the Reed and Muench method [26]. Pathogenicity to SPF chickens Experiment 1. The pathogenicity studies were performed after identification of the reassortant ZJ2000 using plaque purified virus. Five-week-old specific pathogen free (SPF) chickens were randomly divided into two groups with 10 chickens per group. Chickens in the first group were inoculated with 105 ELD50 of strain ZJ2000. Chickens in the second group served as normal controls. All of the chickens were euthanized at seven days post-infection. Bursas and spleens were removed and weighed, and the mean organ/body weight (BW) ratio was determined according to Ismail and Saif [27] using the following formula: (organ weight in grams · 1000)/BW in grams. All of the bursal samples were fixed in 10% formaldehyde and embedded with wax. Thin sections were prepared for Hematoxylin–Eosin (HE) staining and histopathological examination of bursal lesions. Thin sections of bursa were also subjected to RNA extraction and RT-PCR as described above. Segments A and B were sequenced again to confirm that the disease was caused by ZJ2000. Experiment 2. Five-week-old SPF chickens were randomly divided into four groups with 15 chickens per group. Chickens in group 1 were inoculated with a standard virulent IBDV, BC6/85, as a positive control. Chickens in group 2 were inoculated with ZJ2000. Chickens in group 3 were inoculated with an attenuated IBDV strain, JD1, as a non-virulent control. Chickens in group 4 were inoculated with PBS and served as normal controls. Chickens in groups 1–3 were inoculated at a dosage of 2 · 104 ELD50. All of the chickens were euthanized at seven days postinfection. Other procedures were identical to those used in Experiment 1.
279
gen. To further identify the virus, the supernatant was inoculated into CEF cells. Interestingly, CPE was observed at 60 h post-inoculation of the first passage (Fig. 1a), indicating that this strain can directly adapt to cell culture. IBDV positive antiserum specifically neutralized the isolate as determined by a virus-serum neutralization assay in CEF cells. To further confirm that the isolate was IBDV, virus-infected CEF cells were subjected to an indirect immunofluorescence (IF) assay with a polyclonal anti-IBDV antibody. As shown in Fig. 1c, immunofluorescence signal was detected in virus-infected cells, but not in uninfected cells (Fig. 1d). After two passages in CEF cells, the virus was plaque purified and a large-scale purification of virus was prepared. Negative stain electron microscopy (EM) examination of the purified virus revealed non-enveloped icosahedral particles of approximately 58–60 nm in diameter (Fig. 1e). No other viral particles were found in the EM. Taken together, these results confirm that the isolate ZJ2000 is an IBDV strain. Cloning and sequencing of the full-length genomic segments A and B
The mean organ/body weight ratios of the challenged groups were compared with those of the control groups for statistical significance using analysis of variance followed by Fisher’s least significance difference.
The dsRNA was isolated from plaque purified virus particles or original sick bursa tissue by Proteinase K treatment in the presence of SDS. Two specific bands (segment A and segment B) of 3.3 kb and 2.9 kb were visualized in 1.0% agarose gels (data not shown). Using a primer complementary to the 3 0 -terminus of the genomic segments, the first-strand cDNA was synthesized using an M-MULV reverse transcriptase system. The full-length genomic cDNA of ZJ2000 was amplified by a long and optimal PCR, and cloned into the PMD18-T vector. Three independent clones of segment A and segment B from purified virus or bursa tissue were sequenced. No sequence differences were observed between the clones, and the sequence data obtained from plaque purified virus were completely identical to those from original bursa tissue containing the ZJ2000 isolate. The full-length segments A and B of ZJ2000 were 3259 and 2827 bp, respectively, and the sequences were deposited in the GenBank database (Accession Nos. AF321056 and DQ166818).
Results
Segment A of virulent ZJ2000 is highly homologous to those of attenuated strains
Statistical analysis
Isolation and identification of ZJ2000 Bursas with clinical signs of IBD from commercial flocks were isolated, homogenized in PBS, and the supernatant was collected after low-speed centrifugation. A strong immunoprecipitation line was formed between the homogeneous supernatant and standard IBDV positive antiserum as judged by the AGP assay (data not shown), indicating that the supernatant may contain IBDV anti-
Segment A of ZJ2000 contains 3259 bp and is highly homologous to that of attenuated strains. The 5 0 -NCR and 3 0 -NCR of ZJ2000 segment A contain 96 and 50 nucleotides, respectively, which are almost identical to those of attenuated strains JD1, HZ2, CEF94, P2, and CT. In the 3 0 -NCR, a nucleotide (C) deletion at position 3235 was found in the ZJ2000 strain. Within segment A, two partially overlapping ORFs encode VP5 and polyprotein VP2/4/ 3. The amino acid sequence of VP5 of ZJ2000 is completely
280
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
Fig. 1. Isolation and identification of IBDV strain ZJ2000. Phase contrast microscopy showing CPE in ZJ2000-infected cells at 60 h post-inoculation in the first passage (a) and uninfected CEF cells (b). Immunofluorescence detection of ZJ2000-infected CEF cells (c) and uninfected CEF cells (d) with an IBDV antibody. Negative stain electron microscopy of ZJ2000 virus particles (e).
Fig. 2. Unique amino acid substitutions in polyprotein and VP1 protein of vvIBDV and attenuated strains. V, virulent; VV, very virulent; AT, attenuated.
identical to that of attenuated strains HZ2, JD1, NB, CEF94, CT, and P2. However, VP5 of ZJ2000 is less related to vvIBDV strains, which have an N-terminal extension of four amino acids (MLSL) and two unique amino acid changes (R49G and W137R). The polyprotein of ZJ2000 is also highly related to the attenuated strains HZ2, NB, JD1, CEF94, P2, and CT, with amino acid identity between 97.7 and 99.0%. The polyprotein of ZJ2000 shares less homology with those of vvIBDV and variant strains, with amino acid identity between 89 and 95%. Fifteen amino acids, 222P, 242V, 253H, 256V, 279N, 284T, 299N, 330R, 451I, 541V, 680C, 715P, 751H, 980A, and 1005T, which are conserved in all attenuated strains, are also found in ZJ2000 (Fig. 2 and Supplementary Fig. 7a). In addition, six amino acids (positions 30G, 200N, 280S, 511R, 540R, and 952D) are unique to ZJ2000 (Supplementary Fig. 7a).
Segment B of virulent ZJ2000 is highly homologous to those of very virulent strains Notably, segment B of ZJ2000 contains 2827 bp and is highly homologous to that of very virulent strains. The 5 0 -NCR and 3 0 -NCR of ZJ2000 segment B contain 96 and 50 nucleotides, respectively, and are completely identical to vvIBDV strains D6948, HK46, UK611, and BD399, with the exception of one unique nucleotide substitution at position 83 (T to C) in the 5 0 -NCR of ZJ2000. As in most IBDV strains, VP1 of ZJ2000 contains 879 amino acids. VP1 of ZJ2000 showed the highest sequence homology (99.4%) to vvIBDV strain D6948 and was also highly homologous (98–99.3%) to other known vvIBDV strains including UK661, HK46, BD399, GZ96, OKYM, and Harbin-1. However, VP1 of ZJ2000 is less related to that of classic virulent and attenuated strains, with identities
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
281
between 96.9 and 97.6%. As shown in Fig. 2 and Supplementary Fig. 7b, eight amino acids (4V, 61I, 145T, 287A, 508K, 511S, 646S, and 687P), which are conserved in all vvIBDV strains, are also conserved in ZJ2000, strongly suggesting that VP1 of ZJ2000 is derived from a vvIBDV. There are five amino acids (positions 14M, 57S, 74D, 437P, and 584H) which are unique in ZJ2000 (Supplementary Fig. 7b). Phylogenetic analysis further confirms that ZJ2000 is a reassortant virus As shown in Fig. 3, the phylogenetic trees of 15 IBDV strains are split into two major branches: one encompassing attenuated and classic virulent strains (HZ2, NB, JD1, CEF94, D78, CEF94, CT, P2, IM, and GLS) and the other containing vvIBDV strains (D6948, UK661, HK46, OKYM, and BD399). Strikingly, the polyprotein of ZJ2000 formed a cluster with all attenuated and classical virulent strains (Fig. 3a). However, VP1 of ZJ2000 formed a cluster with all very virulent strains (Fig. 3b). These results further support the conclusion that segment A of ZJ2000 is derived from an attenuated strain and segment B is derived from a very virulent strain.
Fig. 4. Replication kinetics of ZJ2000 in CEF cells. CEF cells were infected with each virus at MOI = 1. Viral growth was determined by TCID50 on CEF cells as described in Materials and methods. Values are the average of three independent experiments.
24 h post-infection, ZJ2000 had a titer approximately 3 log lower than JD1 and HZ2 strains (Fig. 4). Pathogenicity of ZJ2000 in SPF chicken embryos
Replication kinetics of ZJ2000 in CEF cells To examine the efficiency of viral replication in vitro, CEF cells were infected with each virus and the virus titers were analyzed by TCID50. Fig. 3 shows the single step growth curve of each virus (expressed as log10 TCID50/ ml) in CEF cells at different time intervals post-infection. Clearly, ZJ2000 exhibited delayed replication kinetics as compared to attenuated JD1 and HZ2 strains (Fig. 4). At
The supernatant suspension containing the virus isolate ZJ2000 was inoculated into 10-day-old SPF chicken embryos at a dosage of 0.4 ml per embryo. The normal control group was injected with 0.4 ml PBS solution. In the first passage, the embryo mortality was 100% (10/10) at one week post-inoculation. The allantoic fluids of the dead embryos were harvested for further passage in embryos. In the following five
Fig. 3. Phylogenetic tree analysis of amino acid sequences of polyprotein (a) and VP1 protein (b). Phylogenetic trees were generated by the Jotun Hein method using MEGA3.1 software.
282
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
in the ZJ2000 group were significantly higher than in normal chickens (P < 0.05). Experiment 2: It has been established that mortality and morbidity caused by virulent IBDV is highly dependent on the challenge dose [8,27,29]. To determine whether the high mortality seen in Experiment 1 was due to an excessive dose of virus, a similar experiment was performed by infecting chickens with the lower dose of 2 · 104 ELD50. As the critical controls, chickens were also infected with the standard classic virulent strain BC6/85 (group 1) or the attenuated strain JD1 (group 3). The recommended challenge dose for this standard virulent strain is 2 · 104 ELD50 per chickens. At this dose, chickens in group 1 (BC6/85) and group 2 (ZJ2000) showed typical clinical signs of IBD as previously described. By day 5 post-infection, 4 out of 15 of the BC6/85 infected chickens were dead and 4 out of 15 of the ZJ2000 infected chickens were dead (Table 2). The clinical symptoms and mortality caused by ZJ2000 at 2 · 104 ELD50 were less severe than at the higher dose of 105 ELD50. All the bursae infected with BC6/85 and ZJ2000 exhibited gross lesions including inflammation, swelling, and edema (Fig. 5B, panels a and b). The bursa/ body and spleen/body ratio in both groups were significantly higher than normal chickens (P < 0.05). As a nonvirulent control, chickens in group 3, which received the attenuated IBDV strain JD1, did not show any clinical symptoms or gross lesions caused by IBDV (Fig. 5B, panel c). All of the chickens in this group survived, and the bursa/body and spleen/body ratios were not significantly different from normal chickens (P < 0.05). All of the bursal samples from Experiments 1 and 2 were fixed in 10% formaldehyde and embedded with wax. Thin sections were prepared for Hematoxylin–Eosin (HE) staining and histopathological examination. All of the bursae showed severe necrosis, atrophy, B lymphocyte depletion in follicles, aggregation of heterophils and macrophages, and loss of the outline of follicular architecture such that
Table 1 Pathogenicity of ZJ2000 in SPF chicken embryos Passagesa
ZJ2000 Control
P1
P2
P3
P4
P5
P6
10/10 0/10
10/10 0/10
10/10 0/10
8/10 0/10
10/10 0/10
9/10 0/10
a Ten-day-old SPF-embryonated eggs were inoculated with 0.4 ml of the suspension via the chorioallantoic membrane (CAM) route and were incubated at 37 C for a week. Number of dead chicken embryos at seven days post-inoculation during the six passages.
passages, 80–100% embryo mortality was recorded with a dose of 0.4 ml of the infected allantoic fluids (Table 1). The gross lesions of the infected embryos were hemorrhage on the legs, swelling, dwarfing, and curling. All of the chicken embryos in the normal control group survived. Pathogenicity of ZJ2000 in SPF chickens Experiment 1: The reassortant ZJ2000 provided a good model to study the pathogenicity of IBDV. SPF chickens were infected with ZJ2000 at a dosage of 105 ELD50 using plaque purified virus. Chickens inoculated with ZJ2000 showed typical clinical signs of IBD at day 2 post-challenge, including watery diarrhea, dehydration, depression, and death. By day 5 post-infection, 7 out of 10 chickens were dead (Table 2). The dead chickens were severely dehydrated and showed muscular hemorrhages at necropsy. At the gross pathological level, all the bursae from chickens infected with ZJ2000 exhibited significant inflammation, swelling, and edema, and were surrounded by thickened gelatinous serosa and exudates (Fig. 5A, panel a). On the bursa mucosal surface, the plicae were gray-white and edematous with occasional hemorrhage or bleeding. The spleen also showed significant inflammation and edema (Fig. 5A, panel c). The bursa/body and spleen/body ratios
Table 2 Pathogenicity of ZJ2000 in SPF chickens Exp.
Exp. 1 Exp. 2
a
Group
1 2 1 2 3 4
Virusa
ZJ2000 PBS BC6/85 ZJ2000 JD1 PBS
Dosage 5
1 · 10 / 2 · 104 2 · 104 2 · 104 /
Bursa lesions scoreb 0
1
2
3
0 10 0 0 5 15
0 0 0 0 10 0
0 0 0 4 0 0
10 0 15 11 0 0
B/BW ratioc
S/BW ratiod
Deade
Morbidityf
ADg
6.14* 3.45 6.83* 6.91* 5.22 4.91
3.01* 1.87 3.29* 2.95* 2.34 2.10
7/10 0/10 4/15 4/15 0/15 0/15
10/10 0/10 15/15 15/15 0/15 0/15
10/10 0/10 15/15 15/15 15/15 0/15
Chickens were inoculated with virus at the age of five weeks. Chickens in the control group were injected with PBS. The details of bursal lesion scores were described in our previous report [25,28]. Score 0, no damage; score 1, mild damage; score 2, moderate damage; score 3, severe damage. c B/BW, (average bursa in grams · 1000)/total body weight in grams. Values show the averages for 10 or 15 chickens. *Values are significantly different from the normal control (P < 0.05). d S/BW, (average spleen in grams · 1000)/total body weight in grams. Values show the averages for 10 or 15 chickens. e Number of dead chickens at 5 days post-inoculation. f Number of chickens with clinical symptom of IBD at 5 days post-inoculation. g AD, antigen detection. RT-PCR was performed to detect IBDV antigen from the bursa. b
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
283
Fig. 5. (A) Bursa and spleen samples from chickens inoculated with ZJ2000 at a dosage of 105 ELD50 using plaque purified virus. Severe gross lesions were observed in bursa from chickens inoculated with ZJ2000 (a), but not from normal chickens (b). Spleens from chickens inoculated with ZJ2000 (c) were significantly swollen as compared to those from normal chickens (d). (B) Bursa samples from chickens inoculated with BC6/85 (a), ZJ2000 (b), JD1 (c), and PBS (d) at a dosage 2 · 104 ELD50. Gross lesions were observed in bursa from chickens inoculated with BC6/85 (a) and ZJ2000 (b), but not from chickens inoculated with JD1 (c) and PBS (d). (C) Histopathological examination of bursa samples. (a–c) Bursal samples from chickens inoculated with ZJ2000; (d), normal bursa sample.
it is replaced by proliferated connective tissue and fibroplasias (Fig. 5C, panels a–c). None of the chickens in the normal control showed any sign of disease or bursa lesion. To confirm that the mortality, morbidity, and bursa lesions observed in these experiments were indeed caused by ZJ2000, total RNA was extracted from each bursa sample and genomic segments A and B were amplified by RTPCR as described above. Subsequent sequencing showed that all segment A sequences were classic IBDV-like and all segment B sequences were vvIBDV-like.
Discussion ZJ2000 is a novel natural reassortant virus Double-stranded RNA viruses have a high rate of mutation, recombination, and reassortment. Sequence analysis of limited region of the IBDV genome has suggested that genome recombination and/or reassortment may occur among different pathotypes of IBDV, but direct evidence is still lacking [14,22,30]. Most recently (2006), Le Nouen
284
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
et al. [23] described a reassortant strain which had a vvIBDV-like segment A but a classical IBDV-like segment B. However, their study was based solely on the ORF sequences of segments A and B, and did not distinguish between single gene recombination and genome reassortment. Based on the full-length genomic sequence (including NCRs), we have identified a natural reassortant IBDV strain (ZJ2000) which has a vvIBDV-like segment B but attenuated IBDV-like segment A. ZJ2000 is different from Le Nouen’s isolate, which represents a reciprocal reassortant strain. ZJ2000 is a natural genome reassortant isolate, but not a reassortant virus between attenuated and vvIBDV during the subsequent passage in tissue culture. First of all, it is well documented that vvIBDV (bursalderived field strains) cannot directly adapt to tissue culture [7,31–33]. Theoretically, vvIBDV cannot provide its genome segments as a reassortment donor during the cell culture because it was not able to replicate in cell culture. Second, ZJ2000 was able to directly adapt to cell culture even at the first passage in CEF cells. Subsequently, we plaque purified the virus at the first passage. This should limit the possible reassortment in cell culture to the minimum level. Finally, the sequences of segments A and B were determined from the original sick bursa tissue as well as multiple plaque purified viruses. All segment A sequences were identical and attenuated IBDV-like, and all segment B sequences were vvIBDV-like. This further confirmed that ZJ2000 is a natural reassortant IBDV strain, and not a mixed population of vvIBDV and attenuated IBDV strains, and also ruled out genome reassortment during passage in cell culture. Therefore, we conclude that ZJ2000 is a novel natural reassortant IBDV strain, which provides a useful tool to study the replication and pathogenicity of IBDV. VP2 is not the sole determinant of IBDV virulence It has been shown that VP2 is the major determinant of virulence, cell tropism, and pathogenic phenotype of IBDV. Amino acid substitution at positions 253 (Q to H), 279 (D to N), and 284 (A to T) of the VP2 protein was necessary and sufficient to adapt antigenic variant strains (E/Del and GLS) and vvIBDV strains (HK46 and UK661) to cell culture [7,31–33]. However, there is some evidence in the literature that VP2 is not the sole determinant of virulence. For example, replacement of the VP2 gene of an attenuated strain (CEF94) by a VP2 from a vvIBDV strain (D6948) resulted in a mosaic IBDV which caused increased damage to the bursa, but induced neither morbidity nor mortality in young chickens [34]. VP2 of ZJ2000 contains these characteristic amino acids (253H, 279N, and 284T). Consistent with this property, ZJ2000 was able to directly adapt to CEF cells. Extensive CPE was observed until 60 h post-inoculation at the first passage (Fig. 1a). The virulence of IBDV does not correlate with its replication in cell culture. For example, vvIBDV replicates
very fast in bursa, but cannot adapt to cell culture directly. In contrast, attenuated IBDV strain grows to very high titer in cell culture, but does not replicate as efficiently as vvIBDV in bursa. ZJ2000 was much more virulent than JD1 and HZ2, however, virus growth curve showed that ZJ2000 has delayed growth kinetics in cell culture as compared to these attenuated strains (Fig. 4), indicating that vvIBDV-derived segment B has a negative effect on viral replication in cell culture. It has also shown that VP1 of virulent IBDV carries the determinant for cell-specific viral replication [7]. Thus, VP1 may accelerate virus replication in vivo, but diminish the replication in vitro. However, the virulence of ZJ2000 in chickens indicates that VP2 is not the sole virulence determinant. VP1 protein plays an important role in IBDV virulence An increasing amount of evidence indicates that the RNA-dependent RNA polymerase (RdRp) protein contributes to virulence. Indeed, many viruses such as Measles virus [35], rinderpest virus [36], human parainfluenza virus type 2 [37], influenza virus [38], and vesicular stomatitis virus [39] have been attenuated by modification of their RdRp. IBDV VP1 protein is the RdRp responsible for viral genome replication and RNA synthesis. It is also linked to both ends of the dsRNA genome, as well as to the internal surface of the capsid. The formation of VP1–VP3 complexes plays a critical role in IBDV replication and the morphogenesis of IBDV particles. Thus, VP1 may contribute to virulence by altering its polymerase function and by affecting morphogenesis of IBDV particles. Direct evidence of VP1 involvement in viral replication and virulence was shown by Liu and Vakharia [40]. Using a reverse genetics system, they generated a reassortant virus rGLSBDB containing segment B from GLSBD (bursa-derived) and segment A from GLSTC (tissue culture-adapted). This virus exhibited delayed replication kinetics in CEF cells. However, it propagated efficiently in the BF and caused severe bursal lesions, whereas the tissue culture-adapted GLSTC virus replicated less efficiently and induced only mild bursal lesions. By constructing genomic reassortant viruses from attenuated CEF94 and vvIBDV D6948, Boot et al. [41] further confirmed that the enhanced virulence of vvIBDV is partly determined by its B segment. In this study, we showed that natural reassortant ZJ2000 is virulent although segment A is highly homologous to that of attenuated strains, providing another evidence that VP1 contributes to the virulence. Comparison of the VP1 sequences of classical and vvIBDV strains identified eight conserved amino acid differences (Fig. 2 and Supplementary Fig. 7b). Identification of specific amino acids in the VP1 protein that contribute to virulence remains an interesting problem. Multiple factors may contribute to the virulence of ZJ2000 In this study, we showed that ZJ2000 was pathogenic to SPF chickens and embryos. At a dosage of 105 ELD50 per
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
chicken, ZJ2000 caused 70% (7/10) mortality. At a lower dosage (2 · 104 ELD50), ZJ2000 cause 26.7% mortality (4/15), which was similar to the mortality caused by the standard virulent strain BC6/85. At both dosages, both ZJ2000 and BC6/85 caused severe bursa damage. This suggests that ZJ2000 is of comparable virulence to the standard classic virulent strain. Actually, ZJ2000 has also been used as a challenge virus to evaluate the efficacy of vaccine [28]. Multiple factors may be involved in the enhanced virulence of ZJ2000. First, the vvIBDV-derived VP1 protein may modulate the virulence of ZJ2000. The vvIBDV-derived VP1 protein of ZJ2000 has six additional unique amino acid substitutions, which may also contribute to the virulence of ZJ2000. Second, ZJ2000 has unique characteristics in segment A. For example, in the second smaller hydrophilic domain of VP2, ZJ2000 has an amino acid substitution at 280, which may influence the antigenicity and increase the virulence. In addition, there are two unique amino acid substitutions at positions 511 (L to R) and 540 (G to R) which are close to the VP2–VP4 cleavage site and may affect virulence through the processing of the polyprotein by the viral cleavage protease. Finally, other amino acid changes unique for ZJ2000 in VP3 and VP4 may also be involved in the increased virulence. For example, it has been shown that exchange of the C-terminus of VP3 from vvIBDV resulted in attenuation of the virus [42]. Possible model of genomic reassortment of IBDV in nature Fig. 6 shows a possible mechanism of genomic reassortment of IBDV in nature. It has been documented that live
285
vaccine has poor efficacy against vvIBDV [14,27,43]. Therefore, the outbreak of vvIBDV in live vaccine vaccinated flocks could lead to the co-infection of the two viruses, vvIBDV, and the attenuated live vaccine strain, in a single chicken. Replication of the two viral strains in the same target organ, the bursa, may lead to exchange of the double-stranded genomic RNA segments to generate new reassortant viruses. The discovery of a reassortant virus in nature suggests an additional risk of using live IBDV vaccines. Live vaccine could not only lead to virulence reversion due to continued passage in flocks, but could also act as a genetic donor for genome reassortment. Therefore, design of a safer and more effective vaccine is necessary to control IBDV. The rate of genomic reassortment event in nature To date, thousands of IBDV strains have been isolated worldwide in the GeneBank. However, almost all known strains were not reassortant virus, indicating the reassortment rate is low in nature. Also, it has not been documented that researchers can isolate a reassortant virus from vvIBDV and attenuated IBDV co-infected chickens in laboratory, raising the possibility that the occurring of reassortment is not simply due to co-infection of two types of IBDV. Many factors may contribute to the reassortment process. These factors may include: vaccine pressure (using different kind of vaccine such as inactivated vaccine and different attenuated live vaccines); environment (such as temperature, light, humidity, and breeding density) and immune system of chickens (such as response to vaccine
Fig. 6. Model of genomic reassortment of IBDV in nature.
286
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287
and environment). Also, reassortment event may require virus-specific factors such as specific amino acid substitutions in viral proteins. Most recently, it was shown that non-structural proteins (NSP2) of Rotavirus, a doublestranded RNA virus, plays an important role in reassortment restriction between different Rotavirus groups [44]. Finally, reassortment may take a long time. Virus population may require transferring from chickens to chickens and generation to generation. In conclusion, we have identified a natural genetic reassortant IBDV, whose genomic segment A is derived from an attenuated strain, and whose segment B is derived from a very virulent strain. This reassortant virus and the classic virulent IBDV have an equivalent virulence in SPF chickens. Our results also provide further evidence that the VP1 protein plays an important role in determining the virulence of IBDV. Acknowledgments We thank Weijie Chen for help in collecting the bursa tissues from IBDV field outbreaks in Zhejiang Province. We thank Dr. Yaowei Huang for help in animal challenge experiment. We thank Professor Wei Jin for their kind help in histopathological examination. We thank Melina Agosto and Marco Morelli for critical reading of the manuscript. This research was supported by the grants from the National High Technology and Research and Development Program of China (No. 2004BA757C) and the Key Project of Zhejiang Province (No. 2003C22002). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bbrc. 2006. 09.040. References [1] H. Mu¨ller, C. Scholtissek, H. Becht, The genome of infectious bursal disease virus consists of two segments of double-stranded RNA, J. Virol. 31 (1979) 584–589. [2] E. Lombardo, A. Maraver, I. Espinosa, A. Fernandez-Arias, J.F. Rodriguez, VP5, the nonstructural polypeptide of infectious bursal disease virus, accumulates within the host plasma membrane and induces cell lysis, Virology 277 (2000) 345–357. [3] E. Mundt, J. Beyer, H. Mu¨ller, Identification of a novel viral protein in infectious bursal disease virus-infected cells, J. Gen. Virol. 76 (1995) 437–443. [4] K. Yao, V.N. Vakharia, Induction of apoptosis in vitro by the 17kDa nonstructural protein of infectious bursal disease virus: possible role in viral pathogenesis, Virology 285 (2001) 50–58. [5] A.A. Azad, S.A. Barrett, K.J. Fahey, The characterization and molecular cloning of the double-stranded RNA genome of an Australian strain of infectious bursal disease virus, Virology 143 (1985) 35–44. [6] A.B. Sanchez, J.F. Rodriguez, Proteolytic processing in infectious bursal disease virus: identification of the polyprotein cleavage sites by site-directed mutagenesis, Virology 262 (1999) 190–199.
[7] M. Brandt, K. Yao, M. Liu, R.A. Heckert, V.N. Vakharia, Molecular determinants of virulence, cell tropism, and pathogenic phenotype of Infectious Bursal Disease Virus, J. Virol. 75 (2001) 11974–11982. [8] K.J. Fahey, K. Erny, J. Crooks, A conformational immunogen on VP-2 of infectious bursal disease virus that induces virus-neutralizing antibodies that passively protect chickens, J. Gen. Virol. 70 (1989) 1473–1481. [9] U. Spies, H. Mu¨ller, H. Becht, Properties of RNA polymerase activity associated with infectious bursal disease virus and characterization of its reaction products, Virus Res. 8 (1987) 127–140. [10] U.I. von Einem, A.E. Gorbalenya, H. Schirrmeier, S.E. Behrens, T. Letzel, E. Mundt, VP1 of infectious bursal disease virus is an RNAdependent RNA polymerase, J. Gen. Virol. 85 (2004) 2221–2229. [11] N. Ismail, Y.M. Saif, P.D. Moorhead, Lack of pathogenicity of five serotype 2 infectious bursal disease viruses in chickens, Avian Dis. 32 (1988) 757–759. [12] D.J. Jackwood, R.J. Jackwood, Infectious bursal disease viruses: molecular differentiation of antigenic subtypes among serotype 1 viruses, Avian Dis. 38 (1994) 531–537. [13] D.B. Snyder, D.P. Lana, P.K. Savage, F.S. Yancey, S.A. Mengel, W.W. Marquardt, Differentiation of infectious bursal disease viruses directly from infected tissues with neutralizing monoclonal antibodies: evidence of a major antigenic shift in recent field isolates, Avian Dis. 32 (1988) 535–539. [14] M.D. Brown, M.A. Skinner, Coding sequences of both genome segments of a European ‘very virulent’ infectious bursal disease virus, Virus Res. 40 (1996) 1–15. [15] A. Wallensten, V.J. Munster, J. Elmberg, A.D. Osterhaus, R.A. Fouchier, B. Olsen, Multiple gene segment reassortment between Eurasian and American lineages of influenza A virus (H6N2) in Guillemot (Uria aalge), Arch. Virol. 150 (2005) 1685–1692. [16] M.F. Saeed, H. Wang, M. Suderman, D.W. Beasley, A. Travassos da Rosa, L. Li, R.E. Shope, R.B. Tesh, A.D. Barrett, Jatobal virus is a reassortant containing the small RNA of Oropouche virus, Virus Res. 77 (2001) 25–30. [17] M.D. Bowen, S.G. Trappier, A.J. Sanchez, R.F. Meyer, C.S. Goldsmith, S.R. Zaki, L.M. Dunster, C.J. Peters, T.G. Ksiazek, S.T. Nichol, RVF Task Force, A Reassortant Bunyavirus isolated from acute hemorrhagic fever cases in Kenya and Somalia, Virology 291 (2001) 185–190. [18] M.J. Roossinck, Evolutionary history of Cucumber mosaic virus deduced by phylogenetic analyses, J. Virol. 76 (2002) 3382–3387. [19] D.J. Robinson, W.D.O. Hamilton, B.D. Harrison, D.C. Baulcombe, Two anomalous tobravirus isolates: evidence for RNA recombination in nature, J. Gen. Virol. 68 (1987) 2551–2561. [20] P.S. White, F.J. Morales, M.J. Roossinck, Interspecific reassortment in the evolution of a cucumovirus, Virology 207 (1995) 334–337. [21] G.J. Miranda, O. Azzam, Y. Shirako, Comparison of nucleotide sequences between northern and southern Philippine isolates of rice grassy stunt virus indicates occurrence of natural genetic reassortment, Virology 266 (2000) 26–32. [22] L.L. Kong, A.R. Omar, M. Hair-Bejo, I. Aini, H.F. Seow, Sequence analysis of both genome segments of two very virulent infectious bursal disease virus field isolates with distinct pathogenicity, Arch. Virol. 149 (2004) 425–434. [23] C. Le Nouen, G. Rivallan, D. Toquin, P. Darlu, Y. Morin, V. Beven, C. de Boisseson, C. Cazaban, S. Comte, Y. Gardin, N. Eterradossi, Very virulent infectious bursal disease virus: reduced pathogenicity in a rare natural segment-B-reassorted isolate, J. Gen. Virol. 87 (2006) 209–216. [24] L. Yu, J.R. Li, Y.W. Huang, J. Dikki, R. Deng, Molecular characteristics of full-length genomic segment A of three infectious bursal disease viruses in China: two attenuated strains and one virulent field strain, Avian Dis. 45 (2001) 862–874. [25] J. Li, Y. Huang, X. Liang, M. Lu, L. Li, L. Yu, R. Deng, Plasmid DNA encoding antigens of infectious bursal disease viruses induce protective immune responses in chickens: factors influencing efficacy, Virus Res. 98 (2003) 63–74.
Y. Wei et al. / Biochemical and Biophysical Research Communications 350 (2006) 277–287 [26] L. Reed, H. Muench, Simple method of determining fifty percent endpoints, Am. J. Hyg. 27 (1938) 493–497. [27] N. Ismail, Y.M. Saif, Immunogenicity of infectious bursal disease viruses in chickens, Avian Dis. 35 (1991) 460–469. [28] J. Li, X. Liang, Y. Huang, S. Meng, R. Xie, R. Deng, L. Yu, Enhancement of the immunogenicity of DNA vaccine against infectious bursal disease virus by co-delivery with plasmid encoding chicken interleukin 2, Virology 329 (2004) 89–100. [29] G.A. Abdel-Alim, Y.M. Saif, Immunogenicity and antigenicity of very virulent strains of infectious bursal disease viruses, Avian Dis. 45 (2001) 92–101. [30] T. Yamaguchi, M. Ogawa, M. Miyoshi, Y. Inoshima, H. Fukushi, K. Hirai, Sequence and phylogenetic analyses of highly virulent infectious bursal disease virus, Arch. Virol. 142 (1997) 1441–1458. [31] B.L. Lim, Y. Cao, T. Yu, C.W. Mo, Adaptation of very virulent infectious bursal disease virus to chicken embryonic fibroblasts by site-directed mutagenesis of residues 279 and 284 of viral coat protein VP2, J. Virol. 73 (1999) 2854–2862. [32] E. Mundt, Tissue culture infectivity of different strains of infectious bursal disease virus is determined by distinct amino acids in VP2, J. Gen. Virol. 80 (1999) 2067–2076. [33] A.A. van Loon, N. de Haas, I. Zeyda, E. Mundt, Alteration of amino acids in VP2 of very virulent infectious bursal disease virus results in tissue culture adaptation and attenuation in chickens, J. Gen. Virol. 83 (2002) 121–129. [34] H.J. Boot, A.A. ter Huurne, A.J. Hoekman, B.P. Peeters, A.L. Gielkens, Rescue of very virulent and mosaic infectious bursal disease virus from cloned cDNA: VP2 is not the sole determinant of the very virulent phenotype, J. Virol. 74 (2000) 6701–6711. [35] M. Takeda, A. Kato, F. Kobune, H. Sakata, Y. Li, T. Shioda, Y. Sakai, M. Asakawa, Y. Nagai, Measles virus attenuation associated with transcriptional impediment and a few amino acid changes in the polymerase and accessory proteins, J. Virol. 72 (1998) 8690–8696.
287
[36] D.D. Brown, B.K. Rima, I.V. Allen, M.D. Baron, A.C. Banyard, T. Barrett, W.P. Duprex, Rational attenuation of a morbillivirus by modulating the activity of the RNA-dependent RNA polymerase, J. Virol. 79 (2005) 14330–14338. [37] S.M. Nolan, S.R. Surman, E. Amaro-Carambot, P.L. Collin, B.R. Murphy, M.H. Skiadopoulos, Live-attenuated intranasal parainfluenza virus type 2 vaccine candidates developed by reverse genetics containing L polymerase protein mutations imported from heterologous paramyxoviruses, Vaccine 23 (2005) 4765–4774. [38] M. Quinlivan, D. Zamarin, A. Garcia-Sastre, A. Cullinane, T. Chambers, P. Palese, Attenuation of equine influenza viruses through truncations of the NS1 protein, J. Virol. 79 (2005) 8431–8439. [39] J. Li, E.C. Fontaine-Rodriguez, S.P. Whelan, Amino acid residues within conserved domain VI of the vesicular stomatitis virus large polymerase protein essential for mRNA cap methyltransferase activity, J. Virol. 79 (2005) 13373–13384. [40] M. Liu, V.N. Vakharia, VP1 protein of infectious bursal disease virus modulates the virulence in vivo, Virology 330 (2004) 62–73. [41] H.J. Boot, A.J. Hoekman, A.L. Gielkens, The enhanced virulence of very virulent infectious bursal disease virus is partly determined by its B-segment, Arch. Virol. 150 (2005) 137–144. [42] H.J. Boot, A.A. ter Huurne, A.J. Hoekman, J.M. Pol, A.L. Gielkens, B.P. Peeters, Exchange of the C-terminal part of VP3 from very virulent infectious bursal disease virus results in an attenuated virus with a unique antigenic structure, J. Virol. 76 (2002) 10346–10355. [43] G.A. Abdel-Alim, Y.M. Saif, Immunogenicity and antigenicity of very virulent strains of infectious bursal disease viruses, Avian Dis. 45 (2001) 92–101. [44] Z.F. Taraporewala, X. Jiang, R. Vasquez-Del Carpio, H. Jayaram, B.V. Prasad, J.T. Patton, Structure-function analysis of Rotavirus NSP2 octamer by using a novel complementation system, J. Virol. 80 (2006) 7984–7994.