key: cord-0683758-wet0go35 authors: Jia, W.; Karaca, K.; Parrish, C. R.; Naqi, S. A. title: A novel variant of avian infectious bronchitis virus resulting from recombination among three different strains date: 1995 journal: Arch Virol DOI: 10.1007/bf01309861 sha: 91f76039e4f3b0ce03e16d1011fd064634f370d4 doc_id: 683758 cord_uid: wet0go35 An antigenic variant of avian infectious bronchitis virus (IBV), a coronavirus, was isolated and characterized. This strain, CU-T2, possesses a number of unusual features, which have not been previously observed in IBV. The S1 glycoprotein of CU-T2 carries virus-neutralizing and serotype-specific epitopes of two IBV serotypes, Arkansas (Ark) and Massachusetts (Mass). Sequence analysis revealed that the virus, originally an Ark serotype, has acquired the Mass-specific epitope by mutation(s). This provides evidence that point mutations may lead to generation of IBV antigenic variants in the field. It was further observed that two independent recombination events involving three different IBV strains had occurred in the S2 glycoprotein gene and N protein gene of CU-T2, indicating that genomic RNA recombination in IBV may occur in multiple genes in nature. It was especially significant that a sequence of Holland 52 (a vaccine strain) had replaced half of the N gene of CU-T2. This proves that recombination among vaccine strains is contributing to the generation of IBV variants in the field. Based on these observations it is predicted that every IBV field isolate could have unique genetic nature. Therefore, several recently reported diagnostic and serotyping methods of IBV which are based on dot-blot hybridization, restriction fragment length polymorphism (RFLP), and polymerase chain reaction (PCR), may not reveal the true antigenic and/or genetic nature of IBV isolates, and may in fact yield misleading information. results in heavy economic losses to the commercial poultry industry, worldwide [24] . IBV has a single-stranded, positive sense RNA genome. It is 27.6 kilobases long, encodes three major structural proteins, and is organized as 5'-POL-S-Gene3-M-Gene5-N-3' (Fig. 1) [9, 35, 36] . The three structural proteins of IBV are the spike glycoprotein (S protein), the membrane glycoprotein (M protein) and the nucleocapsid protein (N protein). The S protein of IBV is cleaved post-translationally into N-terminal S1 and C-terminal $2 proteins [13] . The S1 protein carries antigenic epitopes which induce virus neutralizing (VN) antibody, and also determine the virus serotype [14, 15] . The N protein may induce cellular immune responses against IBV [6] . IBV isolates have been grouped into serotypes based on VN test [16, 18] . Although more than twenty distinct IBV serotypes have been reported in the USA [17, 32] , most isolates causing disease in the field belong to Massachusetts (Mass), Connecticut (Conn), or Arkansas (Ark) serotypes [24] . Live vaccines containing strains of those three serotypes have been routinely used in the field. However, more frequently than ever, outbreaks of the disease are being observed in vaccinated flocks. These vaccine failures, however, are not always due to infection by distinct serotypes, but could be caused by antigenic variants of IBV emerging from wild-type or vaccine viruses by point mutation(s) or genomic RNA-RNA recombination [12, 25] . Although point mutation(s) are believed to contribute to the generation of new antigenic variants of IBV, no direct evidence for that has been presented. RNA-RNA recombination has been shown to occur at a high frequency both in vivo and in vitro in mouse hepatitis virus (MHV), a coronavirus I- 3, 22, 23, [27] [28] [29] 31 ], yet, no isolation of recombinant MHV from natural disease in mice has been reported. In the case of IBV, one Japanese and three European isolates have been suggested to be possible recombinants [11, 25] . However, the origin of ~-~'~Unknown-oriqin sequence[-------]Hol152-1ike sequence ~ HVR Fig. 1 . Schematic characterization of the CU-T2 genome. POL RNA polymerase gene. S S protein gene; $1 S1 protein gene; $2 $2 protein gene. N N protein gene. U TR 3' end untranslated region. HVR Hypervariabte region the recombination fragments in those isolates was unknown. More recently, two American field isolates were found to contain fragments of Mass-like sequences in the S1 gene, which were 94% to 95% homologous to IBV strain Mass41 [38] . All the recombinations suggested in IBV so far have been in the S gene, and none of those reports have addressed whether the recombinations had altered the antigenic characteristics of the viruses. Here we describe an IBV serotype and recombination variant, CU-T2, which reacted with two monoclonal antibodies (MAbs) specific to the VN epitopes of IBV serotypes Mass and Ark, respectively. We present evidence that the CU-T2 was originally an Ark serotype which subsequently acquired a Mass-specific epitope on the S1 protein through mutation(s). We also show that CU-T2 had undergone two independent genomic RNA recombination events within the $2 and N protein gene (N gene), respectively, involving three different IBV strains. Furthermore, we provide evidence that CU-T2 had acquired a sequence of Holland 52 (Ho1152), a vaccine strain of Mass serotype used in the field. Implications of such genetic and antigenic changes in diagnosis and serotyping of IBV are discussed. Four previously characterized IBV strains, Ark99, Hol152, Corm46, and Mass41 [19, 20] were obtained from our own laboratory stocks. CU-T2 was isolated from an adult commercial chicken flock in New York State, which was experiencing reduced egg production and poor egg shell quality associated with IBV infection. Virus isolation from the flock was performed through the use of "sentinel" chickens. Briefly, ten specificpathogen-free (SPF) chickens were placed among the affected birds in separate cages for ten days. Subsequently, the "sentinel" chickens were euthanized and the tracheas were collected for passage in SPF chicken embryos. CU-T2 strain was recovered from the allantoic fluid of inoculated embryos, and was subsequently propagated and plaque purified in chicken kidney cell (CKC) culture monolayers [16] . MAbs 1318, 1588, and 940 specific to Sl-associated serotype-specific epitopes of IBV prototype Ark, Mass, and Conn, respectively, have been described elsewhere [21] . The protocols for both antigen-capture and competitive-binding ELISAs have also been reported previously [21, 34] . The ELISAs were performed using whole virions, or purified S1 protein of CU-T2 according to the procedures used previously [21] . Viruses were propagated in chicken embryos and purified by sucrose gradient as described previously [37] . Viral genomic RNAs were extracted with phenol-chloroform and used for cDNA synthesis and RNA sequencing. cDNA synthesis, cIonin9 and sequencin 9 For cDNA synthesis, an Uni-ZAP II kit (Stratagene, La Jolla, CA) was used in conjunction with a synthetic primer complementary to a sequence located downstream of the S genes of 262 W. Jia et al. several Mass serotype strains [5] (5'-GAACTAGTCTCGAGGAAGGACGTGGGACT-TTG-Y). The cDNA library was screened by hybridization with a 5' end-radiolabeled probe prepared from a StyI fragment derived from a cDNA clone (pBSM41M) of S gene of Mass41 obtained from Solvay Animal Health, Inc., Mendota Heights, MN. Phagemids containing IBV genes were prepared by in vivo excision in Escherichia coli (E. coli) using a protocol furnished by Stratagene (La Jolla, CA). DNA sequencing was carried out by the dideoxy method, using a Sequenase 2.0 DNA sequencing kit (United States Biological (USB), Cleveland, OH). cDNA clones pARK99-8pl, pBARK99-8p2, pBT2-t223, pBT2-t229, and pBT2-t237 were used for sequencing. Both strands of each cDNA clone were sequenced. Sequence data were obtained from cDNA clones, except for the first 1100 bases of the S gene, the last 200 bases of the N gene, the 3' end non-coding region of CU-T2, and the partial Gene3 of Ark99 and Ho1152, which were obtained from direct sequencing of IBV genomic RNA. Direct sequencing of IBV genomic RNA was also performed at least twice for those regions, using an RNA sequencing kit (USB, Cleveland, OH). Sequence alignments and analysis were performed using the Genetics Computer Group (GCG) program. The nucleotide sequences reported here have been deposited with the GenBank. The accession numbers are as follows: S protein gene of IBV Ark99, L10384; S protein gene of IBV CU -T2, U04739; N protein gene of IBV CU-T2, U04805; 3' end non-coding region of IBV CU-T2, U04804. Tables 1 and 2 summarize results of the three types of ELISAs performed with the Ark, Conn, and Mass serotype-specific MAbs. In both the indirect and antigen-capture ELISA, the Ark-specific and Mass-specific MAbs reacted with both the whole virions and the purified S1 protein of CU T2 strain (Table 1) . In the competitive-binding ELISA, unlabeled Mass-and Ark-specific MAbs blocked reaction of the homologous labeled MAbs, but not with the heterotogous MAbs. In order to determine the genetic basis for the expression of the Ark-and Mass-specific epitopes on the $1 protein ofCU T2, the entire S gene of CU T2 was sequenced. The entire S gene of Ark99 was also sequenced for comparison. All base positions indicated in this paper are calculated from the S gene start codon [4] . Amino acid positions indicated are of the S protein precursor, including the signal peptide. The gene of Ark99 was found to be 3480 bases long, with a capacity to code 1159 amino acids. The S1 and $2 genes were 1629 and 1851 bases long, respectively, and were predicted to encode 543 and 616 amino acids, respectively. Comparison of the deduced amino acid sequences of the S proteins of Ark99 and Mass41 [5] , revealed an 88.2~o homology (1022/1159). However, the S1 protein showed only 78.1~o homology (424/543), compared to 97.1% (598/616) for the $2 (data not shown). Amino acid variations in the S1 proteins of the two viruses were more than 7 times as frequent as for the $2 protein. Alignment of the $1 gene of Ark99 with that of Mass and several European IBV isolates revealed that fifteen bases (5'-TGGAAGTGCTACGCC-3') between bases 419 and 435 of the Ark99 strain were unique to that virus (data not shown). A 95.3~ homology was observed between the first 1100 bases of the $2 gene of Ark99 and Mass41, while the last 751 bases of the two genes were identical (Fig. 2) . The nucleotide sequences of S1 genes of CU-T2 and Ark99 exhibited over 98~o homology° The fifteen-base insertion in the S1 gene of Ark99 described above was also observed in the CU-T2 $1 gene, but its location was between bases T AGGGTCTTAA TGACTC-TT-ATAGACCTTG AAA-A-T-TC AATACTCAAA ACTTATATTA AGTGGCCTTG GTATGTGTGG TTAGCCATAG CTTT-GCCAC 3330 3429 T C A A A GC TC (~ GT C Q A TATTATCTTC ATC-TAATA-TAGG-TGGGT TTTCTTCATG ACTGG-TGTT GTGGTTGTTG TTGTGGATGC TTTGGCATTA TGCCTCTAAT GAGTAAGTGT 3430 3529 T C T T 416 and 432, as compared to bases 419 and 435 for Ark99. The difference in the position of the insertion was due to a three-base deletion (TAA) between bases 356 and 360 of the S1 gene of CU-T2 compared to the S1 gene of Ark99. A single noncoding substitution was noted within the fifteen-base insertion of CU-T2 (T versus C at position 428). A 96.7% homology was observed between the amino acid sequences of the S1 proteins of CU-T2 and Ark99. Seventeen amino acids were found to be different in the S1 protein of CU-T2 compared to those in Ark99, and those differences were distributed throughout the gene (data not shown). A 96.6~o homology was observed in the first 1100 bases of the $2 genes of Ark99 and CU-T2, but, the following 751 bases had only 87.63/o homology (Fig. 2) . Comparison with published sequences showed that the sequence of this 751-base region of CU T2 was more similar to a fragment in the 3' end of the $2 gene of a Japanese isolate KB8523 [37] than to those of all other strains. Although 76 nucleotide substitutions were found in that fragment of the CU-T2 and KB8523, 66 of those were synonymous substitutions (Fig. 2) . The downstream crossover site of this fragment was about 12 to 13 bases downstream of the stop codon of the $2 gene of CU-T2. Through a substitution of T by G in the TAA stop codon, the $2 gene of CU-T2 was 27 bases longer than that of Ark 99, but of the same length as the $2 gene of KB8523 (Fig. 2) , indicating that a fragment of approximately 751 bases was incorporated into the $2 gene of CU-T2 by recombination. To further study the possible evolutionary origin of the CU-T2 strain, the first 268 bases of Gene3, the N gene and the 3' non-coding region of CU-T2 were sequenced subsequently. The same regions of Gene3 of Ark99 and Ho1152 were also sequenced for comparison. The first 268 bases of the Gene3 of CU-T2 matched more closely with those of Ark99 Gene3 (97.6~o homology), than of Mass41 [7] , KB8523 [37], or Ho1152 (Table 3) . The N gene of CU-T2 was t230 bases long, with a capacity to code 410 amino acids. Around 436 bases at the 5' end and 150 bases at the 3' end of i I00 G A G G A C T A G T G ATGGC~GCG GT~GGC~C TGGAAAGACA GACGCCCCAG CGCCAGTCAT CAAACTAGGA GGACCAAA-C CACCTAAAGT TGGTTCTTCT GGAAATGCAT 101 200 C T T C TT T A T T T CTTGGTTTCA AGC~T#~ GCC~GAAGC TAAATTCACC TCCACCT~G TTTG~GGTA GCGGTGTTCC TGAT~TG~ ~TCTTAAAA C~GCCAGCA 201 300 G T C AC T C A T T A G C T G T A T ACATGG-TAC TGGAGACGCC ~GC-AGGTT T~GCCAGGT AAAGGCGGAA GAAAACCAGT CCCAGATGCT TGGTACTTCT ATTATACTGG AACAGGACCA 301 400 T G C C GT C A T A GCCGCTGACC TG~TTGGGG TGATAGCC~ GATGGTATAG TGTGGGTTGC TGCAAAGGGT GCTGATACTA ~TCTAGATC T~CCAGGGT AC~GGGACC 401 500 T T TT G A C GAA AT T T T ATC~CAGC-GCTTCATCAG CAGCATCTAG TAGAGCACCG TCGCGTGATG GCTCGCGTGG -CGTAG~G-GG-GCTG~G ATGATCTTAT TGCTCGTGCA 601 700 T A ~ AG T C C C T GC~J~AGATCA TTCAGGATCA GCAG~G~G GGTTCTCGCA TTACTAA-GC T~GGC-GAT GAAATGGCTC ATCGCCGGTA TTG- T T C C A -GTTACAGCA ATGCTCAACC TAGTCCCTAG CAGCCATGCT TGTCTTTTTG GAAGTAGAGT GACGCCCAAA CTTC~CCAG ATGGGCTGCA CTTGA-ATTT CC T T A A T Q T AG T T G~TTTACTA CTGTGGTTTC ACGTGATGAT CCGCAGTTTG AT~TTATGT GAAAATTTGT GATCAGTGTG T- G C C A T A G C G G TA T T G~GCAGGAT GATGAAGTAG ATAAAGCATT GACCTCAGAT GAGGAGAGGA ACAATGCACA GCTGGAATTT GATGATG~C CC~GGTGAT T~-TGGGGG 1201 1230 AC T A GT G A TT A TT A C AC T A GATTCAGC-- T-GGTGAG~ TG~CTTTGA CUT2 N gene were very similar to the comparable regions of Ark99. However, two crossover sites were observed around bases 436 and 1090 (Fig. 3) . Between those two sites only five bases were different (99.2% homology) between the CU-T2 and the Ho1152 strains, and of those, only two were non-synonymous (Fig. 3) . Within the same region, 42 bases were different between CU-T2 and ARK99 (93.6% homology), and eight of those changes were non-synonymous (Fig. 3 ). There were many more differences in that region between CU-T2 and other IBV strains, including Mass 41, Gray, KB8523 and Beaudette (data not shown). The 3' end untranslated region (UTR) of the IBV genome was previously shown to contain an approximately 190-base hypervariable region (HVR) [39] (Fig. 1) . That region of the CU-T2 sequence had a high homology with Ark 99 (96.4~), but lesser homology with KB8523 (89.3%), Ho1152 (77.3%) and Mass41 (52.4~) (data not shown). The results show that CU-T2 is a serotype and recombination variant which originated from an Ark-like IBV. This is based on both the sequence data and MAb studies of the virus. The nucleotide sequence homologies between the S 1 gene, most of $2 gene, Gene3 and the 3' UTR of CU-T2 and Ark99 all attest to the origin of CU-T2 from an Ark-like virus. The fifteen-base insert observed in the $1 genes of both Ark99 and CU-T2, which was absent in the Mass serotype strains, further strengthens this conclusion (Fig. 1) . The seventeen amino acids in the S 1 protein of CU-T2 which were different from those in Ark99 were distributed throughout the gene, indicating that these amino acid changes were not due to recombination. Therefore, it may be concluded that CU-T2 acquired the Mass-specific serotype epitope on the S1 protein as a result of point mutation(s). The results of competitive-binding ELISAs indicated that the Ark-and Mass-specific epitopes on CU-T2 are independent. These facts provide direct evidence that point mutation(s) can result in the generation of antigenic variants during a natural infection. The results show that an approximately 751 base-long fragment of unknown origin was incorporated into the 3' end of CU T2 $2 gene by recombination. Interestingly, this recombinant fragment was similar to one previously observed in KB8523, a possible recombinant strain of IBV [25] . Recombination fragments in other two European isolates (6/82 and D207) were also found within the same regions [25] , although they were of shorter length than those of CU-T2 and KB8523 strains. In an in vitro transfection experiment it was observed that RNA recombination in MHV occurred preferentially at certain selected sites (hot spots) [1] . In IBV, a secondary structure most likely exists at the 5' end of Gene3 [30] . We found that the downstream crossover sites of both CU-T2 and KB8523 were located around that secondary structure (data not shown). The clustering of recombinations in IBV suggests that a recombination 'hot spot' most likely exists around the 3' end of the S2 gene and the 5' end of the Gene3, although the clustering may be a result of selection [2] . An additional recombination which resulted in the replacement of half (52.4~) of the N gene was also observed, indicating that genomic RNA recombination of IBV may involve more than two strains and may occur in multiple genes during natural infection. More interestingly, that recombination fragment in the CU-T2 is almost identical to a sequence (99.2~o homology) in the N gene of Ho1152 strain. Since both Ark99 and Ho1152 strains have been used simultaneously as live vaccines in the USA [24] , this finding provides convincing evidence that recombination among vaccine strains is contributing to the emergence of variants in the field. The significance of recombination in the N gene must also be viewed in the light of a recent report that the N protein may induce cellular immune responses against IBV [6] . The sequence of S1 portion of Ark99 S gene obtained from a cloned PCR product has been published more recently [38] . When the S1 portion of the entire S gene sequence of Ark99 obtained in this study was compared with that sequence, nine bases were found to be different, which would result in four amino acid changes in the S1 protein at positions 143, 335, 348 and 356. Our data suggest that both inter-strain recombinations and mutation(s) are contributing to the generation of IBV variants in the field. We have shown here that genomic recombination between IBV strains may lead to replacements of large RNA fragments in multiple genes. The replacement, however, may not result in change of antigenic nature of IBV. In contrast, mutation(s) on the S1 protein alone may generate new VN epitope(s). The data imply that every field isolate of IBV could be unique in RNA sequence. Therefore, recently reported diagnostic and serotyping methods of IBV, such as dot-blot hybridization [33] , restriction fragment length polymorphism (RFLP) and polymerase chain reaction (PCR) [26, 40] , may not reflect the true antigenic and/or genetic nature of the virus. Indeed, an exclusive use of such assays may yield misleading information on the antigenic and/or genetic characteristics of the virus. A clustering of RNA recombination sites adjacent to a hypervariable region of the peplomer gene of murine coronavirus Random nature of coronavirus RNA recombination in the absence of selection pressure Establishing a genetic recombination map for murine coronavirus strain A59 complementation groups Cloning and sequencing of the gene encoding the spike protein of the coronavirus IBV Comparison of the spike precursor sequences of coronavirus IBV strains M41 and 6/82 with that of IBV Beaudette Induction of anti-viral immune responses by immunization with recombination-DNA encoded avian coronavirus nucleocapsid protein Sequencing of coronavirus IBV genomic RNA: three open reading frames in the 5' 'unique' region of mRNA D Sequence of the nucleocapsid genes from two strains of avian infectious bronchitis virus Completion of the sequence of the genome of the coronavirus avian infectious bronchitis virus Sequence analysis of strains of avian infectious bronchitis coronavirus isolated during the 1960s in the Infectious bronchitis virus: evidence for recombination within the Massachusetts serotype Location of the amino acid differences in the S1 spike glycoprotein subunit of closely related serotypes of infectious bronchitis virus Coronavirus IBV: partial amino terminal sequencing of spike polypeptide $2 identifies the sequence Arg-Arg-Phe-Arg-Arg at the cleavage site of the spike precursor polypeptide of IBV strains Beaudette and M41 Coronavirus IBV: virus retaining spike glycopolypeptide $2 but not S1 is unable to induce virus-neutralizing or haemagglutination inhibiting antibody, or induce chicken tracheal protection Coronavirus IBV: removal of spike glycopolypeptide S1 by urea abolishes infectivity and haemagglutination but not attachment to cells Serotyping of avian infectious bronchitis viruses by the virus-neutralization test Variant serotypes of infectious bronchitis virus isolated from commercial layer and broiler chickens Serological comparisons of strains of infectious bronchitis virus using plaque purified isolates The neutralizing characteristics of strains of infectious bronchitis virus as measured by the constant-virus variable-serum method in chicken tracheal cultures Immunologic differences in strains of infectious bronchitis Production and characterization of monoclonal antibodies to three infectious bronchitis virus serotypes In vivo RNA-RNA recombination of coronavirus in mouse brain RNA recombination of murine coronaviruses: recombination between fusion-positive mouse hepatitis virus A59 and fusion-negative mouse hepatitis virus 2 Avian infectious bronchitis In: Calnek BW Sequence evidence for RNA recombination in field isolates of avian coronavirus infectious bronchitis virus Differentiation of infectious bronchitis virus serotypes using polymerase chain reaction and restriction fragment length polymorphism analysis RNA recombination in animal and plant viruses Recombination between nonsegmented RNA genomes ofmurine coronaviruses RNA recombination in a coronavirus: recombination between viral genomic RNA and transfected RNA fragments Internal entry of ribosomes on a tricistronic mRNA encoded by infectious bronchitis virus High-frequency RNA recombination of murine coronaviruses An overview of infectious bronchitis virus serotypes Dot-blot hybridization using digoxigenin-labeled cDNA probe complementary to the S1 gene of avian infectious bronchitis virus permits discrimination between virus strains A monoclonal antibody-based antigen capture enzyme-linked immunosorbent assay for identification of infectious bronchitis virus serotypes Rapid detection and identification of avian infectious bronchitis virus We thank Ms. Beverley Bauman, and Ms. Alice Adriguetto for their technical assistance, and Dr. Bruce Calnek and Dr. Randall Renshaw for critical reading of the manuscript. The clone of Mass41 S gene was kindly provided by Dr. Georges Thiry, Sotvay Animal Health, Inc. This work was partially supported by a grant from Sotvay Animal Health, Inc.Recombination among IBV strains