key: cord-0688472-n75ldyy2
authors: Motokawa, K.; Hohdatsu, T.; Aizawa, C.; Koyama, H.; Hashimoto, H.
title: Molecular cloning and sequence determination of the peplomer protein gene of feline infectious peritonitis virus type I
date: 1995
journal: Arch Virol
DOI: 10.1007/bf01718424
sha: dfcd3c7178d37774f5e1f1bbdc64bfd8257a7d15
doc_id: 688472
cord_uid: n75ldyy2

cDNA clones spanning the entire region of the peplomer (S) gene of feline infectious peritonitis virus (FIPV) type I strain KU-2 were obtained and their complete nucleotide sequences were determined. A long open reading frame (ORF) encoding 1464 amino acid residues was found in the gene, which was 12 residues longer than the ORF of the FIPV type II strain 79–1146. The sequences of FIPV type I and mainly FIPV type II were compared. The homologies at the N- (amino acid residues 1–693) and C- (residues 694–1464) terminal halves were 29.8 and 60.7%, respectively. This was much lower than that between FIPV type II and other antigenically related coronaviruses, such as transmissible gastroenteritis virus of swine and canine coronavirus. This supported the serological relatedness of the viruses and confirmed that the peplomer protein of FIPV type I has distinct structural features that differ from those of antigenically related viruses.

Feline infectious peritonitis (FIP) is a virus-induced chronically progressive and usually fatal disease in domestic and wild Felidae. The causative agent of this disease is FIP virus (FIPV) which belongs to the family Coronaviridae. FIPV forms a related antigenic cluster with feline enteric coronavirus, transmissible gastroenteritis virus (TGEV) of swine, canine coronavirus (CCV), and porcine respiratory coronavirus (PRCV) [12, 18, 23] . FIPV has two serological subtypes, type I and type II, which cause similar diseases in animals [-8, 9, 20] . When we surveyed antibody positive cats in Japan, about 70 and 30% with FIP were infected with types I and II, respectively [10] .

FIPV is an enveloped RNA virus with a single-stranded positive-sense RNA genome. The FIPV virions consist of three main structural proteins, peplomer (S) protein, membrane (M) protein and nucleocapsid (N) protein. A comparison of the antigenicity of M and N proteins of FIPV types I and II revealed that 470 K. Motokawa et al. they are serologically closely related. However, their peplomer proteins were not homologous [5, 8, 9] . The virus genome is about 20 kilobases (kb) tong and three major structural protein genes are located in the 3' half of the genome [3] . The complete nucleotide sequences of these genes have been established with FIPV type II strain 79-1146 [2, 4, 27] . Several other serologically related coronaviruses have also been investigated molecularly, such as TGEV [13, 22] , and CCV [11, 28] .

To date, immunization against FIP has not been available because of the antibody-dependent enhancement (ADE) phenomenon due to humoral antibody [7, 15, 17, 19, 30] . To clarify the involvement of virion antigens in this phenomenon, Vennema and his collaborators cloned cDNAs for the S, M and N protein genes of FIPV and constructed recombinant vaccinia viruses, each of which contained one of these cDNAs. These recombinant viruses were used for immunization and subsequent virus challenge of kittens. Those authors indicated that only the peplomer protein might be responsible for this phenomenon [26, 27] . FIPV type II has been preferentially used for experimental materials, probably because the proliferative efficiency of FIPV type II in cultured cells is much higher than that of type I. Thus, type I has been little investigated, especially at the molecular biological level. Therefore, it seems important to study FIPV type I, which is more prevalent in Japan than type II. In this study cDNA clones and sequences of the peplomer gene of FIPV type I were compared with those of related viruses. These studies confirmed the serological relatedness between FIPV type I and other coronaviruses.

Virus and its genomic RNA FIPV type I, strain KU-2, isolated by , was propagated in Felis catus whole fetus cells (fcwf 4). Virus-infected cells were homogenized with a Dounce type homogenizer and celt nuclei were removed by low speed centrifugation. Virions in the supernatants were pelleted through layers of 35% sucrose by centrifugation at 200 000 x g for 2h. Virus genomic RNA was extracted from the virus pellets with SDS-phenol and precipitated with ethanol.

DNA primers were custom-synthesized by Bex Corp. (Tokyo) by using di-amidite chemistry and an automatic DNA synthesizer. Two minus-sense primers were prepared for each cloning study, one of which, located downstream, was used for reverse transcription, and the other, located upstream, for subsequent PCR. The nucleotide sequences and positions of the primers are shown in Table 1 . Primer IIMPr-1 and IIMPr-2 were created with reference to the nucleotide sequence [27] of the M protein gene of FIPV strain 79-1146.

Genomic template RNA of FIPV type I strain KU2 was reversely transcribed with a negative-sense downstream DNA primer. The template RNA was digested with 0.2 M NaOH at 55 °C for 1 h and a poly dA tail was added to the single-stranded cDNA by using terminal deoxynucleotidyl transferase. The poly dA-tailed cDNA was amplified by PCR, primed with another 5' upstream primer and the oligo dT21 primer, which have recognition sites for restriction enzymes BamHI, EcoRI and PstI. DNA fragments larger than 1-kb pairs were obtained by electrophoresis with low melting point agarose, digested with restriction enzymes and cloned into the same restriction sites of pUCI8. The cDNA clones were selected by sequencing the nucleotides of the 3' end that contained the primer sequence.

The cloned cDNA was subcloned into M13mplS/19 and single-stranded DNA was sequenced. The restriction sites used to obtain the cDNA fragments are shown in Fig. 1 . The DNA was sequenced by means of dideoxynucleotide chain termination using the Dye Primer cycle sequencing kit (Applied Biosystems). The sequence was resolved with an automated DNA sequencer (Applied Biosystems model 373A). The sequences determined were then analyzed with the GENETYX computer program (Software Development Co., Ltd.). Homology including the deleted sequence was calculated. The nucleotide sequence data reported in this paper will appear in the GSDB, DDBJ, EMBL and NCBI nucleotide sequence databases with the accession number D32044.

Since FIPV type I proliferates more slowly than type II, it was not easy to obtain a large quantity of purified viruses, We used PCR to generate sufficient cDNA from a viral template. When the first cDNA was synthesized, the minus-strand DNA primers for the 3' end necessary for reverse transcription (IIMPr-1) and PCR (IIMPr-2) were prepared on the basis of the nucleotide sequence of the FIPV strain 79 1146 M gene. A poly dA tail was added to the single-strand cDNA by using terminal deoxynucleotidyl transferase, and oligo dT21 was used as the plus-strand primer for PCR. The amplified cDNA was molecularly cloned into the multiple cloning sites of pUC18 and nucleotide sequences at both ends of the clones were determined. Those clones containing the nucleotide sequence of the primer at the 3' end were selected and those containing it at the 5' end were ret'erred to prepare DNA primers for the next cDNA cloning. The cDNA cloning experiments were hence performed sequentially. Finally, five cDNA clones (pFPSI-1, 2, 3, 4 and 5), four of which covered the entire peplomer gene, were isolated as shown in Fig. t . The complete nucleotide sequence of the peplomer gene of FIPV type I was determined by sequencing these cDNA clones. At least five clones for one type of cDNA were sequenced to avoid artifact mutations due to misreading by reverse transcriptase and Taq polymerase for PCR.

A long open reading frame (ORF) was found extending from the second to the fifth cDNA, which was considered to be the peplomer protein of FIPV type I. This ORF is 4392 bases (1464 amino acid residues) long, which is equivalent Table 1 to a predicted protein of 163.5 kDa. This gene is 36 bases longer (12 amino acids) than that of FIPV type II reported by De Groot et al. [2] . The total nucleotide and deduced amino acid sequences in the ORF of the peplomer gene of FIPV type I are shown in Fig. 2 .

The peplomer protein of FIPV type ! also has two hydrophobic segments characteristic of a type I membrane protein with an N-terminal signal sequence (residues 1-28) [29, 31] and a transmembrane domain (residues 1406-1426). The amino acid sequence homology of the latter is completely conserved but that of the former is not. Downstream of the transmembrane domain, there are many cysteine residues. A similar N-terminal signal sequence, C-terminal transmembrane domain and cysteine cluster are present in the peplomer proteins of FIPV type II, CCV and TGEV.

Forty-one potential N-glycosylation sites (NXS,T) are present in the overall peplomer protein, which is six sites fewer than in type II. Their locations are in good accordance at the C-terminal part but less so at the N-terminal part. The peplomer proteins of FIPV type II and TGEV are not cleaved by host cell proteinases [25] . In the peplomer protein of FIPV type I, no sequence was identified as the cleavage motif RRFRR for avian infectious bronchitis virus [1] except for the tetranucleotide sequence RRSR (residues 787-790) as a vestigial cleavage site.

The 5' and 3' non-coding regions (nucleotides) and the ORFs (amino acids) of the FIPV types I and II peplomer protein gene are aligned in Fig. 3 . The 5' non-coding region is well conserved, but the 3' non-coding region is not, with G~GGTAAAATACTCA~AG~T~TGGT~G~A CT~C~GGTMT CAC~G~CACAC CAT~TA~ CAT~TAC~ACACT CC~AGC G I~ *""**'*** M I F I I L T L L S V 11 NGAGg~GT~TM~A~AGG~GMGAT~A~GTAG~A~TT~FTGT~CATCCTATACTATGCCAC~TlTTA~ACAT~CAAA~AT?~ 

TGAAAGATTCAATGCCACAGCTTTAGGTGGTGAAAAGCTAGGC GGITTATA I I I I GATGGC CTGAGCAGTCTATTAC C GC CTAAAA~GGTAAGAGGTC G 38@@ A L G G E K L G G L Y F D G L S S L k P P K I G K R S 977 E R F GCTGTTGAAGATCTATTGTTCAATAAAGTGGTGAC CAGCGGTCTTGGCACTGTTGATGATGACTATAAAAAGTGCTCTTCC GGCACTGACGTTGCAGATC 31~ A V E D L L F N K V V T S G L G T V D D D Y K K C S S G T D V A D L 1011 TAGTTTGTGC CCAATA~ACAATGGCATAATGGTTTTACCTGGTG~GT GGATG GTAATAAGATGTCTATGTACACTGCATCTTTAATTGGCGGTATGGC 32@~ V C A

Y G F C G GGAAGAAGTGAC GG•ATGGTCAGGAATATGTGTrAATGATACTTATGCATATGTGTTGAAAGA•-FTTGATcATT••ATl-•TCAGCTACAATGGcACGTAT 39@@ E E V T A W S G I C V ~ D TI

TT CT CTTTGTA GTAGAAG GCAATTTGAAAC CTAT GAA C C CATTGAAAAG GTT CA CATTCAI-FAACTA GAC GAl-rTATG GATA CTGT CAAGT CTATTG G CA 45@@

TCTCTGTGGACGCTGTACI -FGACGAGTTAAATTCCG 4536 homology is very low in the N-terminal half, but higher in the C-terminal half. The sequence homology data of the peplomer proteins were analyzed and are summarized in Table 2 and Table 3 , in which FIPV type I is compared with FIPV type II, TGEV and CCV. Amino acid sequence homology of the entire peplomer protein area was only about 45% between each virus and FIPV type I ( Table 2) . Table 3 shows homology at the N-terminal (residues 1-693) amino acid sequence and C-terminal (694-1464) amino acid sequence of FIPV type I. While the C-terminal half showed high homology with FIPV type II, CCV and TGEV, FIPV type I showed homology of only about 60% for the other viruses. This table also shows that FIPV type I is distinct from the other three coronaviruses in terms of amino acid sequence homology of the peplomer protein.

In this study, we first established cDNA clones and sequenced the nucleotides of the peplomer protein gene of FIPV type I, to compare the sequence of FIPV type I with those of antigenically related viruses in the family Coronaviridae.

FIPV forms a related serological cluster with TGEV, CCV and PRCV in the coronavirus family [12, 18, 23] . In this cluster, FIPV type I is considered to be very discriminative to TGEV, CCV and even FIPV type II, based upon the reactivity with established monoclonal antibodies to the peplomer protein [5, 8, 9] . Jacobs et al. have reported that there is a great divergence (30% homology) between FIPV II and TGEV at the first N-terminal part of the peplomer protein despite a high level of conservation (94%) with only 74 amino acid substitutions at the residual C-terminal part. They stated that this divergence could not be derived from selection of neutralizing antibodies but would be generated by recombination with a related virus [13] . However, a much larger diversity was found in the corresponding region (residues 1-290) between FIPV types I and II (25.4%) and the homology in the residual N-terminal part (residues 291-693) was also very low (32.9%). Furthermore, as shown in Table 2 , the amino acid sequence homology was high in the peplomer proteins among FIPV type II, TGEV and CCV. That is, the homology of the entire peplomer protein of FIPV type II with that of CCV is 90.9%, and that between FIPV type II and TGEV is 80.9%. In contrast, the homology between FIPV types I and II is 45.9% over the entire region, with 60.7% in the C-terminal half (residues 694-1464) and only 29.8% in the N-terminal half (residues 1-693). Consequently, the amino acid sequence of the FIPV type II peplomer protein is much more homologous with those of CCV and TGEV than with that of FIPV type I. These results support the serological relatedness by revealing the extensive heterogeneity of the peplomer protein of FIPV type I within the cluster.

We found greater diversity in the corresponding region between FIPV types I and II. Such divergent N-terminal domains are probably involved in the construction of the globular head structures of the coronavirus peplomer protein. These structures must have important roles for virus infection, when viruses attach to cellular receptor sites. It is unlikely that virions of both FIPV types bind to the same receptors on the cellular membrane.

The ADE phenomenon has hindered the control of FIP by immunization with vaccines. In fact, neither inactivated virions nor recombinant antigens in the vaccinia virus vector induced protection in vaccinated animals against subsequent virus challenge [24] . Vennema et al. have postulated that only the peplomer protein is responsible for this phenomenon [26, 27] . This protein is also thought to be the most suitable for use as an FIP vaccine. Whether the epitopes responsible for ADE and protection against viral infection are separable or not will be a fundamental problem in developing usable FIP vaccines. At present, we have no conclusive information about this. Even if a protective vaccine for either type of FIPV is developed in the future, one for the other type will also be required, because the peplomer proteins are quite different. Since FIP is usually a fatal disease and its morbidity rate is not low, effective vaccines are desirable for control of this disease.

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 propolypeptide of IBV strains Beaudette and M41

cDNA cloning and sequence analysis of the gene encoding the peplomer protein of feline infectious peritonitis virus

Intracellular RNAs of the feline infectious peritonitis coronavirus strain 79-1146

Sequence analysis of the 3'-end of the feline coronavirus FIPV 79-t146 genome: comparison with the genome of porcine coronavirus TGEV reveals large insertions

Antigenic comparison of feline coronavirus isolates: evidence for markedly different peplomer glycoproteins

Functional differences in the peplomer glycoproteins of feline coronavirus isolates

A study on the mechanism of antibody-dependent enhancement of feline infectious peritonitis virus infection in feline macrophages by monoclonat antibodies

Characterization of mononoclonal antibodies against feline infectious peritonitis virus type II and antigenic relationship between feline, porcine, and canine coronaviruses

Antigenic analysis of feline coronaviruses with monoclonal antibodies (MAbs); preparation of MAbs which discriminate between FIPV strain 79-1146 and FECV strain 79-1683

The prevalence of types I and II feline coronavirus infections in cats

Analysis of a 9.6 kb sequence from the 3' end of canine coronavirus genomic RNA

Antigenic relationships among homologous structural polypeptides of porcine, feline, and canine coronaviruses

The nucleotide sequence of the peplomer gene of porcine transmissible gastroenteritis virus (TGEV): comparison with the sequence of the peplomer protein of feline infectious peritonitis virus (FIPV)

A simple method for displaying the hydropathic character of a protein

Monoclonal antibodies to the spike protein of feline infectious peritonitis virus mediate antibodydependent enhancement of infection of feline macrophages

A review of feline infectious peritonitis virus: molecular biology, immunopathogenesis, clinical aspects, and vaccination

Identification of antigenic sites mediating antibody-dependent enhancement of feline infectious peretonitis virus infectivity

Antigenic relationship of the feline infections peritonitis virus to coronaviruses of other species

Sequence of FIPV type I

Immunologic phenomena in the effusive form of feline infectious peritonitis

Pathogenic differences between various feline coronavirus isolates

Pathogenicity studies of feline coronavirus isolates 79-1146 and 79-1683

The predicted primary structure of the peplomer protein E2 of the porcine coronavirus transmissible gastroenteritis virus

Antigenic homology among coronaviruses related to transmissible gastroenteritis virus

Immunization against feline coronaviruses

Coronaviruses: structure and genome expression

Early death after feline infectious peritonitis virus challenge due to recombinant vaccinia virus immunization

Primary structure of the membrane and nucleocapsid protein genes of feline infectious peritonitis virus and immunogenicity of recombinant vaccinia viruses in kittens

Genomic organization and expression of the 3' end of the canine and feline enteric coronaviruses

Patterns of amino acids near signal-sequence cleavage sites

Antibody-mediated enhancement of disease in feline infectious peritonitis: comparisons with dengue hemorrhagic fever

Multiple mechanisms of protein insertion into and across membranes

This work was supported by private contributions of Ajinomoto General Foods, Inc., Japan.

Authors' address: Dr. T. Hohdatsu, Department of Veterniary Infectious Diseases, School of Veterinary Medicine and Animal Sciences, Kitasato University, Towada, Aomori 034, Japan.Received September 5, 1994