Molecular characterization of a major outer capsid protein encoded by the Threadfin aquareovirus (TFV) gene segment 10 (S10)

Page created by Terrence Stevens
 
CONTINUE READING
Arch Virol (2005) 150: 2021–2036
DOI 10.1007/s00705-005-0550-9

               Molecular characterization of a major outer
                capsid protein encoded by the Threadfin
               aquareovirus (TFV) gene segment 10 (S10)

                    E. K. Seng1 , Q. Fang2 , Y. M. Sin3 , and T. J. Lam1
    1 Department   of Biological Sciences, National University of Singapore, Singapore
                   2Wuhan   Institute of Virology, CAS, Wuhan, P.R. China
                    3 Singapore Fish Breeding and Immunization Center,
                         Teo Way Yong & Sons, Pte. Ltd., Singapore

                    Received December 27, 2004; accepted April 4, 2005
                   Published online June 3, 2005 
                                                 c Springer-Verlag 2005

Summary. Genome segment 10 (S10) of Threadfin aquareovirus (TFV) was
cloned, sequenced, analyzed and found to be 987 bp long encoding a protein of
298 aa with a predicted molecular mass of 32.0 kDa. The TFV S10 gene possesses
terminal motifs, (5 GTTTTA and ATTCATC 3 ) which are also conserved in the
S6 and S11 TFV gene segments. Sequence comparison revealed that the TFV
S10 gene was similar to the Striped bass reovirus (SBR) VP7 outer capsid protein
(OCP). A conserved putative zinc-finger motif, CCHC, present in the mammalian
reovirus (MRV) δ3 protein, was identified in TFV and other aquareovirus VP7 pro-
tein. Phylogenetic analysis of the TFV VP7 protein indicated that TFV is closely
related to SBR and Chum salmon reovirus (CSV) and possibly belong to the same
species Aquareovirus A as SBR and CSV. The TFV VP7 protein was expressed in
E. coli, purified and injected into mice. Serum specific antibodies were generated,
however, the serum showed weak neutralizing activity. In contrast, co-incubation
of this serum with another serum obtained from mice immunized with another
OCP encoded by the TFV S6 gene segment resulted in a highly elevated antibody
neutralization titer.

                                      Introduction
Members of the genus Aquareovirus, a relatively new genus included in the family
Reoviridae, contain viruses that infect fish and shellfish [44] and are among

    The nucleotide sequence data reported in this paper have been deposited to the GenBank
database under accession no. AY236219.
2022                                E. K. Seng et al.

one of the more frequently isolated fish viruses [19]. Like all the members of
the family Reoviridae, the genome of aquareovirus is composed of segmented
double-stranded RNA (dsRNA). The fish aquareovirus genome has 11 seg-
ments of dsRNA (S1–S11), similar to isolates of the genus Rotavirus, family
Reoviridae [24]. Using cryoelectron microscopy, Shaw et al. [39] revealed that
the aquareovirus genome is contained in a core surrounded by a double-layered
icosahedral capsid (coat protein) that physically resembles the capsids of
mammalian orthoreovirus (MRV), family Reoviridae. In the past, molecular
characterization of aquareoviruses was achieved by comparing the relative mo-
bilities of the dsRNA segments of different isolates, reciprocal RNA–RNA hy-
bridization and serological tests [5, 6, 17, 18, 34, 41]. These studies classified
aquareovirus isolates into six genogroups, Genogroups A–F. Recently, the In-
ternational Committee on Taxanomy of Viruses (ICTV) revised the old classifi-
cation of aquareovirus isolates from Genogroups A–F to species: Aquareovirus
A to F (ARV A–F), and this includes various tentative species [25]. The type
species is represented by Striped bass reovirus (SBR), species ARV-A. Other
members of the species ARV-A, are Chum salmon reovirus (CSV), Angelfish
reovirus (AFR) and many others. The species ARV-B is represented by Coho
salmon reovirus (CSR) and six other isolates. Species ARV-C has one member
the Golden shiner virus (GSV). Members of the species Aquareovirus D, is
represented by Channel catfish reovirus (CRV) while the species Aquareovirus
E contains the Turbot reovirus (TRV). The last species, Aquareovirus F, has
two members, Chum salmon reovirus (PSR) and Coho salmon reovirus (SSR)
[25].
    Aquareoviruses grow in various fish cell lines and produce a plaque-like cyto-
pathic effect (CPE) during the course of infection. Most aquareoviruses have been
isolated from seemingly healthy fish but some have been recovered from diseased
fish [19]. The typical clinical symptom of aquareovirus infection is hemorrhage
in the abdominal area, followed by lethargic behavior, which eventually results
in death [10, 38]. Hence, the diseases caused by such aquareovirus isolates are
a potential threat to the aquaculture industry. For example, Grass carp reovirus
(GCRV) has been reported to cause severe hemorrhages in fingerlings and yearling
grass carp, and can cause up to 80% mortality [16].
     In 1998, a reo-like virus was isolated from homogenates of brain, eye, liver,
kidney and spleen tissues from diseased threadfin fish (Eleutheronema
tetradactylum) collected from a fish farm in Singapore [4]. The virus was propa-
gated on bluegill fry fish cell line (BF-2), then further characterized and found to
belong to the genus Aquareovirus and was designated as Threadfin aquareovirus
(TFV) [38]. Experiments revealed that TFV is able to infect threadfin fingerlings
and cause 100% mortality and is able to infect sea bass fingerlings causing
60% mortality [38]. Therefore, physical, biochemical and molecular character-
ization of each new aquareovirus isolate is vital in order to quickly identify
the virus so that countermeasures can be taken to prevent the spread of the dis-
ease. As of today, no commercial vaccine or drug against aquareovirus has been
reported.
Characterization of TFV VP7 capsid protein and neutralization studies             2023

    In lieu of the lack of such knowledge on fish Aquareovirus, we have cloned,
characterized, analyzed and investigated the antigenicity of the TFV outer coat
protein(s) which maybe suitable candidates for vaccine development.

                                 Materials and methods
                                       Viruses and cells
TFV was propagated in BF-2 fish cell line as previously described [38]. Guppy reovirus (GPV)
and Grass carp reovirus (GCRV) was propagated and purified as previously described [36].
TFV infectivity was assayed using 96-well microtiter plates (IWAKI) of BF-2 cells. Virus
titer was determined using the 50% tissue culture infective dose (TCID50 ml−1 ) assay [31]
with endpoints calculated by the method of Reed and Muench [33].

                     Virus purification and extraction of genomic dsRNA
TFV particles were obtained from the supernatant of virus infected BF-2 cells and purified
by sucrose gradient centrifugation using the method of Seng et al. [38]. Viral double stranded
RNA (dsRNA) was then extracted from TFV particles as previously described [36] and stored
at −80◦ C until use.

                      Cloning and sequencing of TFV genome segments
Synthesis of cDNA from TFV dsRNA was carried out according to the method of Lambden
and Clarke [14] with slight modifications. A detailed protocol is described by Seng et al. [37].
After screening the cDNA generated by EcoRI endonuclease digestion and sequencing the
inserts, several clones containing the complete full-length sequence (987 bp) of the TFV S10
gene was identified. The 987 bp TFV S10 gene segment sequences was then submitted to
GenBank and was assigned an accession no. AY236219. The complete full-length sequence
of another TFV S6 gene (2056 bp) was also submitted to GenBank and was assigned an
accession no. AY235428.

                 Analysis of nucleotide and deduced amino acid sequences
The NCBI BLAST N program was performed to seek for sequences similar to the TFV
sequences. Further, sequence analysis was carried out using two computer programs,
DNASTAR (Lasergene, Inc.) and DNASIS ver.2.5 (Hitachi Software Engineering Company,
Ltd). Multiple sequence alignment of the TFV S10 gene (AY236219) with other S10 gene
sequences from various isolates of aquareoviruses: CSV, GSV, GCRV, SBR (GenBank ac-
cession nos. AF418303, AF403407, AF236688, U83396) and Coho salmon aquareovirus
(CSR) (GenBank accession no. U90430), and the S4 gene sequence of mammalian
orthoreovirus serotype 3 (MRV) (GenBank accession no. NC004276) was performed using
CLUSTAL W [42]. Hydropathy plots of the deduced amino acid of the aquareovirus genome
segment 10 and MRV S4 genome segment were carried out according to the method of
Kyte and Doolittle [12] using the DNASTAR program. Phylogenetic analysis was performed
using the maximum parsimony method as implemented in the phylogeny inference package,
PHYLIP [8]. The phylogenetic tree was constructed using Treeview 1 [29]. Bootstrap analysis
was performed using 1000 data re-samplings.

              Subcloning of TFV S10 genome segment into expression plasmid
The ORF of the S10 genome segment of TFV was amplified from purified plasmid harboring
the complete TFV S10 nucleotide sequence using appropriate linker primers and Expand
2024                                      E. K. Seng et al.

High Fidelity Taq DNA polymerase (Roche). The linker primer sequences are: 5 GAC
GACGACAAGATGGAGACCAAACCAATTC 3 , upstream (sense) primer, and 5 GAGGA
GAAGCCCGGTCACGGCAATGGGTTGGGC AG 3 downstream (antisense) primer. PCR
was carried out as follows: one cycle of denaturation (94 ◦ C, 2 min) followed by 30 cycles
(94 ◦ C, 30 s); (55 ◦ C, 30 s); (72 ◦ C, 1 min) and one cycle (72 ◦ C, 5 min). Amplified DNA was
electrophoresed on agarose gel, correct band was excised and purified using GFX PCR DNA
& Gel Band Purification Kit (Amersham), and cloned into the pET-30 Ek/LIC plasmid vector
(Novagen) and transformed into NovaBlue E. coli (Novagen). Transformed E. coli harboring
recombinant plasmids with correct insert was screened by restriction enzyme digestion. Three
recombinant plasmids were amplified, purified and the corresponding insert sequences were
checked by DNA sequencing. A recombinant plasmid with the correct in frame insert was
then transformed into the expression host, E. coli BL-21 (DE3) (Novagen) and transformants
were selected by colony PCR.

               Expression and identification of recombinant TFV VP7 protein
Transformed E. coli BL-21 (DE3) containing the insert was grown LB-Kan (50 µg ml−1 ) and
protein expression was induced by the addition of IPTG (Promega) to a final concentration
of 1 mM. Protein expression was checked using SDS-PAGE gel electrophoresis using 12%
acrylamide gel [13] and protein bands were stained using 1% (w/v) Coomassie brilliant
blue R-250 and destained using a 40% methanol, 7% acetic acid solution. Solubility of
the recombinant protein was determined as detailed in the QIA expressionistTM handbook
(Qiagen). Three hours after induction, induced bacterial cells were pelleted by centrifugation at
7000 × g for 20 min, 4 ◦ C and kept at −80 ◦ C until further purification. The TFV recombinant
outer capsid protein (rVP7) was purified under denaturing conditions using Ni-NTA beads
(Qiagen) followed by serial dialysis in decreasing urea concentrations of 6 M, 4 M, 2 M, 1 M
and 0.5 M Urea containing 500 mM NaCl, 20% glycerol and 20 mM Tris-HCL, pH 7.4 to allow
proteins to refold. A final dialysis step in 1X PBS was performed to remove remaining traces
of urea and the refolded protein was concentrated by dialysis using PEG-30 000 (Sigma).
Protein concentration was measured using the BioRad Protein Assay Kit (BioRad). rVP7
was identified by Western blot analysis [43] using a mouse anti-histidine serum (Roche)
and an alkaline phosphatase labeled goat anti-mouse IgG (H + L) conjugate (Sigma), with
NBT/BCIP (Boehringer, Manheim, GmbH, Germany) as the substrate. In another blot, the
primary antibody, mouse anti-rVP7 serum (prepared in this study) was used.

                    Generation of anti-VP7 polyclonal antiserum in mice
Nine female adult Swiss Albino mice (20–25 g) were purchased from the Animal Holding Unit
(AHU), National University of Singapore. Purified rVP7 (50 µg) was mixed with Freund’s
complete adjuvant (FCA) in a 1:1 (v/v) ratio to form a 0.5 ml emulsion. Three mice were
each injected intraperitoneally (i.p.) with this protein-adjuvant mixture. The other three mice
received only FCA and the remaining three mice did not receive any injection (act as controls).
Two weeks later, a booster injection which consist a mixture of purified rVP7 (50 µg) and
Freund’s incomplete adjuvant in a 1:1 ratio was similarly administered. A third injection
comprising of only purified rVP7 (50 µg) in PBS was administered two weeks after the
second injection. The mice were then bled seven days after the last injection. Blood was left
to clot overnight at 4 ◦ C and serum was collected after centrifugation at 1000 × g for 10 min,
and inactivated at 56 ◦ C for 30 min, and stored at −20 ◦ C until further use.

                                    Viral neutralization test
Serum neutralization titer was determined using a modified method of Payment and Trudel
[31], which assess the neutralization of infectivity i.e., inhibition of cytopathic effect (CPE)
Characterization of TFV VP7 capsid protein and neutralization studies         2025

by serum. Serial two-fold diluted samples of mice anti-rVP7, mice anti-rVP6 serum (obtained
from mice immunized with protein encoded by the TFV S6 gene segment) [37] were incubated
with an equal volume of TFV (103 TCID50 ml−1 ) at 25 ◦ C followed by overnight incubation
at 4 ◦ C. This antiserum: virus mixture was then inoculated onto BF-2 cells grown in 96-
well microplates and incubated at 25 ◦ C and examined daily for CPE. In all the experiments,
positive control: 103 TCID50 ml−1 of TFV, and negative control: serum from control mice
was included. Neutralizing antibody (NeuAb) titers were expressed as the reciprocal value of
the highest serum dilution showing a 50% reduction of CPE.

                                Results and discussion
                           TFV S10 gene sequence analysis
The complete nucleotide sequence and deduced amino acid sequence of the S10
genome segment of TFV is shown in Fig. 1. The nucleotide sequence was derived
from three independent recombinant plasmids and represents the consensus se-
quence. The S10 gene of TFV is 987 bp long with an open reading frame (ORF)
encoding a protein of 298 amino acids in length. The first initiation codon, ATG

Fig. 1. Complete nucleotide sequence (presented in cDNA form) and deduced amino acid
sequence of the S10 RNA segment of Threadfin aquareovirus (TFV). The conserved 5 and
3 -terminal nucleotide sequences are boxed and the inverted repeats are underlined. In the
amino acid sequence, the putative zinc-finger motifs, CCHC (aa 50–71) is in italics with a
 grey background. Nucleotide and amino acid positions are numbered on the left and right
2026                                     E. K. Seng et al.

(aa 28–30), is consistent with the optimal consensus initiation sequence
(A/G)NNATG identified by Kozak [11]. The first stop codon, TAG, is located
at position 922–924 leaving 63 untranslated nucleotides (UTR) at the 3 end. The
calculated molecular mass of the deduced protein is 32.0 kDa and is in close
agreement to the molecular mass of a protein (∼33 kDa) that we previously
described in the TFV virion [38]. The terminal ends of the TFV S10 gene seg-
ment display the nucleotide sequences, 5 GTTTTA and ATTCATC 3 , motifs
which are also conserved in the TFV genome segment 6 (GenBank accession
no. AY235428) and segment 11 (GenBank accession no. AF524892), both cloned
and sequenced by our group. Further analysis revealed that the 3 -end terminal
sequence, TTCATC, is conserved in other aquareovirus S10 genome segments
such as in SBR, CSV, GCRV and GSV, and is also identical to those of the
mammalian reovirus serotype 3 (MRV), genus Orthoreovirus, another member
of the family Reoviridae. Adjacent to the 5 end terminal of the S10 TFV gene,
nucleotides CACGCC at position 12–17, was found to be complementary to its
3 end inverted repeat sequence, GGCGTG at position 971–976 (Fig. 1). The
presence of conserved terminal sequences and inverted repeats adjacent to the
end terminals have been well established to be a common feature of the genome
segments of members of the family Reoviridae [24, 25].
    BLAST search [1] of the TFV S10 gene sequence showed a high sequence
similarity to the S10 genome segment sequence of Striped bass reovirus (SBR)
and Chum salmon reovirus (CSV), which codes for the major outer capsid protein
(VP7) of the virion [2, 20, 23, 27, 40]. Hence, we deduced that the S10 gene
of TFV most likely represents the major outer capsid protein, VP7 of the TFV
virion. Sequence comparison between TFV, SBR and CSV S10 genome segment
sequences revealed that these three isolates exhibited comparable properties in

Table 1. Nucleotide and amino acid identities of the gene segment 10 of TFV, SBR, CSV,
       CSR, GCRV, GSV and gene segment S4 of mammalian orthoreovirus, MRV

Identity (%)
Virus isolate      TFV        SBR        CSV         CSR        GCRV         GSV         MRV

TFV                           75.7       64.6        36.5       23.3         24.5        21.0
SBR                86.3                  68.8        36.1       22.7         23.4        20.5
CSV                78.9       80.6                   34.5       22.3         24.3        21.0
CSR                32.7       32.3       30.3                   24.0         23.1        20.3
GCRV               13.7       13.4       12.6        15.5                    90.6        21.7
GSV                13.4       13.4       13.4        15.2       95.7                     21.0
MRV                11.7       10.4       11.0        13.9       12.6         12.6
    Deduced nucleotide sequence identities are in the upper right portion. Deduced amino acid
sequence identities are in the lower left portion. Identities are in percentage (%). Percentage
identity values were determined using the CLUSTAL software in DNASTAR (Lasergene
Inc.).Abbreviations: TFV, Threadfin reovirus; SBR, Striped bass reovirus; CSV, Chum salmon
reovirus; CSR, Coho salmon reovirus; GCRV, Grass carp hemorrhage reovirus; GSV, Golden
shiner reovirus, MRV, mammalian orthoreovirus serotype 3
Characterization of TFV VP7 capsid protein and neutralization studies    2027

terms of total length of the gene segment and the protein products it encodes for,
a protein of ∼32 kDa in molecular mass. Comparison of the nucleotide sequence
and predicted amino acid sequence of the TFV S10 with other reported isolates
of aquareovirus and the mammalian reovirus (MRV) δ3 protein revealed that the
TFV VP7 protein showed high amino acid sequence identity to VP7 protein of
SBR (86.3%), CSV (78.9%), moderate identity to the CSR VP7 protein (32.8%),
relatively low identity to GCRV, GSV VP7 protein, and MRV δ3 protein (encoded
by the MRV S4 gene segment) with values of only 13.7%, 13.4%, and 11.7%
respectively (Table 1). Although the deduced protein sequence from CSR, GCRV
and GSV S10 gene and MRV S4, showed moderate to low homology to the
deduced protein sequence of TFV S10 (Table 1; Fig. 2), hydrophathy plots (data
not presented) calculated using the method of Kyte and Doolittle [12], clearly
showed that the predicted proteins encoded by all the Aquareovirus isolates, TFV,
SBR, CSV, GCRV, and GSV, as well as the δ3 protein of mammalian reovirus
(MRV) showed very similar profiles especially in the N-terminal part of the
protein, with four domains in the order hydrophilic-hydrophobic-hydrophilic-
hydrophobic (data not presented). Hence, it is highly suggestive that the VP7
protein from these aquareoviruses maybe analogues to the δ3 protein of MRV.
This observation is in agreement to Attoui et al.’s [2] finding which showed that
the SBR and CSV gene segment 10 (S10), encoding the VP7 protein is most likely
the equivalent to the MRV δ3 protein encoded by the S4 gene segment of MRV.
    Additional confirmation of the relatedness of the protein encoded by TFV
S10 genome segment to the MRV δ3 protein was gained from the discovery of a
zinc-finger motif found in both proteins (Fig. 2). Analysis of the deduced amino
acid sequence of TFV segment 10 revealed the presence of a CX2 CX15 HX1 C
motif (Fig. 1, shaded; Fig. 2) within the N-terminal, residues 50–71, analogous
to the zinc-finger motif (CCHC), CX2 CX15 HHX1 C (residue 51–73) identified
by Mabrouk and Lemay [22] in the mammalian reovirus (MRV) δ3 protein. The
CX2 CX15 HX1 C motif identified in TFV was also found at the same location
in the deduced SBR and CSV VP7 protein with the exception that in the CSV
VP7 protein, the supposed cysteine (C) residue at position 53 was substituted
for a tyrosine (Y) instead (Fig. 2). This discrepancy could be unchecked errors
encountered during the sequencing of CSV. Further analysis revealed that the
deduced VP7 protein encoded by the GCRV, GSV and CSR S10 gene also posses
a putative zinc-finger motif, CX2 CX16 HX1 C.
    By analogy, the CCHC zinc-finger motif, found in all the aquareovirus isolates
as well as in the MRV protein, is similar to that reported in retroviral nucleocapsid
proteins, which proposed function was to direct the specific packaging of retroviral
RNA [7, 9]. Recently, modifications of zinc-binding residues inside the conserved
CCHC motif of human immunodeficiency virus Type 1 NCp7 into CCHH, induced
a complete loss of infectivity [32]. Studies conducted by Mabrouk and Lemay [22]
indicated that the zinc-finger motif in MRV δ3 protein is important in conferring
stability to the protein. So far, the exact function(s) of the CX2 CX15−16 HX1 C
motif found in all the aquareovirus isolates is unknown and its significance in
viral replication or pathogenesis remains to be elucidated.
Fig. 2. Multiple alignment of deduced amino acid sequence of protein encoded by S10 RNA
segment of Threadfin aquareovirus (TFV) with deduced proteins from other aquareovirus
S10 RNA segment; Striped bass reovirus (SBR), Chum salmon reovirus (CSV), Coho salmon
reovirus (CSR), Golden shiner reovirus (GSV), grass carp reovirus (GCRV) and the δ3 protein
from mammalian reovirus Dearing strain (serotype 3) (MRV). See Methods section for
accession numbers of each isolate. Amino acid positions for each individual sequence are
numbered on the right. Identical amino acids are indicated by an asterisk. The putative zinc-
finger domain (CCHC form), identified by Mabrouk and Lemay [22] within the mammalian
reovirus δ3 protein, is boxed with a continuous line (amino acids 51–73) and the corresponding
motif within the Threadfin aquareovirus protein is boxed with a dotted line (amino acids
50–71). Potential N-linked glycosylation sites, Nx1 S/T, where x represent any amino acid
residue except proline (P), are underlined and can be found in all aquareovirus isolates except
                 CSR. No glycosylation sites are present in MRV δ3 protein
E. K. Seng et al.: Characterization of TFV VP7 capsid protein            2029

Fig. 3. Phylogenetic tree of the gene segment 10 of TFV, SBR, CSV, CSR, GCRV, GSV
and gene segment S4 of mammalian orthoreovirus (MRV). Phylogeny was inferred from the
complete amino acid sequences of the indicated viruses using the PHYLIP suite program [8].
Bootstrap confidence levels following 1000 replicates are listed next to each branch node.
       The scale bar is proportional to genetic distance. ARV – Aquareovirus species

    The presence of the putative zinc-finger motif in all deduced protein of the
S10 segment of all the aquareovirus isolates and MRV δ3 protein illustrates that
the VP7 protein coded by the S10 genome segment of aquareoviruses may be
functionally and structurally conserved. Hence, we concluded that the TFV S10
gene segment, encodes a viral protein, VP7, which is alike the δ3 protein of MRV
that also represent the outermost virus capsid protein. The conservation of the 3
ATTCATC gene sequence between TFV S10 and MRV S4 partly implied that both
viruses might have evolved from a common ancestral precursor. A phylogenetic
analysis we conducted supported this claim (Fig. 3) as all aquareovirus isolates
were segregated into three separate clusters with MRV forming the fourth cluster.
The separation of MRV could stem from the fact that the MRV genome consist of
only 10 dsRNA genome segments as oppose to 11 in aquareoviruses. Among the
aquareoviruses, the TFV VP7 protein is more closely related to SBR and CSV VP7
protein, both belonging to the genogroup A (species, ARV-A). CSR of genogroup
B (ARV-B) formed an individual clade while GCRV and GSV, both of genogroup
C (ARV-C) formed the other clade. This observation is well received as earlier
classification methods of aquareovirus isolates by way of reciprocal RNA–RNA
hybridization and nucleotide analysis revealed a similar structure of grouping.
Thus, TFV should be placed into genogroup A (species ARV-A) due to its close
relationship in terms of the VP7 protein sequence of SBR and CSV. Additional
evidence for the placement of TFV into the same species as SBR and CSV was
2030            E. K. Seng et al.: Characterization of TFV VP7 capsid protein

gained in our previous results that showed that the protein encoded by the TFV
S6 gene segment showed high sequence identity to the VP4 protein encoded by
CSV S6 gene segment [37].
    Further thorough sequence analysis of the TFV VP7 protein sequence revealed
that there are three potential N- linked glycosylation sites located at aa residues
245–247 (NQS), 269–271 (NLS) and 276–278 (NKS) (Fig. 2). In SBR and CSV,
similar potential N-linked glycosylation sites were discovered. However, in GCRV
and GSV, only one potential N-linked glycosylation (aa 244–246) site was found
while none was found in CSR (Fig. 2).
    Generally, the carbohydrate groups of glycoproteins have been known to
facilitate proper protein folding, confer protease resistance, maintain structural
integrity and conformational stability and affect surface charges and water binding
capacity [30]. In recent years, much interest has cumulated investigations into the
importance of oligosaccharide chains of many glycoproteins in biological recog-
nition, protein trafficking, host-pathogen, and cell–cell interaction. In Rotavirus,
the VP7 protein, an outer capsid protein, analogous to the SBR VP7 and the
TFV VP7 protein, has been shown to be glycosylated [26]. The major function of
carbohydrates on the rotavirus VP7 is to facilitate correct disulfide bond formation
and protein folding [26]. In addition, glycosylation of the rotavirus VP7 has also
been shown to affect the antigenic specificity of the protein [3, 15]. As such, we can
only speculate that due to the close percentage similarity between the TFV VP7
and SBR VP7 protein, in which the latter has been demonstrated to be glycosylated
[35] it is most likely that the TFV VP7 is also glycosylated. However, we have yet
to provide evidence for the presence of oligosaccharides in the TFV VP7 protein.
It will be interesting to see whether the TFV and SBR VP7 protein show similar
functional properties as the rotavirus VP7 protein in future experiments.

                TFV VP7 protein expression and immunogenicity
Thus far, sequence determination and analysis of genome segment 10 (S10) of
TFV only revealed that this gene codes for an outer capsid protein of the virus. In
order to verify that the S10 gene codes for a protein present on the TFV virion,
the ORF of the S10 TFV gene was cloned into the pET-30 expression vector
and a double tagged recombinant outer capsid protein, rVP7, harboring histidine
(His) and S-Tag sequences was expressed in E. coli. SDS-PAGE analysis revealed
that 1 h and 3 h post induction, the rVP7 protein was overexpressed (Fig. 4A).
Overexpressed rVP7 protein accumulated in inclusion bodies and the yield of
the rVP7 was 40.0 mg L−1 (∼32% of total bacterial cellular protein content). Im-
munoblot analysis using monoclonal mouse anti-His antiserum (Fig. 4B) showed
that the rVP7 protein expressed has a molecular mass of ∼36 kDa, much larger
than the native TFV VP7 protein (∼33 kDa), but is consistent with the expected
fusion protein. The higher molecular mass of rVP7 protein was attributed by the
extraneous polypeptide contributed by His and S-tags of the vector.
    To study the authenticity and antigenicity of the rVP7 protein produced in
E.coli, antiserum against the rVP7 was raised in mice and its reactivity with the
Fig. 4. A) SDS-PAGE analysis of recombinant TFV VP7 outer capsid protein expressed in
E.coli. Lane 1, control (no IPTG); 2, 1 h after IPTG induction; 3, 3 h after IPTG induction; M,
molecular weight marker (BioRad). The rVP7 protein is indicated by an arrow. B) Western
blot analysis of purified TFV rVP7 with mouse anti-HIS serum. 1, purified TFV rVP7 protein;
M, molecular weight marker (BioRad). Arrowhead indicates the position of the TFV rVP7
protein. C) Western blot analysis of purified TFV, GPV, GCRV and E.coli expressed TFV
rVP7 with mouse anti-rVP7 serum. Preparations of purified viruses were separated on SDS-
PAGE and transferred to a nitrocellulose membrane. The membrane was then treated with
mouse anti-rVP7 serum generated as described in this manuscript. 1, purified TFV virus; 2,
purified GPV virus; 3, purified GCRV virus; 4, purified and dialyzed rVP7; M, molecular
weight marker (BioRad). Arrowhead indicates the position of the VP7 outer capsid protein of
TFV virion. Arrow indicates the position of the purified full-length product of the TFV rVP7
         protein. Unfilled arrow indicates a probable dimer form of the rVP7 protein
2032                                E. K. Seng et al.

TFV, Guppy reovirus (GPV) and Grass carp reovirus (GCRV) VP7 protein was
assayed by immunoblotting. As shown in Fig. 4C, both the native TFV VP7 and
rVP7 protein ( , ; in Fig. 4C, respectively) was detected by the mice antiserum.
In addition, a weak band at ∼72 kDa ( , in Fig. 4C) was also detected. This
product could possibly be a dimer form of the rVP7 protein. Dimer formation has
been observed in the δ3 protein of MRV [28]. The dimer of the MRV δ3 protein
was proposed by Olland and colleagues [28] to be the likely protein form that
is involved in binding dsRNA. To our knowledge, this is the first report of the
possible formation of a ‘δ3-like’ dimer in an aquareovirus protein but its function
remains to be investigated. In Fig. 4C, it was observed that the mice anti-rVP7
serum also reacted to the GPV VP7 protein (Lane 2), but did not react with any of
the proteins of the GCRV virus. This clearly showed that TFV, a seawater isolate,
is antigenically related to GPV, a freshwater isolate but is not related to GCRV,
another freshwater isolate. We next investigated the immunogenicity of the mice
anti-rVP7 serum that we generated by conducting viral neutralization tests. A low
TFV neutralizing antibody (NeuAb) titer: 16, 32, 32 was recorded from serum
recovered from the three mice injected with rVP7 as compared to the NeuAb
titer in control mice (≤8, ≤8, ≤8). This indicated that TFV was ineffectively
neutralized. This result is in agreement with Lupiani et al.’s [21] observation
whereby antiserum from rabbit immunized with recombinant SBR VP7 outer
capsid protein expressed in a baculovirus system also showed no neutralizing
activity. Collectively, these two independent results suggest that the outer capsid
protein, VP7, encoded by TFV S10 genome segment or other aquareoviruses may
not be the major neutralizing antigen.
     In this study, we also investigated the neutralizing activity of antiserum ob-
tained from mice immunized with another TFV outer capsid protein, VP4, en-
coded by the S6 genome segment, which our group recently cloned, sequenced
and expressed in E. coli [37]. Molecular analysis revealed that the TFV S6
gene sequence codes for a protein analogous to the VP4 protein of SBR and
CSV and the MRV µ1 protein [2, 37]. In this study we found that antiserum
recovered from three mice injected with recombinant VP4 protein (rVP4) also
showed low neutralizing activity, NeuAb titer: 16, 16, 32. Two possible reasons
for the low neutralization activities of the mice antiserum against rVP7 and
rVP4 could be due to the loss of important neutralizing epitopes as a result of
incorrect protein folding during our protein purification procedure or the ab-
sence of glycosylation as we used E. coli derived proteins for the production of
antibodies.
     On the basis of the low neutralizing titer of both mice antiserum, we ra-
tionalized that if we combined both the antiserum against rVP7 and rVP4, the
neutralization titer of the antibodies should be in the range of 32 to 64. However,
when antiserum from mice injected with rVP7 and antiserum from mice injected
with rVP4 were mixed in a 1:1 vol:vol ratio, a highly elevated NeuAb titer
(128, ≥256; ≥256) was observed. To our knowledge, this is the first report
showing that antibodies raised against outer capsid proteins of an aquareovirus
isolate did not possess any neutralizing properties when applied individually, but
Characterization of TFV VP7 capsid protein and neutralization studies    2033

exhibited a high NeuAb activity when used in concert. This unique phenomenon
of synergism between two antibodies generated from two different viral outer
capsid proteins that resulted in neutralization of virus certainly warrants further
detailed investigation.
     A probable mechanism to explain our observations could be indirectly gained
from previous studies using monoclonal antibodies (mAb) directed against the
mammalian orthoreovirus serotype 3 (MRV) capsid proteins, δ3 and µ1, which
showed that some antibodies are non-neutralizing but yet protective, i.e., able
to hinder viral replication [45]. Virgin et al. [45] showed that non-neutralizing
monoclonal antibody (mAb) directed against the MRV δ3 outer capsid protein
inhibits intracellular proteolytic uncoating of the reovirus outer capsid while
mAb directed against MRV µ1 outer capsid protein inhibit membrane penetra-
tion by infectious subviral particles (ISVP) and as a result block MRV repli-
cation. Perhaps, antiserum against the TFV VP7 and VP4 outer coat protein
may employ similar mechanisms but must complement each other in order
to totally curb TFV infection. Nonetheless, the exact mechanism of how
antibody-mediated inhibition of TFV replication remains to be verified in future
experiments.
     In summary, we have cloned, sequence and analyzed the TFV gene segment
10 (S10) which codes for the TFV outer viral coat protein (VP7). Based on amino
acid sequence similarities and phylogenetic analysis we demonstrated that TFV
likely belongs to the species, ARV-A, of which SBR and CSV are representative
members. In addition, we demonstrated that the serum obtained from mice injected
with recombinant TFV VP7 protein had a low neutralizing antibody titer. In
contrast, by combining the mice anti-VP7 and VP4 serum, a highly elevated
neutralizing antibody titer was achieved. A direct implication for this observation
is that multiple antibodies directed against several viral coat proteins are required
to effectively neutralize the TFV virus. This unique observation, of synergism
between two antibodies to neutralize a virus would definitely be interesting to
investigate in future experiments.
     While considerable fundamental and applied research has been carried out on
human isolates from the genus Orthoreovirus and Rotavirus, very little progressive
research has been done on members of the genus Aquareovirus. As a result, it
has lead to the lack of basic fundamental knowledge of aquareovirus genetics,
although much can be inferred from works already done on mammalian reoviruses.
Hence, our work on cloning and expression of recombinant aquareovirus outer
capsid protein will hopefully provide reagents for future works like structural
function related studies, production of monoclonal antibodies, determination of
epitopes responsible for virus neutralization and possibly explore the mechanisms
of antibody mediated virus neutralization.

                                Acknowledgments
This work was supported by a research grant (R-154-000-068-112) from the National
University of Singapore, Singapore.
2034                                    E. K. Seng et al.

                                       References
 1. Altschul SF, Madden TL, Schiffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ (1997)
    Gapped BLAST and PSI-BLAST: a new generation of protein database search programs.
    Nucleic Acids Res 25: 3389–3402
 2. Attoui H, Fang Q, Jaafar FM, Cantaloube JF, Biagini P, de Micco P, de Lamballerie X
    (2002) Common evolutionary origin of aquareoviruses and orthoreoviruses revealed by
    genome characterization of Golden shiner reovirus, Grass carp reovirus, Striped bass
    reovirus and Golden ide reovirus (genus Aquareovirus, family Reoviridae). J Gen Virol
    83: 1941–1951
 3. Caust J, Dyall-Smith ML, Lazdins I, Holmes IH (1987) Glycosylation, an important
    modifier of rotavirus antigenicity. Arch Virol 96: 123–134
 4. Chang SF, Ngoh-Lim GH, Kueh LFS, Qin QW, Seng EK, Sin YM (2002) Initial
    investigations into two viruses isolated from marine food fish in Singapore. Vet Rec
    150: 15–16
 5. Dopazo CP, Toranzo AE, Samal SK, Roberson BS, Baya A, Hetrick FM (1992) Antigenic
    relationships among rotaviruses isolated from fish. J Fish Dis 15: 27–36
 6. Dopazo CP, Bandin I, Rivas C, Cepeda C, Barja JL (1996) Antigenic differences
    among aquareoviruses correlate with previously established genogroups. Dis Aquat Org
    26: 159–162
 7. Dupraz P (1990) Point mutation in the proximal Cys-His box of Rous sarcoma virus
    nucleocapsid protein. J Virol 64: 4978–4987
 8. Felsenstein J (1995) PHYLIP (Phylogentic Inference Package), version 3.57c.
    Department of Genetics, University of Washington, Seattle, WA, USA
 9. Green LM, Berg JM (1989) A retroviral Cys-Xaa2-Cys-Xaa4-His-Xaa4-Cys peptide
    binds metal ions: spectroscopic studies and proposed three-dimensional structure. Proc
    Natl Acad Sci USA 86: 4047–4051
10. Jiang Y, Ahne W, Li Y, Ogawa M (1991) Isolation of an aquareovirus from common carp
    (Cyprinus carpio) in the PR China. In: Proceedings Second International Symposium
    on Viruses of Lower Vertebrates, Oregon State University Press, Corvallis, USA,
    pp 287–292
11. Kozak M (1981) Possible role of flanking nucleosides in recognition of the AUG initiator
    codon by eukaryotic ribosomes. Nucleic Acids Res 6: 5233–5250
12. Kyte J, Doolittle RF (1982) A simple method for displaying the hydropathic character of
    a protein. J Mol Biol 157: 105–132
13. Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of
    bacteriophage T4. Nature 227: 680–685
14. Lambden PR, Clarke IN (1995) Cloning of viral double-stranded RNA genomes by
    single primer amplification. In: Adolph KW (ed) Viral gene techniques. Academic Press,
    New York, pp 359–372
15. Lazdins I, Coulson BS, Kirkwood C, Dyall-Smith M, Masendycz PJ, Sonza S, Holmes
    IH (1995). Rotavirus antigenicity is affected by the genetic context and glycosylation of
    VP7. Virology 209: 80–89
16. Li J, Wang T, Yi Y, Liu H, Lu R, Chen H (1997) A detection method for grass carp
    hemorrhagic virus (GCRV) based on a reverse transcription-polymerase chain reaction.
    Dis Aquat Org 29: 7–12
17. Lupiani B, Hetrick FM, Samal SK (1993) Genetic analysis of aquareoviruses using
    RNA–RNA blot hybridization. Virology 197: 475–479
18. Lupiani B, Hetrick FM, Samal SK (1994) Identification of the angelfish, Pomacanthus
    semicirculans aquareovirus as a member of aquareovirus genogroup A using reciprocal
    RNA–RNA blot hybridization. J Fish Dis 17: 667–672
Characterization of TFV VP7 capsid protein and neutralization studies         2035

19. Lupiani B, Subramanian K, Samal SK (1995) Aquareovirus. Annu Rev Fish Dis
    5: 175–208
20. Lupiani B, Reddy SM, Samal SK (1997a) Sequence analysis of genome segment 10
    encoding the major outer capsid protein (VP7) of genogroup B aquareovirus and its rela-
    tionship with the VP7 protein of genogroup A aquareovirus. Arch Virol 142: 2547–2552
21. Lupiani B, Reddy SM, Subramanian K, Samal SK (1997b) Cloning, sequence analysis
    and expression of the major outer capsid protein gene of an aquareovirus. J Gen Virol
    78: 1379–1383
22. Mabrouk T, Lemay G (1994) Mutations in a CCHC zinc-binding motif of the reovirus
    δ3 protein decreases its intracellular stability. J Virol 68: 5287–5290
23. McPhillips TH, Dinan D, Subramanian K, Samal SK (1998) Enhancement of
    aquareovirus by treatment with proteases: Mechanism of action. J Virol 72: 3387–3389
24. Mertens P (2004) The dsRNA viruses. Virus Res 101: 3–13
25. Mertens PPC, Arella M, Attoui H, and 41 other authors (2000) Family Reoviridae. In:
    Van Regenmortel MHV, Fauquet CM, Bishop DHL, Carstens EB, Estes MK, Lemon SM,
    Maniloff J, Mayo MA, McGeoch DJ, Pringle CR, Wickner RB (eds) Virus taxonomy.
    Seventh Report of the International Committee for the Taxonomy of Viruses. Academic
    Press, San Diego, pp 395–480
26. Mirazimi A, Svensson L (1998) Carbohydrates facilitate correct disulfide bond formation
    and folding or rotavirus VP7. J Virol 72: 3887–3892
27. Nason EL, Samal SK, Prasad VV (2000) Trypsin-induced structural transformation in
    Aquareovirus. J Virol 74: 6546–6555
28. Olland AM, Jane-Valbuena J, Schiff LA, Nibert ML, Harrison SC (2001) Structure
                                                                          ◦
    of the reovirus outer capsid and dsRNA binding protein δ3 at 1.8 A resolution. EMBO
    J 20: 979–989
29. Page RDM (1996) TREEVIEW: An application to display phylogenetic trees on personal
    computers. Comput Appl Biosci 12: 357–358
30. Paulson JC (1989) Glycoproteins: What are the sugar chains for? Trends Biochem Sci
    14: 272–276
31. Payment P, Trudel M (1993) Methods and techniques in Virology. Marcel Dekker,
    New York
32. Ramboarina S, Druillennec S, Morellet N, Bouaziz S, Roques BP (2004) Target specificity
    of human immunodeficiency virus type 1 NCp7 requires an intact conformation of its
    CCHC N-terminal zinc finger. J Virol 78: 6682–6687
33. Reed LJ, Müench H (1938) A simple method of estimating fifty percent endpoints. Am
    J Hyg 27: 493–497
34. Samal SK, Dopazo CP, Subramanian K, Lupiani B, Mohanty SB, Hetrick FM (1991)
    Heterogeneity in the genome RNAs and polypeptides of five members of a novel group
    of rotavirus-like viruses isolated from aquatic animals. J Gen Virol 72: 181–184
35. Samal SK, Subramanian K, Lupiani B (1995) Characterization of two aquareovirus
    proteins. Vet Res 26: 388–393
36. Seng, EK, Fang Q, Lam TJ, SinYM (2004) Development of a rapid, sensitive and specific
    diagnostic assay for fish Aquareovirus based on RT-PCR. J Virol Methods 118: 111–122
37. Seng EK, Fang Q, Sin YM, Lam TJ (2005) Molecular characterization, DNA sequence
    analysis, and expression of cDNA sequence of RNA genomic segment 6 (S6) that
    encodes a viral outer capsid protein of Threadfin aquareovirus (TFV). Virus Genes
    30: 209–221
38. Seng EK, Fang Q, Chang SF, Ngoh GH, Qin QW, Lam TJ, SinYM (2002) Characterisation
    of a pathogenic virus isolated from marine threadfin fish (Eleutheronema tetradactylus)
    during a disease outbreak. Aquaculture 214: 1–18
2036            E. K. Seng et al.: Characterization of TFV VP7 capsid protein

39. Shaw AL, Samal SK, Subramanian K, Prasad BV (1996) The structure of aquareovirus
    shows how the different geometries of the two layers of the capsid are reconciled to
    provide symmetrical interactions and stabilization. Structure 4: 957–967
40. Subramanian K, McPhillips TH, Samal SK (1994) Characterization of the polypeptides
    and determination of genome coding assignments of an aquareovirus. Virology
    205: 75–81
41. Subramanian K, Hetrick FM, Samal SK (1997) Identification of a new genogroup of
    aquareovirus by RNA–RNA hybridization. J Gen Virol 78: 1385–1388
42. Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W: Improving the sensitivity
    of progressive multiple sequence alignment through sequence weighting, position
    specific gap penalties and weight matrix choice. Nucleic Acids Res 22: 4673–4680
43. Towbin H, Staehelin T, Gordon J (1979) Electrophoretic transfer of proteins from
    polyacrylamide gels to nitrocellulose sheets: Procedures and some applications. Proc
    Natl Acad Sci USA 76: 4350–4354
44. Van-Regenmortel MHV, Fauquet CM, Bishop DHL, Carstens EB, Estes MK, Lemon SM,
    Maniloff J, Mayo MA, McGeoch DJ, Pringle CR, Wickner RB (2000) Virus taxonomy.
    Classification and nomenclature of viruses, 7th edn, Academic Press, San Diego, CA
45. Virgin HW, Mann MA, Tyler KL (1994) Protective antibodies inhibit internalization and
    uncoating by intracellular proteases. J Virol 68: 6719–6729
   Author’s address: Dr. Seng Eng Khuan, Department of Biological Sciences, National
University of Singapore, Block S1A, 05-02 Virology Lab, 14 Science Drive 4, Singapore
117543; e-mail: dbssek@nus.edu.sg
You can also read