Journal of General Virology (1994), 75, 3423-3430. Printed in Great Britain
3423
Identification of the nucleic acid binding domain of the rotavirus VP2
protein
M . Labbr, P. Baudoux, A. Charpilienne, D. P o n c e t and J. Cohen*
Institut National de la Recherche Agronomique, Laboratoire de Virologie et Immunologie MoldcuIaires,
Domaine de Vilvert, 78350 Jouy-en-Josas, France
The bovine rotavirus VP2 protein is the major component of the core and forms the most internal layer
surrounding the dsRNA genome. We have constructed
recombinant baculoviruses expressing truncated VP2
proteins. The nucleic acid binding activity of these
truncated proteins was tested by North-Western blotting
experiments with single-stranded and double-stranded
probes. The nucleic acid binding domain in VP2 was
localized between amino acids 1 to 132. Recombinant
proteins bound single-stranded and double-stranded
nucleic acids, but showed less affinity for doublestranded RNA and DNA. Interactions o f VP2 with the
genome were investigated in viral single-shelled particles
by u.v.-cross-linking. In these experiments, only VP2
protein bound the genomic RNA in purified singleshelled particles.
Introduction
single-shelled particles. We have constructed recombinant baculoviruses expressing truncated p r o d u c t s o f VP2
that were tested for their ability to bind nucleic acids in
vitro by N o r t h - W e s t e r n blotting experiments. We also
investigated the interactions o f structural proteins with
the rotavirus g e n o m e in purified particles by u.v.-crosslinking.
Rotaviruses are the m o s t i m p o r t a n t cause o f severe
gastroenteritis in h u m a n s and in animals. These viruses
belong to the Reoviridae family and their segmented
d s R N A g e n o m e is encapsidated in three layers o f
structural proteins (for a review, see Estes & Cohen,
1989). The m o s t external capsid consists o f the proteins
VP4 and VP7. The intermediate capsid is constituted by
VP6. VP2, the m a j o r c o m p o n e n t o f the core, forms a
third layer a r o u n d the genomic R N A (Labb~ et al.,
1991).
It has been suggested that VP2 is implicated in the
binding and the bending o f the g e n o m e inside viral
particles ( K a p a h n k e et al., 1986). It has been previously
shown by N o r t h - W e s t e r n blotting experiments that VP2
binds single-stranded and double-stranded nucleic acids
(Boyle & Holmes, 1986). Binding was n o t sequence
specific, but VP2 showed m o r e affinity for s s R N A than
for double-stranded nucleic acids. R e c o m b i n a n t VP2
protein, expressed in insect cells, also showed a nucleic
acid binding activity (Labb6 et al., 1991).
The gene 2 (coding for VP2 protein) o f various strains
has been sequenced ( K u m a r et al., 1989; Mitchel & Both,
1990; Ernst & Duhl, 1989; Tian et al., 1990; B r r m o n t et
al., 1992; Lindsay et al., 1994), and amino acid sequence
analysis revealed several motifs that could be responsible
for interactions with nucleic acids.
The aim o f this study was to better characterize the
binding o f VP2 with nucleic acids either in vitro with
r e c o m b i n a n t truncated VP2 or in vivo within purified
0001-2521 © 1994 SGM
Methods
Viruses and cells. The RF strain of bovine rotavirus was grown in
cultures of MA 104 cells as previously described (L'Haridon & Scherrer,
1976). Rotavirus was extracted with 'Freon 113' and purified by CsC1
density gradients (Cohen, 1977). Wild-type modified baculovirus
AcRP6-SC (Kitts et al., 1990) and Spodopterafrugiperda Sf9 ceils were
grown as described (Summers & Smith, 1987).
Construction of recombinant baculoviruses. A full-length cDNA
corresponding to gene 2 of bovine RF strain rotavirus was cloned in the
pVL941 vector as described (Labb~ et al., 1991). For the construction
of recombinant baculoviruses expressing truncated VP2 proteins,
deleted sequences of the gene 2 were subcloned in the pVL941. A first
set of recombinant transfer vectors was obtained by site-directed
mutagenesis of the gene 2 cloned in the pBS÷ plasmid (Stratagene) and
subcloning of parts of the gene. The recombinant vectors contained
different combinations of one to three sequences that corresponded
approximately to the different thirds of VP2. The oligonucleotides used
for mutagenesis are presented in Fig. 1. Start and stop codons used
for translation of truncated sequences and restriction sites used for
subcloning were inserted by the Kunkel method (1985). Plasmids pB1,
pB2, pBl2 and pB32 were obtained by mutagenesis with oligonucleotides 1, 2, 1+2 and 2+3 respectively. AT877 fragment (aminoterminal third) and CTI821 fragment (the last two thirds) were
obtained by digestion of pB 1 by BamHI or BamHI + KpnI respectively
and then subcloned in the transfer vector pVL941 iinearized by BamHI
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 18:33:00
3424
M. Labb~ and others
Oligonucleotide
Sequence
Polarity / position
Oligo I
AAT GAC GTT AAC [ T - ~ G G A
Oligo 2
TCA CTC ACA TCC[~']GGA ~Ce A e e I ~ G G A
Oligo 3
AAC TAT A T A CTT ACe ~ G q G A C
A1964
C CCG GC,A TCC A C C ~ ' ~ GAG A A A C A A GAT C A A AAC
+ / 92 to 109
A1965
c Cc(~ G(~A TCC A C C ~ ] G A G
+ / 149 to 166
A1602
C CCG OGA ~C
A1966
c CCO ~A
A1595
c cc(~ G G A 9'CC ~
A1596
c CCG ~
A1597
~I'A A G G A T C ~ ' ] G G A
A1476
CCCCd~GCTC, C A G G A T C C G G C T A T T A A A G G T T C A
A1598
T T A AC,G A T C ~ ' A ] G G G TAG CTG TCT CGG TTC
~'~
'ICe Ace ~ " ~ G A C
AGA AAT CTG
+ / 854 to 892
hAT GCA ACe
+ / 1754 to 1792
A G A AAT
G A A GTC G T A ACC GAC
G A A G A A ATT A A A ATT GCT
TCC A C C ~
GTG AAG A A A TCG ACG AAA
÷ / 863 to 889
+ / 173 to 190
+ / 197 to 214
G A A CCA A A A GAG TCA A T A
+ / 311 to 328
TCC ~ - G ] G A A CCG A G A CAG CTA CCA
+ / 395 to 412
TGT GAC T G A AGT TA
/ 1749 to 1765
+ / 1 to 16
/ 395 to 411
Fig. 1. Oligonucleotides used for site-directed mutagenesis of VP2. Oligonucleotides 1, 2 and 3 were used to introduce mutations into
gene 2 cloned in pBS vector by the method described by Kunkel (1985). Other oligonucleotides were used to amplify truncated
sequences of gene 2 by PCR. The initiation or termination codons added are boxed and restriction sites are underlined. Modified
nucleotides are shown in bold.
or BamHI + KpnI. Similarly, the AT1777 fragment (first two thirds)
and CT921 fragment (last third) were cloned in pVL941 after digestion
of pB2 by BamHI or BamHI+KpnI. The middle third (M9) was
obtained by BamHI digestion of pB12 and then subcloned in pVL941
linearized by BamHI. The fragment AN (VP2 with the middle third
deleted) was obtained by excision of the fragment NcoI of pB23 and
subcloning of the fragment BamHI-KpnI in pVL941. A second set of
mutants was obtained by amplification of gene 2 with Taq polymerase
(Promega) using oligonucleotides to introduce initiation or termination
codons and a BamHI restriction site. Gene 2 was amplified with a 3'
antisense primer (A1597) containing a stop codon and 5' sense primers
(A1964, A1965, A1595, A1596) that allowed progressive deletions of
the amino-terminal region and introduction of an ATG codon.
Oligonucleotides A1476 and A1598 were used for amplification of
sequences coding for amino acids 1 to 132. Amplified sequences were
digested by BamHI and cloned into the BamHI restriction site of the
pVL941 vector.
Recombinant baculoviruses were obtained by cotransfection of Sf9
cells with Bsu36I linearized DNA of AcRP6-SC baculovirus (Kitts et
al., 1990) and recombinant transfer vectors using lipofectin reagent
(BRL). Truncated proteins expressed in insect cells by the various
recombinant baculoviruses are shown in Fig. 2.
Proteins analysis. For production of recombinant proteins, Sf9
cells were infected by the recombinant baculoviruses at high m.o.i.
(10 p.f.u./cell) and harvested 72 h post-infection (p.i.). Proteins were
analysed by SDS~PAGE (Laemmli, 1970).
North-Western blotting analysis. Proteins separated by SDS@AGE
as described above were blotted onto PVDF membranes (Problott;
Applied biosystems) in transfer buffer (10mM-CAPS pH 11, 10%
methanol) as described by the manufacturer. PVDF membranes were
saturated for 30min in Standard Binding Buffer (SBB: 10mMTris-HCl pH 7.4, 1 mM-EDTA, 0-04 % BSA, 0.04 % polyvinylpyrrolidone, 0-04 % Ficoll). When indicated, blots were saturated with SBB
containing 250 ~tg/ml of tRNA from Eseherichia coli (Boehringer) used
as a competitor. The PVDF membranes were incubated with probes
(105 c.p.m./ml) diluted in SBB containing 50 mM-NaC1 or tRNA as
indicated. The blots were washed three times for 2 min at 25 °C, dried
and autoradiographed.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 18:33:00
RNA binding domain of rotavirus VP2
3425
P1 kinase homology
--
64
!
76
7 Helix-turn-helix
6 6 - 121
~
*
i
Char~ed helix
~
J
53
Leucine zipper
77
536-557
665
686
880
i
I
]
IrA.
..............
0
I
II
:: ...............
] ......................
i ....................
]
.......
t00
200
3(10
400
]
II
........
500
i .................
] ................
] ..........
6QO
700
8(30
Bac2AT877
Bac2AT1777
I
I i'J
;
I I~I
Bac2CT1821
N~
M~-I
11~'/
II
I ...................................................................... ~
I I
Bac2CT1921
M.,-~
Bac2M9
~88
Bac2AN
I
I:lzl=~
l
Bac2M16.9
29O
.,
I
Bac2M 16.3
51t3
Bac2M 14
M" ~=t
III
"127
Bac2M 13
Bac2A420
I
i..~?---~, I
BacRF2
I
I;~;
583
I
II
II
Fig. 2. Schematic of truncated VP2 expressed in insect cells. The top bar represents the VP2 protein sequence. Putative nucleic acid
binding motifs deduced from the amino acid sequence are indicated. The lower bars represent truncated proteins expressed in insect
cells infected by the recombinant baculoviruses indicated on the left.
Single-stranded RNA probes were obtained by/n vitro transcription
of the gene 2 cDNA cloned in pBS plasmid, downstream of the T3
promoter. Plasmid pBSRF2 was linearized and gene 2 was transcribed
using T3 polymerase as described by the manufacturer (Promega).
Rotavirus non-specific probe was obtained by in vitro transcription of
the PBS polylinker. Rotaviral mRNAs, obtained by viral transcriptase,
were also used as probes.
Double-stranded RNA probes were obtained by reannealing of
transcripts produced by in vitro transcription of gene 2 cloned in pBS
plasmid with T3 and T7 polymerases. For annealing, equimolar
amounts of both transcripts were mixed, boiled, and quickly cooled.
NaCI was added to a final concentration of 200 mM. The sample was
heated to 70 °C, slowly cooled to 30 °C and 1/2 vol of SSC 20 x (3 MNaC1; 3 N-sodium citrate) and RNase A (12-5 gg/ml) were added. The
reaction mixture was incubated at 37 °C for 20 min, then RNA was
extracted using phenol:chloroform and precipitated with ethanol.
Viral genomic RNA labelled with T4 RNA ligase and [32P]pCp also
was used as probe.
U.v.-cross-linking analysis. Interactions between rotavirus proteins
and genomic RNA were analysed according to a protocol obtained
from Dr Claude Meric (Stewart et al., 1990). Purified single-shelled
particles (10 to 15 ~tg)in water were irradiated by u.v.-light at a distance
of 4 cm from the lamp (15 W Philips G5T8) for various times. Particles
were disrupted by boiling for 10 min in 0.5% SDS. P r o t e i ~ R N A
complexes were then immunoprecipitated in 1 ml of 50 mM-Tris-HC1
buffer pH 8.5, 150 mM-NaCI, 1% Triton X-100, 1% sodium desoxycholate, 20 mM-EDTA, with 1 gl of either an anti-rotavirus serum or an
anti-VP2 monoclonal antibody (E22, Roseto et al., 1983). After a
1 h incubation at 37 °C, 40 gl of a suspension of protein A-Sepharose
beads (Pharmacia) in immunoprecipitation buffer (1:1, v/v) were
added. Incubation was continued for 30 min at room temperature, then
the beads were washed three times in immunoprecipitation buffer and
twice in 40 mM-Tris-HC1 buffer pH 7.4, 4 mM-EDTA. RNA was
digested with RNase A (1 pg/ml) for 15 min at 37 °C. The beads were
washed twice with TE (40 mM-Tris-HC1 4 mM-EDTA) and twice with
kinase buffer (50 mM-TribHC1 pH 8, 1.5 mM-Spermidine, 5 mMMgCI2, 1 mM-DTT, 5 % glycerol). RNA, protected from RNase A by
proteins, was labelled with 0.5 HI of [y~eP]ATP (148 TBq/mmol to
0-37 TBq/ml) and 10 U of T4 polynucleotide kinase (Amersham) for
15 min at 37 °C. The beads were washed once in immunoprecipitation
buffer and boiled for 10 min in dissociation buffer (Laemmli, 1970).
P r o t e i ~ R N A complexes were separated by SDS-PAGE.
When indicated, complexes were digested before electrophoresis by
various concentrations of endoproteinase V8 (Sigma) for 1 h at 37 °C
in 125mM-Tris-HCl pH7-4, 0.1% SDS, 1% glycerol, lmM-2mercaptoethanol. Electrophoresis of proteolytic peptides was performed as described by Shfigger & Von Jagow (1987).
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 18:33:00
3426
M. Labbd and others
Results
Localization of the RNA-binding domain in VP2 using
recombinant proteins
Rotavirus VP2 protein from infected MA104 cells binds
nucleic acids as seen by North-Western blotting experiments (Boyle & Holmes, 1986). Recombinant baculovirus expressed VP2 protein also binds nucleic acids in
the same assay (Labb~ et at., 199t). The ability of
recombinant protein to bind nucleic acids in vitro
provided an opportunity for determining which domain
of VP2 was essential for binding activity. To identify this
domain, truncated VP2 proteins were expressed in insect
cells. All the 12 recombinant VP2 proteins were expressed
to a high level varying from 2 to 40 mg/2 x 109 cells (data
not shown). None of these VP2 constructs (with the
exception of the full-length protein) was able to assemble
into core-like particles.
Each recombinant protein was first tested for ssRNA
binding activity as described in Methods, using a 32p_
labelled viral mRNA as the probe (Fig. 3 and Fig. 5). We
also used probes obtained by in vitro transcription of the
full-length cDNA of gene 2, which possessed the
complete non-coding sequence, but not the exact 5' and
Y termini (Fig. 4). To localize the binding domain
106K 80K
49-5K
32-5K
Fig. 3. Single-stranded R N A binding with amino terminal truncated
forms of VP2. Sf9 cells were infected by recombinant baculoviruses and
harvested 72 h p.i. Proteins were separated by electrophoresis and
stained by Coomassie blue or transferred onto PVDF membranes.
Blots were incubated with single-stranded probes (~P-labelled viral
mRNA) diluted in SBB containing 250 gg/ml tRNA, and 50 mMNaC1. Blots were dried and autoradiographed. The left panel shows a
Coomassie blue-stained gel and the right panel shows a blot. The
position of the M,. markers is indicated on the left. Recombinant
baculoviruses used to infect insect cells (4 x 105 cells/well) are indicated
on the top. 0 , Positions of VP2 and truncated proteins.
approximately, a first set of recombinant baculoviruses
was constructed in order to express different combinations of VP2 divided into three parts. Recombinant
proteins VP2, P2AT877 (amino acids 1 to 285) and
P2AT1777 (amino acids 1 to 583) bound ssRNA (Fig. 3
and Fig. 4). It should be noted that a few cellular
proteins, observed in all lanes, also bound nucleic acid
in these conditions. In cells infected by recombinant
baculovirus Bac2AT1777, a protein of approximately
27K bound the probe. This protein was found to be a
degradation product of VP2, as it was recognized by our
monoclonal antibody directed against VP2 (data not
shown). P2AN protein (amino acids 1 to 290 plus 589 to
880) also bound the ssRNA probe (data not shown). In
contrast, P2CT1821 (amino acids 290 to 880; Fig. 3),
P2CT921 (amino acids 589 to 880; data not shown) and
P2M9 (amino acids 290 to 880; data not shown) did not
show any RNA binding activity. All recombinant
proteins containing amino acids 1 to 285 bound ssRNA
probe, indicating that the VP2 RNA binding domain was
localized in the amino-terminal region of VP2. This
binding domain appeared to be independent of the rest
of the protein.
In order to refine our mapping, a second set of
recombinant baculoviruses was constructed to express
proteins that were progressively deleted from the aminoterminal region. Recombinant proteins P2M16.9,
P2M16.3, P2M14 and P2M13 having 26, 45, 99 and 127
amino acids deleted, respectively, were tested for R N A
binding. None of these recombinant truncated proteins
bound ssRNA probe, although they were expressed at
high level in insect cells (left part of Fig. 4). These results
showed that the first amino acids of VP2 are essential for
binding activity.
A recombinant baculovirus that expressed amino acids
1 to 132 of VP2 was constructed and shown to bind
ssRNA probe. Consequently, the RNA binding site was
located between amino acids 1 to 132.
Recombinant proteins that possess binding activity
interact with exact mRNA (viral mRNA, Fig. 3 and Fig.
5) as well as with ssRNA having modified ends (T3
transcripts, Fig. 4). These results clearly indicate that 3'
and 5' termini have no role in the studied activity. The
recombinant proteins P2AT877 and P2AT 1777 also were
tested for their ability to bind a ssRNA probe nonspecific for the rotavirus sequence. Like complete VP2,
these proteins bound non-rotavirus-specific probe in a
North-Western blotting assay (data not shown).
We also analysed the ability of truncated proteins to
bind double-stranded probes. Proteins VP2, P2AT1777
and P2A420 bound double-stranded probes (Fig. 6),
although, to observe binding of double-stranded
probes, tRNA competitor had to be omitted, confirming
that recombinant proteins, like viral VP2, had lower
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 18:33:00
RNA binding domain of rotavirus VP2
3427
JoJ
106K80K
Q
49.5K-
1?
32.5KFig. 4. Single-stranded R N A binding activityof amino terminal deleted
proteins. Recombinant proteins deleted from amino-terminal region
were tested for their ability to bind ssRNA probes (T3 transcripts) as
described in the legend of Fig. 3. 0 , Positions of VP2 and truncated
proteins.
Fig. 6. Localization of the dsRNA binding site. Recombinant proteins
were transferred onto PVDF membrane and tested for their ability to
bind double-stranded probes (reannealed T3-T7 transcripts as described in Methods). 0 , Position of truncated proteins.
106K -
Interactions between rotavirus proteins and the genome
inside viral particles
80K 49.5K 32.5K
27.5K -
18K -
Fig. 5. Localization o f the ssRNA binding domain on VP2 protein.
Proteins from insect cells (4 x 105) infected by bacnlovirus recombinant
BacRF2A (VP2) or Bac2A420 (amino acids 1 to 132) were separated by
electrophoresis (18% acrylamide gel) and transferred onto PVDF
membranes. Blots were incubated with probes (32P-labelled viral
mRNA) as described. 0 , Positions of VP2 and truncated proteins.
affinity for double-stranded nucleic acids than for singlestranded nucleic acids. Any deletion of the aminoterminal region abolished the binding activity of doublestranded probes as observed with single-stranded probes
(data not shown), suggesting that the double-stranded
binding site, like the single-stranded binding site, was
located in the amino-terminal region of VP2.
To investigate the interaction between VP2 and the
rotavirus genome, purified single-shelled particles were
u.v.-irradiated for various lengths of time. Complexes
were immunoprecipitated, radiolabelled and analysed as
indicated in Methods. These experiments revealed that
VP2 protein was cross-linked to RNA within 2 rain of
u.v.-irradiation (Fig. 7 lane 3). Radiolabelling increased
after 6 min of irradiation (Fig. 7 lane 5). In the noncross-linked samples (Fig. 7 lane 1), VP2 was not
labelled; this implied that the labelling was not due to
protein phosphorylation or phosphate exchange. After a
longer exposure to u.v.-light, the signal decreased
probably because of the formation of higher order
complexes (data not shown). Similar experiments were
performed using a polyclonal serum directed against
rotavirus to immunoprecipitate cross-linked particles.
This serum reacts with all rotavirus structural proteins,
but autoradiography revealed that only VP2 was
radiolabelled (results not shown), suggesting that, in
these conditions, VP2 was the only protein to interact
with the genome in purified single-shelled particles.
To verify that a single domain of VP2 interacts with
the genome, the radiolabelled protein-RNA complexes
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 18:33:00
3428
M . Labb{ and others
UV (rain)... 0
E22...
+
2
2
6
6
+
_
+
106K 80K -
49.5K -
32-5K 27.5K
1
2
3
4
5
Fig. 7. U.v.-cross-linking analysis. Single-shelled particles were subjected to u.v.-cross-linking and disrupted by boiling in the presence of
SDS. The duration of u.v.-irradiation is indicated above the figure.
Proteins were immunoprecipitated (+) or not ( - ) with E22 monoclonal antibody. After RNase digestionand labellingwith T4 PNK and
[ya2P]ATP,samples were boiled and fractionated by SDS-PAGE. Gels
were dried and autoradiographed.
Protease V8 (U)...
0-01 0.1 1
5
0
17K
14.4K
107K
8.2K
6-2K
1
2
3
4
5
Fig. 8. Proteolytic cleavage of V P 2 - R N A complexes. V P ~ R N A
complexes obtained as described in the legend of Fig. 6 were digested
by various concentrations of endoproteinase V8. Proteolytic products
were analysed by electrophoresis in a Tris-tricine SDS gel (Sh/igger &
Van Jagow, 1987). O, Position of uncleaved VP2. Mr markers are
indicated on the left.
were digested by the endoproteinase V8 after immunoprecipitation, RNase digestion and labelling. Proteolytic
peptides were separated by electrophoresis in a gel that
allowed resolution of low M r polypeptides (Sh~igger &
Von Jagow, 1987). A high concentration of endoproteinase V8 (Fig. 8 lane 4) generated a unique labelled
peptide of about 7-5K suggesting that the nucleic acid
binding site is localized on a unique sequence of VP2.
Discussion
This report describes interactions between the structural
rotavirus protein, VP2, and nucleic acids. The nucleic
acid binding sites are localized in the N-terminal portion
of VP2 as summarized in Fig. 2. In viral particles, VP2
interacts with genome RNA.
The ability o f the recombinant protein, expressed in
insect cells, to bind nucleic acids, allowed us to localize
the binding domain using truncated recombinant proteins. Recombinant truncated protein consisting of
amino acids 1 to 132 binds both ssRNA and doublestranded nucleic acid. The data presented here show that
the nucleic acid binding domains consist of sequences
located between amino acids 1 to 132 indicating that
these binding sites are not composed of several domains
arranged in a high order structure. The fact that deletions
outside the amino-terminal region did not modify the
binding activity, indicates that this activity of the
terminal region is not dependent on other sequences,
although the conformation of these truncated proteins
may be different. In contrast, deletion of the first amino
acids (1 to 26) abolished the R N A binding activity of the
recombinant protein, suggesting that the amino end of
VP2 either plays an important role in the binding site(s)
themselves or in the conformation of the peptide 1 to 132
that contains the binding sites.
We have previously shown that recombinant VP2
assembles into core-like particles in the absence of other
viral components. When protease inhibitors are not
added before lysis of insect cells, core-like particles are
made of a VP2 cleavage product (Labb6 et al., 199 l) that
is also identified in empty viral particles (Brfissow et al.,
1990). Terminal sequencing of the cleavage product
showed that it lacks amino acids 1 to 92 from these corelike particles (Zeng et al., 1994). These results, and data
presented here, indicate that the site of pseudocore
assembly and the R N A binding sites are located in
different domains of the proteins. Location of the R N A
binding sites at the amino terminus of VP2 could be
related to the proteolytic cleavage of VP2 (mentioned
above) in empty particles. Absence of genomic R N A
could make the cleavage site Gln9~$Lysga accessible to
the proteases either by a conformational change of VP2
or because the R N A molecule is in close contact with
Lysga in particles containing nucleic acid.
Amino acid sequence analysis of VP2 from various
rotavirus strains revealed several motifs that could
mediate VP2 binding to the R N A binding activity. All
these motifs are present in most of the sequenced
rotavirus gene 2, but none of them is strictly conserved in
all strains. Our data have clearly shown that the two
leucine zippers (amino acids 536 to 686) that have been
suspected to be involved in R N A binding are not
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 18:33:00
R N A binding domain of rotavirus VP2
essential for this activity. However, these motifs could
take part in the oligomerization process of the protein.
Three other motifs could play a role in interaction
between VP2 and nucleic acids: (i) a predicted helix-turnhelix structure (amino acids 65 to 121) (Garnier et al.,
1978); (ii) heptad repeats of lysine and glutamic acid
residues (amino acids 53 to 81) (Mitchell & Both, 1990);
(iii) a motif found in several proteins known to bind ssand dsRNA (amino acids 64 to 76) (McCormack et at.,
1992). It should be noted that all these motifs are
included in the RNA binding peptide that we have
mapped (amino acids 1 to 132). Among the four group A
rotavirus strains for which the VP2 gene has been
sequenced so far, there are insertions and deletions in the
region of amino acids 30 to 32, despite an overall high
similarity in the region amino acids 1 to 132. This
situation is fairly unusual when comparing group A
rotaviral structural genes and suggests that these insertions and deletions are located in a region of little
importance to the RNA binding site. When comparing
rotavirus VP2 with the counterparts reovirus lambda 1
and bluetongue virus VP3, no conserved motifs could be
identified, except that these proteins share highly charged
amino-terminal ends. More precise site-directed mutations are needed to confirm that one of these motifs is
effective in the RNA binding activity of VP2. Localization of the RNA binding site in the N-terminal
sequence of the viral capsid proteins is commonly found
in ssRNA viruses. This portion of the capsid protein is
generally basic and helical. Basic amino acids are located
on one face of the helix, that interact with RNA in a
sequence independent manner (Fox et al., 1994).
VP2 is the major component of rotavirus core, and it
forms the inner third layer surrounding the 11 segments
of genomic dsRNA. Biophysical data clearly indicate
that genomic RNA should be bent to be packaged in
particles (Kapahnke et al., 1986). In North-Western
blotting assays, viral VP2 (Boyle & Holmes, 1986) as well
as recombinant VP2 or its truncated forms showed a
higher affinity for ssRNA than for double-stranded
nucleic acids. This result suggests that VP2 might play an
important role in the encapsidation of the ssRNA that is
used as template for dsRNA synthesis in core RI. It is
noteworthy that both ssRNA and dsRNA binding
activities are located in the same N-terminal region of the
protein suggesting that encapsidation of ssRNA and of
genomic RNA could be linked.
Interactions of VP2 with the genome were investigated
in purified viral single-shelled particles by u.v.-crosslinking. These experiments showed that VP2 can be
covalently linked to dsRNA, thus demonstrating that
this protein is in close contact with the genomic RNA.
These results also suggest that VP2 may play a role in
bending the genomic RNA. In the conditions we used in
3429
this work, interactions between VP6, VP1 or VP3 and the
viral RNA were not detected (Fig. 7 and data not
shown). These two latter proteins could be suspected to
interact with the genome since it has been demonstrated
that VPI and VP3 possess polymerase activity and
guanyl transferase activity, respectively (Valenzuela et
al., 1991 ; Pizarro et al., 1991 ; Liu et al., 1992). The lack
of labelling after cross-linking could be due to either the
very low concentration of these proteins in particles or
the low reactivity of the antibodies used in the assay, It
could also be hypothesized that interactions between
VP1, VP3 and the genomic RNA occur only transiently
during RNA polymerization.
The location of the RNA binding site in the core is still
unknown. It seems reasonable to hypothesize that this
site is located inside the core, since it interacts with the
genome. However, this site may be accessible from the
outside of the particles as indicated by the following
observations: (i) The RNA binding site is close to the
protease susceptible site of VP2 that is accessible from
the outside, as discussed above (ii) Rotavirus-like
particles, obtained by co-expression in insect cells of VP2
and VP6 are able to bind poly(U) (M. Labbd and J.
Cohen, unpublished). A similar observation was made
with bluetongue virus core-like particles made of VP3
and VP7 (Loudon & Roy, 1992). These data suggest that
the RNA binding domain of VP2 is located close to the
12 type I holes in the capsid (Prasad et al., 1988) that are
the only contact between VP2 and the external medium.
M. Labb~ was funded in part by the 'Fondation Dufrenoy'. We are
grateful to Mary K. Estes for careful reading of the manuscript.
References
BOYLE, J. F. & HOL~ES, K. V. (1986). RNA-binding proteins of bovine
rotavirus. Journal of Virology 58, 561 568.
BRI~MONT, M., JUSTE-LESAGE, P., CHABANNE-MAUTHEROT, D., CHARPILIENNE, A. & COHEN, J. (1992). Sequences of the four larger
proteins of a porcine group C rotavirus and comparison with the
equivalent group A rotavirus proteins. Virology 168, 684~692.
BefJssow, H., BRUTTIN, A. & MARC-MARTIN, S. (1990). Polypeptide
composition of the rotavirus empty capsids and their possible use as
a subunit vaccine. Journal of Virology 64, 3635-3642.
COHI~N, J. (1977). Ribonucleic acid polymerase associated with purified
calf rotavirus. Journal of General Virology 36, 395-402.
ERNST, H. & DUHL, J.A. (1989). Nucleotide sequence of genomic
segment 2 of human rotavirus Wa. Nucleic Acids Research 17, 4382.
ESTES, M . K . & COHEN, J. (1989). Rotavirus gene structure and
function. Microbiological Reviews 53, 410449.
Fox, J.M., JOHNSON, J.E. & YOUNG, M.J. 0994). RNA/protein
interactions in icosahedral virus assembly. Seminars in Virology 5,
51-60.
GARNIER, J., OSGUTHORPE, D. J. & ROBSON, B. (1978). Analysis of the
accuracy and implications of the simple methods for predicting
secondary structure of globular proteins. Journal of Molecular
Biology 120, 97 120.
KAPAHNKE, R., RAPVOLO,W., DESSELBERGER,U. & RIESNER, D. (1986).
The stiffness of dsRNA: hydrodynamic studies on fluorescencelabelled RNA segments of bovine rotavigus. Nucleic Acids Research
14, 3215-3228.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 18:33:00
3430
M . Labb~ and others
KIXTS, P. A., AYR~S, M. D. & POSSEE, R. D. (1990). Linearization of
baculovirus DNA enhances the recovery of recombinant virus
expression vectors. Nucleic Acids Research 18, 5667-5672.
KUMAR, A., CHARPILIENNE,A. & COHEN, J. (1989). Nucleotide sequence
of the gene encoding for the RNA binding protein (VP2) of RF
bovine rotavirus. Nucleic Acids Research 17, 2126.
KUNKEL, T.A. (1985). Rapid and efficient site-specific mutagenesis
without phenotypic selection. Proceedings" of the National Academy
of Sciences, U.S.A. 82, 488.
LABBI3, M., CHARPILIENNE, A., CRAWFORD, S.E., ESTES, M . K . &
COHEN, J. (1991). Expression of rotavirus VP2 produces empty
corelike particles. Journal of Virology 65, 29462952.
LAEMMLI, U.K. (1970). Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature, London 227,
68(~685.
L'HARIDON, R. & SCHERRER, R. (1976). In vitro culture of a rotavirus
associated with neonatal calf scours. Annales de Recherche Vdt~rinaires 7, 373 381.
LINDSAY, A. D., VONDERFECHT, S. L. & EIDEN, J. J. (1994). Group B
rotavirus VP2: sequence analysis, expression and gene coding
assignment. Virology 199, 141-150.
LIU, M., MATTION, M.N. & ESTES, M.K. (1992). Rotavirus VP3
expressed in insect cells possesses guanylyltransferase activity.
Virology 188, 77-84.
LOUt)ON, P. T. & RoY, P. (1992). Interaction of nucleic acids with corelike and subcore-like particles of bluetongue virus. Virology 191,
231-236.
McCORMACK, S.J., THOMAS, D.C. 8,~ SAMUEL, C. E. (1992). Mechanism of interferon action: identification of a RNA binding domain
within the N-terminal region of the human RNA-dependent P 1/elF2a protein kinase. Virology 188, 47-56.
MITCHELL, D. B. & BOTH, G. W. (1990). Completion of the genomic
sequence of the simian rotavirus S A l l : nucleotide sequences of
segments 1, 2 and 3. Virology 177, 324-331.
PIZARRO, J. L., SAND1NO, A. M., PIZARRO, J. M., FERNANDEZ, J. 8z
SPENCER, E. (1991). Characterization of rotavirus guanylyltransferase activity associated with polypeptide VP3. Journal of
General Virology 72, 325-332.
PRASAD, B. V. V., WANG, G. J., CLERX, J. P. M. & CHIL~,W. (1988).
Three-dimensional structure of rotavirus. Journal of Molecular
Biology 199, 269-275.
ROSETO, A., SCHERRER, R., COHEN, J., GUILLEMIN, M.C., CHARPILIENNE, A., FEYNEROL, C. & PERILS, J. (1983). Isolation and
characterization of anti-rotavirus immunoblobulins secreted by
cloned hybridoma cells lines. Journal of General Virology 64,
237-240.
SCH.~GGER, H. & VON JAGOW, G. (1987). Tricine sodium dodecyl
sulfate-polyacrylamide gel electrophoresis for the separation of
proteins in the range from 1 to 100 kDa. Analytical Biochemistry 166,
368-379.
STEWART, L., SCHATZ, G. & VOCT, V. M. (1990). Properties of avian
retrovirus particles defective in viral protease. Journal of Virology 64,
5076-5092.
SUMMERS, M. D. & SMITH, G. E. (1987). A manual of methods for
baculovirus vectors and insect cell culture procedures. Texas Agricultural Experiment Station Bulletin 1555.
TIAN, Y., TARLOW, O. & McCRAE, M. A. (1990). Nucleotide sequence
of gene 2 of the UK tissue culture adapted strain of bovine rotavirus.
Nucleic Acids Research 18, 4015.
VALENZUELA,S., PIZARRO, J., SANDINO,A. M., VASQUEZ,M., FERNANDEZ, J., HERNANDEZ, O., PATTON, J. & SPENCER, E. (1991).
Photoaffinity labeling of rotavirus VP1 with 8-azido-ATP:
identification of the viral RNA polymerase. Journal of Virology 65,
3964-3967.
ZENG, C. Q.-Y., LABBI~,M., COHEN, C., PRASAD, B. V. V., CHEN, D.,
RAMIG, R. F. & ESTES, M. K. (1994). Characterization of rotavirus
VP2 particles. Virology 201, 55-65.
(Received 25 March 1994; Accepted 29 July 1994)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 18:33:00