Download The complete nucleotide sequence of apple mosaic virus (ApMV

Journal
of General Virology (2000), 81, 273–278. Printed in Great Britain
..........................................................................................................................................................................................................
SHORT COMMUNICATION
The complete nucleotide sequence of apple mosaic virus
(ApMV) RNA 1 and RNA 2 : ApMV is more closely related to
alfalfa mosaic virus than to other ilarviruses
P. J. Shiel and P. H. Berger
Plant Pathology Division/Department of Plant, Soil and Entomological Sciences, University of Idaho, Moscow, ID 83844-2339, USA
The complete nucleotide sequences of apple mosaic
virus RNA 1 and 2 have been characterized. Apple
mosaic virus RNA 1 is 3476 nucleotides in length
and encodes a single large open reading frame
(ORF), whereas apple mosaic virus RNA 2 is 2979
nucleotides in length and also encodes a single
ORF. The amino acid sequences encoded by RNA 1
and 2 show similarity to all of the other ilarviruses
for which sequence data are available, but both are
more closely related to alfalfa mosaic virus (AMV)
than to other ilarviruses. Points of similarity include
the absence of ORF 2b, present on the RNA 2 of all
previously characterized ilarviruses. The close relationship to AMV also occurs in the movement
protein, encoded by RNA 3, but not with the coat
protein. These data suggest that the present taxonomy should be revised, and that AMV should be
considered an aphid-transmissible ilarvirus.
Apple mosaic virus (ApMV) is in subgroup III of the genus
Ilarvirus, which includes Prunus necrotic ringspot virus
(PNRSV) and rose mosaic viruses (Rybicki, 1995). The genus
Ilarvirus comprises a large group of plant viruses with primarily
woody hosts and is divided into subgroups based upon their
serological characteristics. This genus, along with the genera
Alfamovirus, Cucumovirus, Bromovirus and Oleavirus, constitute
the family Bromoviridae. Genera are distinguished by the
biological properties of the viruses. The Bromoviridae have
positive-sense single-stranded RNA genomes divided into
three components designated RNA 1, 2 and 3, and contain an
RNA 4 which is a subgenomic messenger for the coat protein
Author for correspondence : Philip Berger.
Fax j1 208 885 7760. e-mail pberger!uidaho.edu
The nucleotide sequence data reported in this article have been
deposited with EMBL/GenBank under accession numbers AF174584
and AF174585.
0001-6433 # 2000 SGM
of the virus (Rybicki, 1995). Ilarviruses and alfalfa mosaic virus
(AMV) (the sole member of the genus Alfamovirus) share the
biological trait of genome activation, where mixtures of RNAs
1, 2 and 3 are non-infectious unless either subgenomic RNA 4
or coat protein is added. This function can be fulfilled by RNA
4 or coat protein of heterologous as well as homologous
viruses of these genera, but not from other genera (Gonsalves
& Fulton, 1977). Biological and chemical RNA protection
assays show that both the AMV and ilarvirus 3h non-translated
region (NTR) specifically bind to coat protein (AnselMcKinney & Gehrke, 1998). Additional experiments with
peptides from the coat protein of tobacco streak ilarvirus (TSV)
and the TSV 3h NTR showed that specific binding occurs
between these two viruses despite only limited sequence
homology (Swanson et al., 1998).
AMV is separated from the ilarviruses primarily by its
mode of transmission : it is transmitted noncirculatively by at
least 14 species of aphids (Jaspers & Bos, 1980). TSV is
transmitted in a pollen-mediated manner by thrips (Kaiser et al.,
1982). Experimental evidence suggests that thrips are also
involved in the transmission of PNRSV (Greber et al., 1992).
The mechanism of transmission of many of the ilarviruses,
including ApMV, is still unknown.
We previously reported the complete nucleotide sequences
of ApMV RNA 3 and RNA 4. RNA 3 is bicistronic and
encodes the putative movement protein (Shiel et al., 1995). The
movement protein is directly translated from RNA 3, whereas
the coat protein is translated from the subgenomic mRNA,
RNA 4 (Alrefai et al., 1994). This is in agreement with the
genome organization of other members of the Bromoviridae,
where considerable sequence information is available. If the
remainder of the ApMV genome shares the genomic
organization of the Bromoviridae, then RNAs 1 and 2 should
encode the virus replicases (Rybicki, 1995). Recently, a second
open reading frame (ORF) on RNA 2 was identified on all of
the ilarviruses for which sequence data are available (Xin et al.,
1998). This 2b ORF is expressed in a manner similar to the 2b
ORF present in cucumoviruses, even though it encodes a larger
peptide and has no sequence similarity to the 2b ORF of
cucumoviruses. It is interesting that there is no corresponding
2b ORF in AMV.
CHD
P. J. Shiel and P. H. Berger
While complete nucleotide sequences of AMV and several
bromoviruses and cucumoviruses have been available for some
time (Brunt et al., 1996), it was only recently that any complete
ilarvirus sequences were published (Scott & Ge, 1995 ; Ge et al.,
1997 ; Scott et al., 1998). This is probably due in part to the
difficulty in handling and manipulating ilarviruses, relative to
many other viruses. In addition, most of the known ilarviruses
primarily infect woody perennial hosts, where often they are
present only in low titre. In general, they are difficult to isolate
from their woody hosts and are poorly transmitted to such
hosts by mechanical means.
ApMV is a pathogen with a diverse natural host-range
consisting primarily of woody plants. Besides apple, ApMV
naturally infects rose, hazelnut, filbert, horse chestnut, raspberry, birch and hops (Rybicki, 1995). However, due to the
labile nature of the virus, isolation has been difficult, and the
host-range of the virus has usually been determined using
grafts of infected plants (Sweet & Barbara, 1979). Our objective
was to characterize the entire genome of ApMV to facilitate
detection and further study of this pathogen of woody plants.
ApMV was isolated from infected symptomatic apple trees
(Malus domestica L.) var. ‘ Golden Delicious ’. The virus was
propagated in and purified from inoculated cotyledons of
cucumber (Cucumis sativus L. var. Lemon) as previously
described (Alrefai et al., 1994). The virus was further purified by
10–40 % sucrose density-gradient centrifugation (SDGC),
which separated the virus into an upper band and two closely
spaced lower bands. The lower bands (which contain RNA 1
and 2) were separated from the upper band (which contains
RNA 3) and concentrated for RNA purification.
RNA was extracted from the bottom component of the
SDGC-purified virions by adding virus to dissociation buffer
containing 1 % SDS and then extracting RNA with phenol
heated to 60 mC. Purified ApMV RNA was polyadenylated by
using poly(A) polymerase (Gibco\BRL) and used as a template
to synthesize the first-strand cDNA with oligo d(T) and
Moloney murine leukaemia reverse transcriptase or Superscript
II (Gibco\BRL). The first-strand reaction products were used
for synthesis of second-strand cDNA, according to the
manufacturer’s protocols (Promega). Double-stranded cDNA
was blunt-end ligated into the EcoRV site of pBluescript, and
these reactions were used to transform E. coli strain DH10b.
Plasmids containing sizeable inserts were analysed by restriction enzyme mapping and were further purified for doublestranded DNA sequencing. Each plasmid was treated with
10 mg\ml RNase A (Sigma) at 37 mC for 30 min, extracted
with phenol–chloroform and precipitated with ethanol.
The DNA was sequenced in both directions by dideoxychain termination method using the Sequitherm Long-Read
Cycle Sequencing Kit (Epicentre Technologies) according to
the manufacturer’s protocols and analysed using a Li-Cor
4000L automated DNA sequencing system and IRD-41
labelled primers. The 5h-ends of RNA 1 and 2 were cloned by
5h-rapid amplification of cDNA ends (5h-RACE) (Gibco\BRL
CHE
version 2.0) according to the manufacturer’s protocols and
sequenced as above.
Nucleic acid and amino acid sequence data were analysed
with the University of Wisconsin Genetics Computer Group
sequence analysis software package (GCG version 8) at the
Center for Visualization, Analysis, and Design of Molecular
Sequences at Washington State University. Additional analyses were performed with DNA Strider (version 1.3, CEA,
France). The RNA secondary structure of the 5h-noncoding
regions was analysed with the RNAFOLD program (RNA
Secondary Structure Predictor, version 2.0 ; Scientific and
Educational Software). Phylogenetic analysis was done with
the program PAUP (Phylogenetic Analysis Using Parsimony,
version 4.0).
cDNA clones that contained ApMV RNA 1 and 2 were
initially identified on the basis of sequences that were
homologous to the 3h-NTR of RNA 3 but not to the rest of
RNA 3, because the 3h-NTRs of the three RNAs of ApMV (and
all other bromoviruses) are the only region which has sequence
homology within the genome of each species. In addition to
RNA 3, two other sets of clones were found. Using the BLAST
program, one of these sets of clones was found to encode
peptides similar to the putative polymerase regions of AMV
and Citrus leaf rugose virus (CiLRV), which are encoded by
RNA 2 of these viruses. The other set of clones encoded
peptides similar to the putative helicase region of these viruses,
which is encoded by RNA 1. In addition, both sets of clones
showed a lesser degree of similarity to the putative polymerase
and helicase regions of other members of the Bromoviridae, as
well as several tobamoviruses and tobraviruses. Additional
clones were then identified and characterized based on
homologies to those already obtained and a library of clones
was generated for each set. Sixteen nucleotides of the ApMV
RNA 1 5h-end and 499 nucleotides of the ApMV RNA 2 5hend were cloned and sequenced by 5h RACE with two
sequence-specific primers for each RNA.
ApMV has three genomic RNAs and a subgenomic RNA
which contains the coat protein cistron. The largest RNA,
RNA 1, is 3476 nucleotides long and encodes a single large
polypeptide which is similar to the methyltransferase-like and
helicase-like domains present in many plant RNA viruses. This
RNA encodes a predicted open reading frame of 1046 amino
acids (aa) with a relative molecular mass (Mr) of 118 416. These
are comparable to the protein products of the RNA 1 of AMV,
CiLRV and elm mottle virus (EMoV). These sequences have
homology to molecules that have been demonstrated to act as
methyltransferases and helicases.
The nucleotide sequence of ApMV RNA 2 was also
determined. This sequence is 2979 nucleotides in length and
contains a single large ORF. The predicted amino acid sequence
of this ORF shows similarity to the corresponding protein
product of AMV and other ilarvirus sequences that are
currently available. It encodes a predicted peptide product that
is 875 aa in length with an Mr of 99 866. This is comparable to
(a)
(b)
(c)
CHF
ApMV RNA 1 and 2 sequences
Fig. 1. Phylogenetic trees inferred by parsimony analysis of (a) the putative methyltransferase/helicase protein, (b) the polymerase protein and (c)
the movement protein of members of the Bromoviridae. Horizontal branch lengths are proportional to evolutionary distance. Analysis was done under
exhaustive search conditions. The number on each branch is the result of bootstrap analysis. TMV is used as the outgroup in trees (a) and (b) ; tree
(c) is unrooted. Serogroups denoted with an asterisk (*) contain definable zinc-finger domains. See text and Table 1 for definitions of virus
acronyms.
P. J. Shiel and P. H. Berger
Table 1. GenBank accession numbers of sequences used in this study
(a) Methyltransferase/helicase motifs
Accession no.
Accession no.
Virus
Helicase
Polymerase
Virus
Helicase
Polymerase
AMV
BMV
CiLRV
CMV
EMoV
L00163
X58456
U23715
D16403
U57047
K02702
X58457
U17726
D10209
U34050
PDV
RBDV
SpLV
TMV
TSV
U57648
S51557
U93192
X02144
U80934
*
S51557
U93193
X02144
U75538
* Sequence not published.
(b) Putative movement protein, coat protein, and 3h-NTR of RNA 3
Virus acronyms not otherwise defined in the text : AV, Asparagus virus ; CVV, Citrus variegation
virus ; HdMV, Hydrangea mosaic virus ; LRMV, lilac ring mottle virus.
Virus
AMV-4M
AMV-S
ApMV-A
ApMV-G
AV-2
CiLRV
CVV
EMoV-1
EMoV-2
HdMV
LRMV
PDV-Sc
PDV-Sk
PNRSV-CH9
PNRSV-CH30
Accession no.
K02703
X00819
U15608
S78319
X86352
U17390
U17389
U57048
U85399
U35145
U17391
L28145
U31310
AF034992
AF034994
Virus
PNRSV-CH38
PNRSV-CH39
PNRSV-CH57
PNRSV-CH61
PNRSV-CH71
PNRSV-G
PNRSV-Mi
PNRSV-P5
PNRSV-Pr
PNRSV-Sc
PNRSV-Sp1
PNRSV-Sp2
PNRSV-SW6
SpLV
TSV
the protein products of the RNA 2 of AMV, CiLRV and
EMoV.
The RNA 2 ORFs of ApMV, like AMV, CiLRV,
EMoV, raspberry bushy dwarf virus (RBDV), spinach
latent virus (SpLV) and TSV, share sequence homology,
especially in the region of the consensus signature
motif
DX [FYWLCA]X – DXn[STM]GX TX [NE]XnGDD
$ $
$
!"
found in most virus-encoded RNA-dependent RNA polymerases (Buck, 1997). In addition, all of these proteins appear
to have a predominance of acidic residues near the aminoterminal end, and a similar predominance of basic residues near
the carboxyl terminus. Unlike all of the ilarviruses that have
been sequenced to date, ApMV does not contain a second
ORF (ORF 2b), and is similar to AMV in that respect.
When GenBank was searched with the BLAST and
TFASTA programs for sequences similar to ApMV RNA 1 and
CHG
Accession no.
AF034991
AF034990
AF034993
AF034989
AF034995
S78312
AF013285
L38823
AF013286
U57046
U03857
Y07568
AF013287
U93194
X00435
2, results indicated strong similarities between these sequences
and proteins of AMV, brome mosaic virus (BMV), CiLRV,
cucumber mosaic virus (CMV), EMoV, prune dwarf virus
(PDV), RBDV, SpLV, tobacco mosaic virus (TMV) and TSV.
ApMV appeared to be more closely related to AMV and PDV
than to other viruses. Phylogenetic inferences were obtained
using both parsimony and the distance-based neighbourjoining methods. Since dendrograms generated by either
method were topologically indistinguishable, only results
using parsimony are presented here (Fig. 1). These dendrograms show that both sequences are more closely related to
AMV than to the ilarviruses from other subgroups.
The putative movement protein on RNA 3 shows the
highest degree of homology to the other ilarviruses that infect
rosaceous hosts, particularly PNRSV (Fig. 1). Next in similarity
is the movement protein of AMV, which is clearly between the
ApMV RNA 1 and 2 sequences
Fig. 2. Alignment and secondary structural features of the 5h NTR of ApMV RNA 1 and RNA 2. Initiation codons for the large
ORFs of both RNAs are boxed. Bases involved in secondary structure are underlined.
subgroup III ilarviruses and other ilarvirus subgroups. This is in
contrast to the ApMV coat protein, which shows similarity
only to the other ilarviruses, as attempts to produce a multiple
sequence alignment of the ilarviruses with AMV or the other
viruses were unsuccessful (data not shown).
The 5h-NTR of ApMV RNA 1 is 78 nucleotides long and
the 5h-NTR of ApMV RNA 2 is 79 nucleotides long. The
RNAs show extensive homology (77 %), with the first 31
nucleotides identical to each other. These nucleotides are
involved in base-pairing for predicted stem–loop structures in
the 5h-NTR (Fig. 2). In addition, both RNAs have a second
predicted stem–loop structure immediately before their respective translation initiation codons. These two stem–loops
have similar sequences but have subtle and perhaps significant
differences. The stem–loop of RNA 2 includes two of the four
bases immediately upstream of the initiation codon, whereas in
RNA 1 the stem structure ends six nucleotides away from the
initiation codon. The ‘ context ’ nucleotides immediately surrounding the initiation codon have profound effects on the
ability of messenger RNAs to translate protein products, and
this secondary structure present on the 5h-NTRs of RNA 1 and
2 may affect expression of their proteins (Taylor et al., 1987).
Both loop sequences can potentially interact with other
nucleotides in the 5h-NTR. The GAAAAG loop sequence in
the RNA 2 stem–loop is complementary to CUUUUC and
$)
%$
could form a pseudoknot structure.
There is no detectable similarity between the 5h-NTR of
RNA 1 and 2 and the 5h-NTR of RNA 3. In addition, there does
not appear to be any consensus internal control region (ICR)
on the 5h-NTR of RNA 1 or 2, as was previously observed on
RNA 3 of ApMV and some of the other Bromoviridae (Shiel et
al., 1995). This also appears to be the case for RNA 1 and 2 of
AMV and the known ilarvirus sequences. However, there is an
eight base palindrome sequence (GAUUAAUC) that occurs at
nucleotide 36 of both RNAs of CiLRV and EMoV. ApMV
RNA 1 and 2 also has an eight nucleotide palindromic
sequence in an A–U-rich area immediately after the 5h
stem–loop at base 22 (AAUUAAUU) (Fig. 2). This is in
contrast to the 5h-NTR sequence of RNA 3, which is a 13
nucleotide palindrome, is G–C-rich, and has homology to the
consensus ICR 2 sequence (Shiel et al., 1995).
Since ApMV is more closely related to AMV than to the
other ilarviruses in all of its gene products except for its coat
protein, a strictly linear descent of these viruses from a single
progenitor seems unlikely. This conclusion is most clearly
supported by comparing the viruses that have a known zincfinger domain in their coat protein, and those that either don’t
bind zinc or use a different motif for this purpose. TSV and the
zinc-finger-containing coat proteins are most distantly related
to ApMV, whereas the ilarviruses without a zinc-finger motif
are intermediate. Phylogenetic analysis of movement proteins
results in groupings according to the natural host-range of the
CHH
P. J. Shiel and P. H. Berger
virus. Furthermore, the methyltransferase\helicase as well as
the polymerase domains of ApMV (zinc-finger domain present)
appear to be more closely related to AMV and to PDV than to
any of the other ilarviruses (Fig. 1). These data further support
the idea that functional domains of the virus were obtained
from disparate host genes or other co-infecting ilarviruses
rather than by linear descent of these viruses from a single
progenitor. A polyphyletic origin of the ilarviruses could also
have been the result of recombination events. It is possible that
AMV originated from ancestors in common with ApMV,
PNRSV and PDV, and acquired its coat protein gene from
another virus (or perhaps a plant gene), enabling it to be
transmitted by aphids, but retained its requirement for coat
protein in genome activation.
In conclusion, phylogenetic analysis of available sequences
of the protein products encoded by ApMV RNA 1 and 2 and
the putative movement protein encoded by RNA 3 indicate
that ApMV is more closely related to AMV than to the other
ilarviruses. These relationships with other related viruses differ
from those observed with coat proteins of ApMV, which is
more closely related to the ilarviruses than to AMV. It has
been noted by other researchers that perhaps AMV should be
included as a member of the ilarviruses, rather than as a distinct
genus (Sanchez-Navarro & Pallas, 1997 ; Scott et al., 1998).
Phylogenetic analysis of the ApMV genome supports this
relationship by placing AMV between ApMV and other
ilarviruses.
Ge, X., Scott, S. W. & Zimmerman, M. T. (1997). The complete
sequence of the genomic RNAs of spinach latent virus. Archives of
Virology 142, 1213–1226.
Gonsalves, D. & Fulton, R. W. (1977). Activation of Prunus necrotic
ringspot virus and rose mosaic virus by RNA 4 components of some
ilarviruses. Virology 81, 398–407.
Greber, R. S., Teakle, D. S. & Mink, G. I. (1992). Thrips-facilitated
transmission of prune dwarf and Prunus necrotic ringspot viruses from
cherry pollen to cucumber. Plant Disease 76, 1039–1041.
Jaspers, E. M. J. & Bos, L. (1980). Alfalfa mosaic virus. CMI\AAB
Descriptions of Plant Viruses, no. 229.
Kaiser, W. J., Wyatt, S. D. & Pesho, G. R. (1982). Natural hosts and
vectors of tobacco streak virus in eastern Washington. Phytopathology 72,
1508–1512.
Rybicki, E. P. (1995). The Bromoviridae. In Virus Taxonomy. Sixth Report
of the International Committee on Taxonomy of Viruses, pp. 450–457. Edited
by F. A. Murphy, C. M. Fauquet, D. H. L. Bishop, S. A. Ghabrial, A. W.
Jarvis, G. P. Martelli, M. A. Mayo & M. D. Summers. Vienna & New
York : Springer-Verlag.
Sanchez-Navarro, J. A. & Pallas, V. (1997). Evolutionary relationships
in the ilarviruses : nucleotide sequence of Prunus necrotic ringspot virus
RNA 3. Archives of Virology 142, 749–763.
Scott, S. W. & Ge, X. (1995). The nucleotide sequence of citrus leaf
rugose virus RNA 1. Journal of General Virology 76, 3233–3238.
Scott, S. W., Zimmerman, M. T. & Ge, X. (1998). The sequence of RNA
1 and RNA 2 of tobacco streak virus : additional evidence for the
inclusion of alfalfa mosaic virus in the genus Ilarvirus. Archives of Virology
143, 1187–1198.
References
Swanson, M. M., Ansel-McKinney, P., Houser-Scott, F., Yusibov, V.,
Loesch-Fries, L. S. & Gehrke, L. (1998). Viral coat protein with limited
Alrefai, R. H., Shiel, P. J., Domier, L. L., D’Arcy, C. J., Berger, P. H. &
Korban, S. S. (1994). Nucleotide sequence of apple mosaic virus coat
sequence homology bind similar domains of alfalfa mosaic virus and
tobacco streak virus RNAs. Journal of Virology 72, 3227–3234.
Sweet, J. B. & Barbara, D. J. (1979). A yellow mosaic disease of horse
chestnut (Aesculus spp.) caused by apple mosaic virus. Annals of Applied
Biology 92, 335–341.
Shiel, P. J., Alrefai, R. H., Domier, L. L., Korban, S. S. & Berger, P. H.
(1995). The complete nucleotide sequence of apple mosaic virus RNA-
3. Archives of Virology 140, 1247–1256.
protein gene : no homology with other Bromoviridae coat protein genes.
Journal of General Virology 75, 2847–2850.
Ansel-McKinney, P. & Gehrke, L. (1998). RNA determinants of a
specific RNA-coat protein interaction in alfalfa mosaic virus : conservation
of homologous features in ilarvirus RNAs. Journal of Molecular Biology
278, 767–785.
Brunt, A. A., Crabtree, K., Dallwitz, M. J., Gibbs, A. J. & Watson, L.
(1996). Viruses of Plants : Descriptions and Lists from the VIDE Database.
Cambridge : Cambridge University Press.
Buck, K. W. (1997). Comparison of the replication of positive-stranded
RNA viruses of plants and animals. Advances in Virus Research 47,
159–251.
CHI
Taylor, J . L., Jones, J. D. G., Sandler, S., Mueller, G. M., Bedbrook, J.
& Dunsmuir, P. (1987). Optimizing the expression of chimeric genes in
plant cells. Molecular and General Genetics 210, 572–577.
Xin, H. W., Ji, L. H., Scott, S. W., Symons, R. H. & Wei-Ding, S. (1998).
Ilarviruses encode a cucumovirus-like 2b gene that is absent in other
genera within the Bromoviridae. Journal of Virology 72, 6956–6959.
Received 5 May 1999 ; Accepted 24 September 1999