Arch Virol (1998) 143: 1145–1153

Sequence diversity in the coat protein coding region of twelve

sugarcane mosaic potyvirus isolates from Australia,
USA and South Africa

J. A. Handley1 , G. R. Smith2 , J. L. Dale1 , and R. M. Harding1

1
Centre for Molecular Biotechnology, School of Life Sciences, Queensland University
of Technology, Brisbane, Queensland, Australia
2
David North Plant Research Centre, Bureau of Sugar Experiment Stations,
Indooroopilly, Queensland, Australia

Accepted January 16, 1998

Summary. We have sequenced the coat protein (CP) coding region of 11 field
isolates of SCMV from Australia, USA and South Africa. The differences between
the nucleotide sequences of the isolates was 0.2 to 4.1% and the encoded amino
acid sequences differed by 0.0 to 3.5%. Phylogenetic analysis of the CP coding
sequences of the SCMV isolates and the related potyviruses SCMV-MDB, JGMV,
SrMV, MDMV-A and PVY showed that the SCMV isolates formed a tightly
clustered group, with SCMV-MDB forming a separate branch. This indicated
that (i) the SCMV isolates are of one strain (SCMV-A) and not geographically
distinct species and (ii) SCMV-MDB is clearly distinct, and may represent another
potyvirus species.

Introduction
Sugarcane mosaic potyvirus (SCMV) is an important pathogen of sugarcane (Sac-
charum L. Interspecific hybrids) in Australia and causes significant yield loss in
susceptible varieties [6, 14, 15]. SCMV is a member of the genus Potyvirus in
the family Potyviridae. Potyvirus genomes consist of a single molecule of lin-
ear, positive sense ssRNA of approximately 10 kb which is translated as a single
polypeptide and post-translationally hydrolysed [10].
The taxonomic status of SCMV has undergone several revisions. SCMV was
initially believed to be a single potyvirus comprising a number of strains from
sugarcane, maize or sorghum [8]. This group was subsequently reclassified into
four distinct potyviruses, namely sugarcane mosaic (SCMV), maize dwarf mosaic
(MDMV), Johnsongrass mosaic (JGMV) and sorghum mosaic (SrMV) viruses
based on serology of their amino termini of the coat proteins (CP) and other
1146 J. A. Handley et al.

biological and biochemical properties [8]. In this classification the SCMV sub-
group included American strains A, B, D and E from sugarcane, MDB (formerly
MDMV-B) from maize, Australian strain SC, Isis, and Brisbane from sugarcane
and Australian strains BC, Bundaberg and Sabi from Queensland blue couch
grass, wild sorghum and sabi grass, respectively [8, 19]. The Australian strains
of SCMV from sugarcane have previously been referred to as SCMV strain A
based on symptoms in sugarcane differentials [5].
The spread of SCMV has been traditionally controlled through the use of
conventionally bred resistant cultivars and virus-free planting material. Breeding
for resistance has proven difficult, however, due to the complexity of the sugarcane
genome. As a consequence, susceptibility to SCMV still limits the cultivation of
several agronomically elite sugarcane cultivars [5].
Pathogen-derived resistance (PDR) provides an alternative strategy to control
SCMV, with coat-protein mediated resistance previously reported for many po-
tyvirus/host combinations [13, 16]. In the majority of these cases, it is unclear
whether resistance is mediated by the RNA or the encoded protein. The CP coding
region of SCMV strains SC (Australia) and MDB (USA) have been sequenced
and comprise 939 and 984 nucleotides, respectively [2]. Partial sequences of the
CP coding region of another five Australian strains have also been reported [19].
Potential resistance constructs have been generated in Australia using the CP and
the 30 UTR of SCMV-SC as transgene [12], but the ability of these constructs to
confer resistance has not yet been determined. As the success of PDR strategies
is largely influenced by the amount of sequence difference between the trans-
gene and challenging virus [6], we wished to determine the variability within the
CP-coding region of field isolates of SCMV in Australia.
This paper reports the cloning, sequencing and analysis of field isolates of
SCMV from Australia, USA and South Africa. The relationships between these
isolates and other related potyviruses is also discussed.

Materials and methods

Virus isolates
The eight Australian field isolates were collected from a range of infected commercial sug-
arcane cultivars growing in the Southern Queensland locations of Nambour, Isis, Bundaberg
and Brisbane (Table 1). Leaf samples were stored at −80 ◦ C before use. The US and South
African isolates (SCMV-A) were provided by Dr. M. Irey, United States Sugar Corporation,
Florida; Dr. M. Grisham, USDA, Sugarcane Research Unit, Louisianna; and Dr. R. Bechet,
South African Sugar Association Experiment Station, Natal, respectively (Table 1).

RT-PCR, cloning and sequencing

Nucleic acids were extracted from 250 mg of frozen leaf tissue [11] and were resuspended
in 625 mL dH2 O. First strand cDNA was synthesised and subsequently amplified by PCR as
previously described [3] using a variety of primers (Fig. 1). Amplified products were analysed
in 0.8% agarose gels and cloned directly into pGEM-T [3]. Both DNA strands were cycle
sequenced using either M13 forward or reverse universal primers.
 Coat protein variability in sugarcane mosaic potyvirus 1147

Table 1. Sources of SCMV field isolates

Name Origin Cultivar

l2 Isis, Australia Q146

l3 Isis, Australia Q145
l5 Isis, Australia Q124
l7 Isis, Australia Q155
BRIS Brisbane, Australia Q44
BUND Bundaberg, Australia Q137
N2 Nambour, Australia Q124
N7 Nambour, Australia Q137
SA Natal, South Africa B3337
USFL Florida, USA unknown
USLA Louisiana, USA unknown

Fig. 1. Sequence and location of oligonucleotides used for cDNA synthesis and PCR
amplification

Sequence analysis
Aligned nucleotide sequences were compared using DISCALC (G. Weiller & A. Gibbs, pers.
comm.) to produce distance matrices. DIPLOMO [18] was used to compare pairwise distance
estimates of nucleic acid and amino acid differences. The accession numbers of sequences
used for comparative purposes were SCMV-SC (D00094), SCMV-MDB (D00949), PVY
(M22470), MDMV-A (A34974), SrMV (U07219) and JGMV (A27631). CLUSTALW [17]
was used to align the nucleotide sequences and phylogenetic trees were prepared using
DNADIST, NEIGHBOR then DRAWGRAM programs in PHYLIP (version 3.5c, copyright
J. Felstein, University of Washington).

Results
Sequence of the CP-coding region
The CP-coding regions of the eleven field isolates were amplified by RT-PCR and
cloned. At least three clones of each isolate were sequenced in both directions;
1148 J. A. Handley et al.

at most, there were two nucleotide differences between the clones over the entire
CP-coding region for each isolate. The CP nucleotide sequences were submitted
to Genbank (accession numbers AF006728 to AF006738 inclusive).

Comparison of the SCMV nucleic acid and amino acid sequences

Six of the field isolates had CP coding regions 942 nt in length, whereas those
of the other five ranged from 903 to 921 nt. When compared to the published
SCMV-SC (SC) sequence (942 nt), [2], isolates I5, N2 and N7 had a gap of 21 nt
(nt 134–154) USLA had a gap of 24 nt (nt 184–208) and I2 had a gap of 39 nt (nt
130–169). These gaps corresponded to in-frame gaps of 7, 8 and 13 amino acids
in the respective translated sequences. All the gaps occurred at the 50 terminus of
the CP-coding region. In addition, the majority of non-synonymous nucleotide
and amino acid variability ocurred at the N-terminus of the CP, a region known to
be highly variable [9]. The majority of changes were clustered between residues 1
to 70. The region spanning residues 70 to 313 was almost unchanged between all
eleven field isolates, with only one difference, in the USFL isolate, at residue 127.
In contrast, the SC isolate had four differences from the field isolates between
residues 70 and 313.
The differences between the nucleotide sequences ranged from 0.2% (BUND
and I2; I2 and I7) to 4.1% (SC and USFL). The amino acid differences were
similar, ranging from no variability (I5, N2, and N7) to a maximum variability of
3.5% (SC and USFL; SC and SA).
The nucleotide and amino acid sequence differences were also compared to
each other using DIPLOMO (Fig. 2). This confirmed that most isolates had more
nucleotide than amino acid differences, especially those with fewest differences
e.g. I5, N2 and N7 (0.5 to 1.2% nucleotide variability but identical amino acid
sequences). A few isolates had more variable amino acid sequences than nucleic
acid sequences e.g. BUND and I2, with 0.22% variability in the nucleotide
sequences, and 0.31% variability in the amino acid sequences. However, the
spread of nucleotide/amino acid differences was not greater than that expected for
natural random processes, there were no subsets of isolates, and the dataset
appeared to be coherent.
The sequence most similar to all the others was from the BUND isolate with
0.22 to 2.76% variability when compared to each of the other sequences at the
nucleotide level, corresponding to 0.32 to 2.23% at the amino acid level.

Relationships between the SCMV isolates from sugarcane

Phylogenetic trees were constructed from the nucleotide sequences of the entire
CP coding region, the N-terminus (nt 1–249), the C-terminus (nt 250–942) and the
NIb coding region (Fig. 3). The NIb sequences for the eight Australian sequences
have been published previously [3], while the USFL, USLA and SA sequences
were cloned and sequenced as described by Handley et al. [3] and submitted to
Genbank (accession numbers U84578, U84579 and U84580).
 Coat protein variability in sugarcane mosaic potyvirus 1149

Fig. 2. DIPLOMO plot of nucleotide variability (x-axis) with amino acid variability (y-axis)
for the twelve SCMV isolates

Phylogenetic analysis of the nucleotide sequences of the CP C-terminus and

NIb coding regions of all twelve isolates (Fig. 3c and 3d) grouped the isolates
according to their provenance; the three non-Australian isolates grouped together,
with a further branch separating the two US isolates from the South African isolate.
Similarly, the Australian isolates were grouped according to their provenance
with the Nambour and Brisbane isolates and the Isis and Bundaberg isolates
grouped together, respectively. When phylogenetic analyses were done using the
entire CP and the CP 50 terminal coding regions (Fig. 3a and 3b) the variability
correlated less well with their provenance. The non-Australian isolates were still
grouped together, but the US isolates were not separated from the South African
isolate. The Australian isolates were grouped approximately according to their
provenance when the entire CP coding region was used (with the exception of the
I5 isolate), but when the 50 terminus of the CP coding region alone was used no
correlation with provenance was observed. Finally, the SC isolate, which had been
propagated in maize [2], was joined to the tree between all the other Australian
isolates and the non-Australian group in Fig. 3a, 3c and 3d. The SC isolate is the
most dissimilar of the Australian isolates, perhaps because it was propagated in
maize.

Relationship of the SCMV isolates from sugarcane to related potyviruses

A phylogenetic tree was made using the nucleotide sequences of the CP coding
region of the twelve SCMV isolates from sugarcane, SCMV-MDB (a member of
1150 J. A. Handley et al.

Fig. 3. Phylogenetic neighbor-joining tree of the SCMV nucleotide sequences from a the
CP coding region, b the N-terminus (nt 1–249) of the CP coding region, c the C-terminus of
the CP coding region and d the NIb coding region
 Coat protein variability in sugarcane mosaic potyvirus 1151

Fig. 4. Phylogenetic neighbor-joining tree of the nucleotide se-

quence of the CP coding region from twelve SCMV isolates from
sugarcane, one SCMV isolate from maize (SCMV-MDB), three
related potyviruses (MDMV-A, SrMV and JGMV) and the type
member of the potyvirus group (PVY)

the SCMV subgroup [8]), the related potyviruses SrMV, JGMV, MDMV-A and the
type member of the potyvirus group, potato virus Y (PVY) (Fig. 4). All twelve
SCMV isolates from sugarcane were grouped very closely together, reflecting
their variability of 0.2 to 4.1%. MDMV-A, SCMV-MDB and SrMV were grouped
together, as were JGMV and PVY. The twelve SCMV isolates from sugarcane
showed similar differences to MDMV-A, SCMV-MDB and SrMV (∼ 80%) but
less homology to PVY and JGMV (∼ 65%).

Discussion
This paper examined the variability in the CP-coding region of twelve isolates
of SCMV from Australia, USA and South Africa. The variability between all the
isolates was 0.2 to 4.1% in the nucleic acid sequences, corresponding to between
0.0 and 3.5% in the encoded amino acid sequences.
The stability and level of CP-mediated resistance has been shown to correlate
with the sequence differences between transgene and the challenge virus in many
virus/host combinations [1, 7, 16]. Hull and Davies [4] reported that transgenic
tobacco plants expressing the CP of cucumber mosic cucumovirus-C (subgroup
I) were resistant to infection with CMV-C but susceptible to CMV-WL, a member
of subgroup II. The two CMV subgroups share approximately 80% amino acid
sequence homology. Similarly, transgenic papaya transformed with the CP-coding
region of a US isolate of papaya ringspot potyvirus (PRSV-P) protected against
infection by the US isolate but did not protect against infection with isolates from
Australia and Thailand [16]. The US and Australia isolates showed approximately
96% and 98% similarity in the nucleotide and encoded amino acid sequences,
respectively [1].
In this study the sequence of the BUND isolate was found to be most similar
to all others in the group, with greater than 97% similarity to all other sequences
1152 J. A. Handley et al.

in the CP-coding region. These variability studies have led to the selection of the
BUND isolate for the generation of trangenes from the CP coding region because
it is most homologous to all the other isolates examined. Earlier transgenes were
derived from the SC isolate [12], which was slightly less similar (2.2 to 4.1%)
to the other isolates and hence may not provide effective protection against all
isolates of SCMV.
Phylogenetic analyses showed that the twelve SCMV isolates formed a tightly
clustered group with SCMV-MDB, MDMV-A, SrMV, JGMV and PVY more
distantly related. The average similarity between the SCMV isolates and SCMV-
MDB was approximately 80%, which was similar to that observed between the
SCMV isolates and MDMV-A and SrMV. Although SCMV-MDB does not infect
sugarcane it has previously been classified as a member of the SCMV subgroup
based on serology [8]. We suggest, on the basis of sequence analysis and host
range, that SCMV-MDB may be a distinct potyvirus rather than a strain of SCMV.
Further, the tight clustering of the other twelve SCMV isolates suggests that these
viruses should be classed as one strain and not geographical isolates as previously
described [8, 10]. Koike and Gillaspie [5] initially designated SCMV in Australia
as strain A; the close relationship between the Australian isolates and the three
non-Australian isolates appears to support this classification.

Acknowledgements
This work was funded by the Sugar Research and Development Corporation. We thank the
Bureau of Sugar Experiment Station extension officers for assistance with the collection
of the Australian isolates, Dr. M. Irey and Dr. M. Grisham for the US isolates and Dr R.
Bechet for the South African isolate. We thank Prof. A. J. Gibbs for advice on analysis and
presentation of the phylogenetic data in this paper.

References
1. Bateson MF, Henderson J, Chaleeprom W, Gibbs AJ, Dale JL (1994) Papaya ringspot
virus: isolate variability and the origin of PRSV type P (Australia). J Gen Virol 75:
3 547–3 553
2. Frenkel MJ, Jilka JM, McKern NM, Strike PM, Clark JM Jr, Shukla DD, Ward CW (1991)
Unexpected diversity in the amino terminal ends of the coat proteins of sugarcane mosaic
virus. J Gen Virol 72: 237–242
3. Handley JA, Smith GR, Dale JL, Harding RM (1996) Sequence diversity in the NIb
coding region of eight sugarcane mosaic potyviruses infecting sugarcane in Australia.
Arch Virol 141: 2 289–2 300
4. Hull R, Davies JW (1992) Approaches to nonconventional control of plant virus diseases.
Crit Rev Plant Sci 11: 17–33
5. Koike H, Gillaspie AG (1989) In: Ricaud C, Egan BT, Gillaspie, AG (eds) Diseases of
sugarcane – major diseases. Elsevier, Amsterdam, pp 301–332
6. Lomonosoff GP (1995) Pathogen-derived resistance to plant viruses. Annu Rev Phy-
topathol 33: 323–343
7. Shukla DD, Frenkel MJ, McKern NM, Ward CW, Jilka JW, Tosic M, Ford RE (1992)
Present status of the sugarcane mosaic subgroup of potyviruses. In: Barnett OW (ed)
Potyvirus taxonomy. Springer, Wien New York, pp 363–373 (Arch Virol [Suppl] 5)
 Coat protein variability in sugarcane mosaic potyvirus 1153

8. Shukla DD, Tosic M, Jilka JW, Ford RE, Toler RW, Langham MAC (1989) Taxonomy of
potyviruses infecting maize, sorghum and sugarcane in Australia and the United States
as determined by the reactivities of polyclonal antibodies directed towards virus-specific
N-termini of coat proteins. Phytopathology 79: 223–229
9. Shukla DD, Ward CW (1989) Structure of potyvirus coat proteins and its application in
the taxonomy of the potyvirus group. Adv Virus Res 36: 273–314
10. Shukla DD, Ward CW, Brunt AA (1994) The Potyviridae. Centre for Agriculture and
Biosciences International, Cambridge University Press, Cambridge
11. Smith GR, Van De Velde R (1994) Detection of sugarcane mosaic and Fiji disease virus
in diseased sugarcane using the polymerase chain reaction. Plant Dis 78: 557–561
12. Smith GR, Ford R, Frenkel MJ, Shukla DD, Dale JL (1992) Transient expression of
the coat protein of sugarcane mosaic virus in sugarcane protoplasts and expression in
Escherichia coli. Arch Virol 125: 15–23
13. Stark DM, Beachy RN (1989) Protection against potyvirus infection in transgenic plants:
evidence for broad spectum resistance. Bio/Technology 7: 1 257–1 262
14. Teakle DS, Grylls NE (1973) Four strains of sugarcane mosaic virus infecting cereals
and other grasses in Australia. Aust J Agric Res 24: 465–477
15. Teakle DS, Shukla DD, Ford R (1989) Sugarcane mosaic virus. AAB Descriptions of
Plant Viruses, No 342
16. Tennant PF, Gonsalves C, Ling K-S, Fitch M, Manshardt R, Slightom JL, Gonsalves D
(1994) Differential protection against papaya ringspot virus isolates in coat protein gene
transgenic papaya and classically cross-protected papaya. Phytopathology 84: 1 359–
1 366
17. Thompson JD, Higgins DJ, Gibson TJ (1994) CLUSTALW: improving the sensitivity of
progressive multiple sequence alignment through sequence weighting, position specific
gap penalties and weight matrix choice. Nucleic Acids Res 22: 4 673–4 680
18. Weiller GF, Gibbs AJ (1995) DIPLOMO: The tool for a new type of phylogenetic
analysis. CABIOS 11: 535–540
19. Xiao XW, Frenkel MJ, Teakle DS, Ward CW, Shukla DD (1993) Sequence diversity in
the amino terminal region of the coat proteins of seven strains of SCMV correlates with
their host range and other biological properties. Arch Virol 132: 399–408
Authors’ address: Dr. R. M. Harding, Centre for Molecular Biotechnology, School of Life
Sciences, Queensland University of Technology, G.P.O. Box 2 434, Brisbane, Queensland
4001, Australia.
Received March 7, 1997