S

Hi

Ma

b

a

ARRAA

KVAIWGR

Pntaahwenc

aT

h0

Virus Research 184 (2014) 87–92

Contents lists available at ScienceDirect

Virus Research

j ourna l h o mepa ge: www.elsev ier .com/ locate /v i rusres

hort communication

ardenbergia mosaic virus: Crossing the barrier between native andntroduced plant species

.A. Kehoea,b,∗, B.A. Couttsa,b, B.J. Buirchell a,b, R.A.C. Jonesa,b

School of Plant Biology and Institute of Agriculture, Faculty of Science, University of Western Australia, Crawley, WA 6009, AustraliaCrop Protection and Lupin Breeding Branches, Department of Agriculture and Food Western Australia, Bentley Delivery Centre, Perth, WA 6983, Australia

r t i c l e i n f o

rticle history:eceived 20 January 2014eceived in revised form 17 February 2014ccepted 18 February 2014vailable online 1 March 2014

eywords:irus emergencegro-ecological interface

ndigenous virushole genomes

enetic diversityecombination

a b s t r a c t

Hardenbergia mosaic virus (HarMV), genus Potyvirus, belongs to the bean common mosaic virus (BCMV)potyvirus lineage found only in Australia. The original host of HarMV, Hardenbergia comptoniana, familyFabaceae, is indigenous to the South-West Australian Floristic Region (SWAFR), where Lupinus spp. aregrown as introduced grain legume crops, and exist as naturalised weeds. Two plants of H. comptonianaand one of Lupinus cosentinii, each with mosaic and leaf deformation symptoms, were sampled from asmall patch of disturbed vegetation at an ancient ecosystem–recent agroecosystem interface. Potyvirusinfection was detected in all three samples by ELISA and RT-PCR. After sequencing on an Illumina HiSeq2000, three complete and two nearly complete HarMV genomes from H. comptoniana and one completeHarMV genome from L. cosentinii were obtained. Phylogenetic analysis which compared (i) the four newcomplete genomes with the three HarMV genomes on Genbank (two of which were identical), and (ii)coat protein (CP) genes from the six new genomes with the 38 HarMV CP sequences already on Genbank,revealed that three of the complete and one of the nearly complete new genomes were in HarMV clade I,one of the complete genomes in clade V and one nearly complete genome in clade VI. The complete HarMVgenome from L. cosentinii differed by only eight nucleotides from one of the HarMV clade I genomes from anearby H. comptoniana plant, with only one of these nucleotide changes being non-synonymous. Pairwise

comparison between all the complete HarMV genomes revealed nucleotide identities ranging between82.2% and 100%. Recombination analysis revealed evidence of two recombination events amongst thesix complete genomes. This study provides the first report of HarMV naturally infecting L. cosentinii andthe first example for the SWAFR of virus emergence from a native plant species to invade an introducedplant species.

Crown Copyright © 2014 Published by Elsevier B.V. All rights reserved.

Hardenbergia mosaic virus (HarMV) (genus Potyvirus, familyotyviridae) is a recently described virus from the native peren-ial legume Hardenbergia comptoniana, a plant species endemico the South-West Australian Floristic Region (SWAFR) (Hoppernd Gioia, 2004; Webster et al., 2007; Coutts et al., 2011; Wyliend Jones, 2011). The SWAFR is a species rich global diversityot spot with around 8000 indigenous plant species where thereas no plant cultivation until Europeans arrived in 1829 (Myers

t al., 2000; Hopper and Gioia, 2004). Coat protein (CP) geneucleotide (nt) sequencing placed 30 HarMV isolates into eightlades with up to 21% nt differences between sequences (Webster

∗ Corresponding author at: Crop Protection Branch, Department of Agriculturend Food Western Australia, Bentley Delivery Centre, Perth, WA 6983, Australia.el.: +61 9368 3333.

E-mail address: monica.kehoe@agric.wa.gov.au (M.A. Kehoe).

ttp://dx.doi.org/10.1016/j.virusres.2014.02.012168-1702/Crown Copyright © 2014 Published by Elsevier B.V. All rights reserved.

et al., 2007). A ninth clade was suggested after the CP sequencesof six additional HarMV isolates were added and compared withthe initial 30 (Coutts et al., 2011). When their complete genomeswere sequenced using Illumina GAIIx technology, two HarMV iso-lates found co-infecting a single H. comptoniana plant differed by18% at the nt level (Wylie and Jones, 2011). The high degree of ntdiversity over a small geographic range demonstrated by HarMV ischaracteristic of viruses that co-evolved with native plants locallyover a long period of time (Spetz et al., 2003; Webster et al., 2007;Coutts et al., 2011).

Potyviruses found in Australia fall into two groups, with roughlyhalf of them being isolated from cultivated plants and found inother parts of the world. Potyviruses isolated from Lupinus spp.

so far fall into this category. The other half constitute a potyviruslineage found only in Australia which belongs to the bean com-mon mosaic virus (BCMV) group (Gibbs et al., 2008; Coutts et al.,2011). Members of the Australian potyvirus lineage have only been

8 s Research 184 (2014) 87–92

iavCephY2

1d(cbaoifipiiac2sncSpfa

aatMLc

aprhuawAumi

Atsfibawwtctsrv

oms

and

sequ

ence

dat

a

for

thre

e

leaf

sam

ple

s

from

pla

nts

infe

cted

wit

h

har

den

berg

ia

mos

aic

viru

s.a

ID

Leaf

sym

pto

msb

No.

of

read

sob

tain

edN

o.

of

read

saf

ter

trim

min

g

No.

of

con

tigs

pro

du

ced

(CLC

c )

Sam

ple

sequ

ence

ID

Acc

essi

onn

um

ber

Con

tig

len

gth

(nt)

Ave

rage

cove

rage

No.

of

read

sm

app

ed

toco

nti

g

ofin

tere

st

(CLC

)

Ref

eren

cese

quen

ce

use

dfo

r

map

pin

g(G

enei

ousc )

Len

gth

ofco

nse

nsu

s(n

t)

Ave

rage

cove

rage

Nu

mbe

r

of

read

sm

app

ed

to

ref.

sequ

ence

(Gen

eiou

s)

Fin

al

sequ

ence

len

gth

(nt)

d

rgia

ana

M, L

D

17,6

16,8

43

17,1

82,5

34

1691

MD

2

KJ1

5215

2

9672

5072

499,

921

NC

0153

94

9751

5354

522,

746

9647

iM

, LD

13,1

54,2

90

12,8

40,9

28

677

MD

3

KJ1

5215

3 96

95

3516

347,

261

NC

0153

94

9726

3612

352,

433

9647

nian

aM

, LD

12,5

38,7

04

12,2

39,1

75

1806

MD

4-A

KJ1

5215

4

9648

2975

292,

188

–

–

–

–

9564

MD

4-B

KJ1

5215

5

8975

194

17,7

38

–

–

–

–

8906

MD

4-C

KJ1

5215

6

9635

343

33,7

16

–

–

–

–

9659

MD

4-D

KJ1

5215

7

9600

298

29,1

20

–

–

–

–

9668

pto

mat

ic

pla

nt

sam

ple

s

wer

e

coll

ecte

d

from

a

dis

turb

ed

pat

ch

of

nat

ive

vege

tati

on

at

the

agro

-eco

logi

cal i

nte

rfac

e.

Sam

ple

s

wer

e

sequ

ence

d

on

an

Illu

min

a

HiS

eq20

00.

mp

tom

des

crip

tion

s:

M, m

osai

c;

LD, l

eaf d

efor

mat

ion

.

use

d

for

asse

mbl

y

of

dat

a:

CLC

, CLC

Gen

omic

s

Wor

kben

ch

6.5

(CLC

bio)

;

Gen

eiou

s

6.1.

6

(Bio

mat

ters

).u

ence

s

for

MD

2

and

MD

3

con

sist

ed

of

the

con

sen

sus

betw

een

de

nov

o

asse

mbl

y

and

the

map

ped

con

sen

sus.

Fin

al

sequ

ence

s

for

MD

4A-D

con

sist

ed

of

thei

r

de

nov

o

con

tigs

only

.

8 M.A. Kehoe et al. / Viru

solated from native plants or naturalised weed species introduceds potential pasture species, apart from passionfruit woodinessirus (PWV) from cultivated Passiflora spp. (Gibbs et al., 2008;outts et al., 2011). HarMV is the best studied member of this lin-age. At the CP level, HarMV is most closely related to six otherotyviruses found only in Australia: clitoria chlorosis virus (CliVY),ibbertia virus Y (HiVY), siratro 1 virus Y (S1VY), siratro 2 virus

(S2VY), passiflora mosaic virus (PaMV) and PWV (Coutts et al.,011).

Lupinus spp. were introduced to the SWAFR around the early900s where they were used as sheep feed initially, and lateromesticated as grain legume crops for rotation with cerealsFrench et al., 2008). L. angustifolius (narrow-leafed lupin), L.osentinii (sandplain lupin), L. luteus (yellow lupin) and L. muta-ilis (pearl lupin) became infected by HarMV experimentally in

glasshouse environment (Webster et al., 2007). Also, naturallyccurring aphids spread HarMV from introduced H. comptoniananfector plants to L. angustifolius plants growing in experimentaleld plots (Luo et al., 2011). However, except within these fieldlots, HarMV has not been found infecting any Lupinus spp. or other

ntroduced plant species naturally. We therefore investigated anntroduced lupin-H. comptoniana interface scenario involving anncient ecosystem (i.e. native Australian plants) and a recent agroe-osystem (i.e. introduced species) in the SWAFR (Webster et al.,007; Jones, 2009; Alexander et al., 2014; Vincent et al., 2014). Asuch, we report the first detection of HarMV infecting L. cosentiniiaturally, and the first example of an indigenous virus effectivelyrossing the barrier between native and introduced species in theWAFR. We also present four new complete and two nearly com-lete HarMV genome sequences, including one complete sequencerom L. cosentinii and the remaining sequences from H. comptoni-na.

Leaf tissue from two H. comptoniana and one L. cosentinii plantsll showing leaf mosaic and deformation symptoms, were collectedt the agro-ecological interface from a patch (50 m × 5 m) of dis-urbed native vegetation surrounding experimental field plots at

edina near Perth, Western Australia (WA). Within the patch, the. cosentinii plant sampled was growing as a naturalised weed inlose proximity (4 m) to the sampled H. comptoniana plants.

The samples were tested with generic potyvirus monoclonalntibodies (Agdia, USA) using the antigen-coated indirect ELISArotocol of Torrance and Pead, 1986. Absorbance values (A405) wereegarded as positive when more than three times those of theealthy sap control. For testing by RT-PCR, total RNA was extractedsing a Spectrum Plant Total RNA kit (Sigma-Aldrich, Australia)ccording to manufacturers’ instructions. Reverse transcriptionas performed with Improm-II reverse transcriptase (Promega,ustralia) using the random primers provided according to man-facturer’s instructions. PCR was performed with the GoTaq greenaster mix (Promega, Australia). PCR primers for generic potyvirus

dentification were from Webster, 2008.Total RNA from each potyvirus positive sample was sent to the

ustralian Genome Research Facility (AGRF) for library prepara-ion and barcoding (24 samples per lane) before 100 bp paired-endequencing on an Illumina HiSeq2000. For each sample, reads wererst trimmed using CLC Genomics Workbench 6.5 (CLCGW) (CLCio) with the quality scores limit set to 0.01, maximum number ofmbiguities to two and removing any reads with <30 nt. Contigsere assembled using the de novo assembly function of CLCGWith automatic word size, automatic bubble size, minimum con-

ig length 500, mismatch cost two, insertion cost three, deletionost three, length fraction 0.5 and similarity fraction 0.9. Con-

igs were sorted by length and the longest subjected to a BLASTearch (Altschul et al., 1990). In addition, for samples MD2 and MD3eads were also imported into Geneious 6.1.6 (Biomatters) and pro-ided with a HarMV reference sequence obtained from Genbank Ta

ble

1H

osts

, sym

pt

Plan

t/h

ost

1.H

arde

nbe

com

pton

i2.

Lupi

nus

cose

ntin

i3.

H. c

ompt

o

aTh

ree

sym

bC

oded

syc

Prog

ram

sd

Fin

al

seq

M.A. Kehoe et al. / Virus Research 184 (2014) 87–92 89

Fig. 1. Neighbour-joining relationship phylograms obtained from alignment of complete and nearly complete hardenbergia mosaic virus (HarMV) genomes trimmed to thelength of the shortest nearly complete HarMV genome from this study (MD4-B with 8906 nt). The alignments were generated in MEGA 5.2.1 using ClustalW and tree branchesw of pasr Compn

(mtfa(

w(nGa8waenetamecsbewtRBmSe(

Ltu

i(t(aMcc

ere bootstrapped with 1000 replications. The trees were rooted with a sequence

elative to HarMV from the Australian potyviruses group with a complete genome.

ew isolates from this study shown in grey.

NC015394). Mapping was performed with minimum overlap 10%,inimum overlap identity 80%, allow gaps 10% and fine tuning set

o iterate up to 10 times. A consensus between the contig of interestrom CLCGW and the consensus from mapping in Geneious was cre-ted in Geneious by alignment with Clustal W. Open reading framesORFs) were predicted and annotations made using Geneious.

The new sequences were aligned by Clustal W in MEGA 5.2.1ith those retrieved from Genbank prior to phylogenetic analysis

Tamura et al., 2011). Phylogenetic analysis compared (i) the fourew HarMV genomes with the three HarMV genomes available onenbank (two of which were identical), (ii) all available genomesfter trimming to the length of the shortest genome (MD4-B with868 nt), (iii) complete CP genes from the five of the new genomesith the 38 HarMV CP sequences already on Genbank, and (iv)

ll available CP genes after trimming to the length of the short-st nearly complete HarMV CP from this study (MD4-B with 404t). Neighbour-Joining trees were made using the number of differ-nces model with a bootstrap value of 1000, Maximum Likelihoodrees using the Tamura-Nei model with a bootstrap value of 1000,nd Minimum Evolution trees using the number of differencesodel with a bootstrap value of 1000. Tables of percentage nt differ-

nces were calculated for the complete genomes using the pairwiseomparison function with the number of differences model. Finalequences were submitted to Genbank with their Accession num-ers KJ152152–KJ152157 (Table 1). The RDP4 package (Martint al., 2010) was used to detect recombination between six HarMVhole genomes and one PWV whole genome. Default parame-

ers were used for the six programmes implemented within RDP:DP (Martin and Rybicki, 2000), GENECONV (Padidam et al., 1999),ootscan (Martin et al., 2005), MaxChi (Maynard Smith, 1992), Chi-aera (Posada and Crandall, 2001), 3Seq (Boni et al., 2007) and

iScan (Gibbs et al., 2000). A recombination pattern was consid-red acceptable if detected by four or more of these programmesOhshima et al., 2002; Wylie and Jones, 2009).

Two samples from H. comptoniana (MD2 and MD4) and one from. cosentinii (MD3), returned a generic potyvirus positive result fromesting by ELISA using generic potyvirus antibodies and RT-PCRsing generic potyvirus primers.

The overall numbers of 100 nt reads obtained from sequenc-ng on the Illumina HiSeq2000 were 17,616,843 (MD2), 13,154,290MD3) and 12,538,704 (MD4) (Table 1). After sequences wererimmed the remaining reads were 17,182,534 (MD2), 12,840,928MD3) and 12,239,175 (MD4). The numbers of contigs produced

fter de novo assembly were 1691, 677 and 1806 for samples MD2,D3 and MD4, respectively. Blastn analysis for MD2 revealed a

ontig of 9672 nt, with an average coverage of 5072 times andonstructed from 499,921 reads which was most closely related

sionfruit woodiness virus (PWV) accession number NC01479, which is the closestlete HarMV genomes with isolates retrieved from Genbank shown in bold text and

to HarMV accession no. HQ161081. MD3 Blastn analysis revealeda contig of 9695 nt, with an average coverage of 3516 times andconstructed from 347,261 reads which was most closely related toHarMV (HQ161081). For MD4, Blastn analysis revealed four con-tigs of interest (i) MD4-A with a contig of 9648 nt, with averagecoverage 2975, constructed from 292,188 reads which was mostclosely related to HarMV (HQ161081); (ii) MD4-B with a contig of8975 nt, with average coverage of 194, constructed from 17,738reads which was most closely related to HarMV (HQ161081); (iii)MD4-C with a contig of 9635 nt, with average coverage of 343,constructed from 33,716 reads which was most closely related toHarMV (HQ161081); and (iv) MD4-D with a contig of 9600 nt, withaverage coverage of 298, constructed from 29,120 reads which wasmost closely related to HarMV (HQ161080). MD2 and MD3 weresubjected to reference assembly against HarMV accession num-ber NC015394 obtained from Genbank. MD2 produced a contig of9751 nt with average coverage of 5354 times and constructed from522,746 reads. MD3 produced a contig of 9726 nt, with averagecoverage 3612 times and constructed from 352,433 reads. Theseresults are comparable with those from de novo assembly of thesetwo samples. The final sequences obtained were complete genomeswith lengths of 9647–9668 nt, with the exception of MD4-A (miss-ing 114 nt from the 5′ UTR region) and MD4-B (missing 91 ntfrom the 5′ UTR region and 675 nt from the 3′ UTR region). Finalsequences for MD2 and MD3 consisted of a consensus of the contigfrom de novo assembly and the mapped consensus sequence. Finalsequences for MD4-A to MD4-D consisted of their de novo contigsonly.

Phylogenetic analysis with complete genomes alone placedMD2, MD3 and MD4-C into HarMV clade I while MD4-D proved tobe the first whole genome representing HarMV clade V. When allgenomes were analysed after trimming to the length of the short-est genome (8906 nt), the groupings were the same with MD4-Bplaced in clade I and MD4-A in clade VI (Fig. 1). When their com-plete CP genes were analysed, MD2, MD3 and MD4-C were placedin clade I, MD4-D in clade V and MD4-A in clade VI. When all CPgenes were analysed after trimming to the length of the short-est sequence (404 nt), the groupings were the same with MD4-Bplaced in clade I (Fig. 2). One exception was HQ161080, which wasplaced in clade IV when complete CPs were analysed, but in cladeV for partial CPs. When all genomes or coat proteins were ana-lysed using Maximum Likelihood or Minimum Evolution methods,the tree topologies shown were the same as the Neighbour-Joining

method. A pairwise comparison of the nt percentage identities ofthe complete genomes revealed identities ranging between 82.2%and 100%. When the PWV sequence used as an outgroup in thephylogenetic analysis was included in the comparison, percentage

90 M.A. Kehoe et al. / Virus Research 184 (2014) 87–92

Fig. 2. Neighbour-joining relationship phylograms obtained from alignment of complete and partial hardenbergia mosaic virus (HarMV) coat protein (CP) sequences trimmedto the length of the shortest nearly complete HarMV CP from this study (MD4-B with 404 nt). The alignments were generated in MEGA 5.2.1 using ClustalW and tree branchesw passio hat wt

i6tnfi(

ere bootstrapped with 1000 replications. The trees were rooted with a sequence off HarMV from the Australian potyviruses group. Isolates retrieved from Genbank this study shown in grey.

dentities between PWV and the HarMV sequences ranged from8.6–69.1%. Sequence MD3 from L. cosentinii was 99.9% identicalo that of MD2, which was from a H. comptoniana plant growing

earby (approx. 4 m). MD3 had just eight separate nt differences

rom MD4, only one of which was non-synonymous with a Ser-ne instead of the Glycine at the position 6799 in the nt sequencepositioned within the Nia-Pro gene). The other HarMV sequences

onfruit woodiness virus (PWV) accession number NC01479, which is a close relativeere part of a complete genome sequence shown in bold text and new isolates from

from this same site had sequence identities 82.2–98.8% to eachother.

Recombination analysis conducted on the complete HarMV

genomes (four from this study and three from Genbank) revealedtwo separate recombination events (Fig. 3). The sequence ofHQ161080 was predicted as a recombinant, with MD4-D as a majorparent and MD3 the minor parent in all programmes within RDP4

M.A. Kehoe et al. / Virus Research 184 (2014) 87–92 91

Fig. 3. Recombination patterns found when six hardenbergia mosaic virus (HarMV) and one passionfruit woodiness virus (PWV) complete genomes were analysed. Thel re. Twc

etbHagft

griivgHwJlAegb(

tv(O2nts2btMwbatduis

ocation of the potyvirus genes are represented in the diagram at the top of the figuorresponds to is indicated at the right hand side (i.e. clades I, IV or V).

xcept for Geneconv, with the range in P-values from 1.978 × 10−6

o 1.083 × 10−25. The breakpoint was estimated at a positionetween 9085 and 9173 nt. The second event was located inQ161081 where the major parent was predicted to be MD4-Cnd the minor parent MD4-D. This was predicted in all pro-rammes with the exception of 3Seq, with the range in P-valuesrom 7.395 × 10−3 to 3.354 × 10−45. The breakpoint was estimatedo be at a position between 9084 and 9160 nt.

The addition of four complete and two nearly complete newenomes of HarMV adds to the molecular knowledge about thisecently characterised indigenous Australian virus. When analysedn conjunction with the three sequences available in Genbank (twodentical), we found that the groupings resembled those seen pre-iously in analyses involving the CP alone. There are now wholeenome representatives for four of the nine suggested clades forarMV (Webster et al., 2007; Coutts et al., 2011). Previously, thereere just three sequences representing two clades (Wylie and

ones, 2011). Comparison of HarMV with PWV at whole genomeevel revealed the extent of divergence between these two species.t the CP level, PWV diverged from HarMV by 24.1–28.7% (Webstert al., 2007; Coutts et al., 2011), but with whole genomes the diver-ence was >30%. This is a much higher than the species demarcationorderline of 23–24% suggested for potyviruses at the nt levelAdams et al., 2005).

Recombination is a very successful method of viral evolu-ion with many well studied examples of economically importantiruses, including potyviruses such as bean yellow mosaic virusBYMV), potato virus Y (PVY) and turnip mosaic virus (TuMV) (e.g.,hshima et al., 2002; Wylie and Jones, 2009; Karasev and Gray,013; Ohshima, 2013). Recombination analysis revealed recombi-ation amongst HarMV whole genomes with patterns that matchedheir phylogenetic groupings. Following the addition of six new CPequences to the sequences from Webster et al., 2007, Coutts et al.,011 suggested that the HarMV phylogenetic clades IV and V shoulde combined into one clade. However, based on our recombina-ion analysis, we suggest that they should remain separate because

D4-D (which represents the first genome sequence from clade V)as not a recombinant, whilst HQ161080 from clade IV was recom-

inant (Fig. 3). Moreover, as next generation sequencing techniquesre becoming cheaper, more accessible and more widely used,he number of complete plant virus genomes available in public

atabases is rising rapidly. Given the readiness of potyviruses tondergo recombination both within species and with other species,

t is becoming increasingly important to use recombination analy-is to provide a thorough understanding of the genetic makeup and

o recombination patterns were found, A and B. The HarMV clade that each isolate

phylogenetic placement of a potyvirus. Addition of further com-plete HarMV genome sequences to those studied here would allowother phylogenetic HarMV groupings defined from CP sequencesto be confirmed for whole genomes. Although the recombinationanalysis conducted here only included HarMV and PWV, it would beinteresting to include whole genomes of some of the more closelyrelated Australian potyviruses such as PaMV, ClCV, HibVY, S1VYand S2VY, once they become available. Ongoing study of the evolu-tion of such a geographically distinct group of viruses provides anexciting prospect!

This study provides the first report of HarMV naturally infect-ing L. cosentinii, a lupin species that became naturalised after itsintroduction to the SWAFR. It also constitutes the first example forthe SWAFR of an indigenous virus that has made a host speciesjump (Woolhouse et al., 2005), successfully crossing the ancientecosystem-recent agrosystem interface. Luo et al., 2011 had previ-ously demonstrated HarMV spread to L. angustifolius in an artificialfield plot situation with deliberately introduced H. comptonianainfector plants.

So far, the natural host range of HarMV is limited to H. comp-toniana and H. violacea, with only the former being native to theSWAFR (Webster et al., 2007). So it seems that the nt and aminoacid changes required to facilitate this jump may be few, giventhere were just eight nt and one amino acid differences between theHarMV sequence from L. cosentinii (MD3) and the closest HarMVsequence from H. comptoniana at the same site (MD2). Little isknown regarding the feeding preferences associated with the nat-ural aphid vectors of HarMV. It is possible they have a feedingpreference for H. comptoniana, or that their transmission efficien-cies are lower on their non-preferred hosts (e.g. L. cosentinii).Further experimentation would provide insight and understandingregarding the limited natural host range for HarMV. Its artificialhost range includes species in four additional families (Websteret al., 2007; Coutts et al., 2011; Vincent et al., 2014).

Our research provides the first example for the SWAFR ofvirus emergence from native plants to invade an introduced plantspecies. The symptoms in the L. cosentinii plant infected withHarMV were leaf mosaic and leaf deformation, but more severesymptoms such as plant death and stunting were recorded withexperimental HarMV infection of L. cosentinii (Webster et al., 2007).The natural distribution of H. comptoniana in the SWAFR covers

much the same area as the distribution of L. cosentinii. More-over, this region borders a larger area of south west Australiawhere most of the grain legume crop species L. angustifolius is pro-duced (Western Australian Herbarium, 1998). Further studies are

9 s Resea

wcmvcdniet2e

A

(puab

R

A

A

A

B

C

C

F

G

G

H

J

2 M.A. Kehoe et al. / Viru

arranted to determine whether HarMV infection is cause foroncern for the grain legume industry. Viruses of native plant com-unities are sometimes ignored or poorly researched, even though

iruses co-evolved with wild plants well before plant domesti-ation and these communities are likely to contain potentiallyamaging viral pathogens. Here we have further evidence that asew contact between native plants and introduced crops or weeds

ncreases due to mans’ activities and climate change, the threat ofmerging viruses from indigenous plants to introduced plants is seto increase (Cooper and Jones, 2006; Webster et al., 2007; Jones,009; Jones and Barbetti, 2012; Alexander et al., 2014; Vincentt al.,2014).

cknowledgements

This research was funded by an Australian Postgraduate AwardAPA), and an Australian Grains Research and Development Cor-oration (GRDC) Studentship, Project number GRS10039. It wasndertaken using the facilities at the Department of Agriculturend Food Western Australia. This study forms part of a PhD projecty the first author at the University of Western Australia.

eferences

dams, M.J., Antoniw, J.F., Fauquet, C.M., 2005. Molecular criteria for genus andspecies discrimination within the family Potyviridae. Arch. Virol 150, 459–479.

lexander, H.M., Mauck, K.E., Whitfield, A.E., Garrett, K.A., Malmstrom, C.M., 2014.Plant-virus interactions and agro-ecological interface. Eur. J. Plant Pathol. 138,529–574.

ltschul, S.F., Gish, W., Miller, W., Myers, E.W., Lipman, D.J., 1990. Basic local align-ment search tool. J. Mol. Biol. 215, 403–410.

oni, M.F., Posada, D., Feldman, M.W., 2007. An exact nanoparametric method forinferring mosaic structure in sequence triplets. Genetics 176, 1035–1047.

ooper, I., Jones, R.A.C., 2006. Wild plants and viruses: under-investigated ecosys-tems. Adv. Virus Res. 67, 1–40.

outts, B.A., Kehoe, M.A., Webster, C.G., Wylie, S.J., Jones, R.A.C., 2011. Indigenous andintroduce potyviruses of legumes and Passiflora spp. from Australia: biologicalproperties and comparison of coat protein nucleotide sequences. Arch. Virol 156,1757–1774.

rench, B., Shea, G., Buirchell, B., 2008. Introduction and history. In: White, P., French,B., McLarty, A. (Eds.), Producing Lupins, Department of Agriculture and FoodWestern Australia Bulletin 4720. , 2nd ed. Department of Agriculture and Food,South Perth, Western Australia.

ibbs, M.J., Armstrong, J.S., Gibbs, A.J., 2000. Sister-scanning: a Monte Carlo pro-cedure for assessing signals in recombinant sequences. Bioinformatics 16,573–582.

ibbs, A.J., Mackenzie, A.M., Wei, K.-J., Gibbs, M.J., 2008. The potyviruses of Australia.Arch. Virol 153, 1411–1420.

opper, S.D., Gioia, P., 2004. The southwest Australian floristic region: evolution and

conservation of a global hot spot of biodiversity. Annu. Rev. Ecol. Evol. Syst 35,623–650.

ones, R.A.C., 2009. Plant virus emergence and evolution: origins, new encounterscenarios, factors driving emergence, effects of changing world conditions, andprospects for control. Virus Res. 141, 113–130.

rch 184 (2014) 87–92

Jones, R.A.C., Barbetti, M.J., 2012. Influence of climate change on plant disease infec-tions and epidemics caused by viruses and bacteria. CAB Rev. 7 (22), 1–32http://www.cabi.org/cabreviews

Karasev, A.V., Gray, S.M., 2013. Genetic Diversity of Potato virus Y complex. Am. J.Potato Res. 90, 7–13.

Luo, H., Wylie, S.J., Coutts, B., Jones, R.A.C., Jones, M.G.K., 2011. A virus of an isolatedindigenous flora spreads naturally to an introduced crop species. Ann. Appl. Biol.159, 339–347.

Martin, D.P., Lemey, P., Lott, M., Moulton, V., Posada, D., Lefeuvre, P., 2010. RDP3: aflexible and fast computer program for analyzing recombination. Bioinformatics26, 2462–2463.

Martin, D.P., Posada, D., Crandall, K.A., Williamson, C., 2005. A modified bootscanalgorithm for automated identification of recombinant sequences and recombi-nation breakpoints. AIDS Res. Hum. Retroviruses 21, 98–102.

Martin, D., Rybicki, E., 2000. RDP: detection of recombination amongst alignedsequences. Bioinformatics 16, 562–563.

Maynard Smith, J., 1992. Analyzing the mosaic structure of genes. J. Mol. Evol. 34,126–129.

Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonesca, G.A.B., Kent, J., 2000.Biodiversity hot spots for conservation priorities. Nature 403, 853–858.

Ohshima, K., Yamaguchi, Y., Hirota, R., Hamamoto, T., Tomimura, K., Tan, Z., Sano, T.,Azuhata, F., Walsh, J.A., Fletcher, J., Chen, J., Gera, A., Gibbs, A., 2002. Molecularevolution of Turnip mosaic virus: evidence of host adaptation, genetic recombi-nation and geographical spread. J. Gen. Virol. 83, 1511–1521.

Ohshima, K., 2013. Studies on the molecular evolution of potyviruses. J. Gen. PlantPathol. 79, 448–452.

Padidam, M., Sawyer, S., Fauquet, C.M., 1999. Possible emergence of new gemi-niviruses by frequent recombination. Virology 265, 218–225.

Posada, D., Crandall, K.A., 2001. Evaluation of methods for detecting recombinationfrom DNA sequences: computer simulations. Proc. Natl. Acad. Sci. U. S. A. 98,13757–13762.

Spetz, C., Taboada, A.M., Darwich, S., Ramsell, J., Salazar, L.F., Valkonen, J.P.T., 2003.Molecular resolution of a complex of potyviruses infecting solanaceous crops atthe centre of origin in Peru. J. Gen. Virol. 84, 2565–2578.

Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011. MEGA5:Molecular evolutionary genetics analysis using maximum likelihood, evolution-ary distance, and parsimony methods. Mol. Biol. Evol. 28, 2731–2739.

Torrance, L., Pead, M.T., 1986. The application of monoclonal antibodies to routinetests for two plant viruses. In: Jones, R.A.C., Torrance, L. (Eds.), Developments inapplied biology. 1 Developments and applications on virus testing. Associationof Applied Biologists, Wellesbourne, UK, pp. 103–118.

Vincent, S.J., Coutts, B.A., Jones, R.A.C., Vincent, S.J., Coutts, B.A., Jones, R.A.C., 2014.Effects of introduced and indigenous viruses on native plants: exploring theirdisease causing potential at the agro-ecological interface. PLoS One, PONE-D-13–51662R1 (in press).

Webster, C.G., Coutts, B.A., Jones, R.A.C., Jones, M.G.K., Wylie, S.J., 2007. Virus impactat the interface of an ancient ecosystem and a recent agroecosystem: studies onthree legume-infecting potyviruses in the southwest Australian floristic region.Plant Pathol. 56, 729–742.

Webster, C.G., 2008. Characterisation of Hardenbergia mosaic Virus and Develop-ment of Microarrays for Detecting Viruses in Plants. In: PhD thesis. MurdochUniversity, Perth, Western Australia.

Western Australian Herbarium, 1998. FloraBase – the Western Australian Flora.Department of Parks and Wildlife, Available at: http://florabase.dpaw.wa.gov.au

Woolhouse, D.T., Haydon, D.T., Anita, R., 2005. Emerging pathogens: the epidemiol-ogy and evolution of species jumps. Trends Ecol. Evol. 20, 238–244.

Wylie, S.J., Jones, R.A.C., 2009. Role of recombination in the evolution of

host specialization within Bean yellow mosaic virus. Phytopathology 99,512–518.

Wylie, S.J., Jones, M.G.K., 2011. Characterisation and quantitation of mutant andwild-type genomes of Hardenbergia mosaic virus isolates co-infecting a wildplant of Hardenbergia comptoniana. Arch. Virol 156, 1251–1255.