Palmisanoe al. Vi ology Jou nal (2025) 22:319
h ps://doi.o g/10.1186/s12985-025-02892-7
RESEARCH Open Access
© The Au ho (s) 2025. Open Access This a icle is licensed unde a C ea i e Commons A ibu ion-NonComme cial-NoDe i a i es 4.0
In e na ional License, which pe mi s any non-comme cial use, sha ing, dis ibu ion and ep oduc ion in any medium o o ma , as long
as you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons licence, and indica e i
you modi ied he licensed ma e ial. You do no ha e pe mission unde his licence o sha e adap ed ma e ial de i ed om his a icle o
pa s o i . The images o o he hi d pa y ma e ial in his a icle a e included in he a icle’s C ea i e Commons licence, unless indica ed
o he wise in a c edi line o he ma e ial. I ma e ial is no included in he a icle’s C ea i e Commons licence and you in ended use is no
pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain pe mission di ec ly om he copy igh holde . To
iew a copy o his licence, isi h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/.
Vi ology Jou nal
Bayesian phylogene ic and ecombina ion
analyses o plum pox i us p o ide a e ined
ision o i s e olu iona y his o y
F. Palmisano1,4†, S. Kawakubo2,5†, M. Chiumen i1, P. Leone i1, V. Pan aleo1*, T. Cand esse3 and A. Mina a1*
Abs ac
Backg ound The disco e y o a plum ee isola e o plum pox i us (PPV, Po y i us plumpoxi), done in Eas e n Albania
in 2011 in he ame o an EU- unded su ey, which ep esen s a di e gen s ain named PPV-An, p o ed o be o iginal
and in o ma i e o he un a eling o PPV e olu iona y his o y.
Me hods Maximum likelihood and Bayesian phylogene ic me hods applied on ull-leng h genomes o selec ed
egions analyzed he a ini ies o he PPV-An wi h o he PPV s ains. Po en ial ecombina ion e en s we e also e alu-
a ed. A e ined imeline o PPV e olu iona y his o y in eg a ing ecombina ion e en s, s ains mig a ion and ances al
hos s a e is p oposed.
Resul s Al oge he , he analyses con i m p e ious hypo heses ha PPV-An co esponds o an ances al, non- ecom-
binan PPV s ain. PPV-An likely se ed as he o igin o he PPV-M and T s ains h ough ecombina ion wi h isola e(s)
o he D s ain. Molecula clock analyses da ed he mos ecen common ances o (TMRCA) o PPV a 4546 yea s ago
and phylogeny sepa a ed he main PPV s ains om he che y-adap ed s ains a ound 3100 yea s ago. Meanwhile,
he ecombina ion e en s ha ga e ise o he M and T s ains a e es ima ed o ha e occu ed in he ea ly six een h
cen u y o he common e a (CE).
Conclusions The cha ac e iza ion o he PPV-An s ain enabled a comp ehensi e phylogene ic analysis o PPV. PPV-
An is con i med o be he p e iously uniden i ied p ogeni o , which, oge he wi h PPV-D led h ough ecombina ion
o he eme gence o he cu en ly p e alen and e olu iona y success ul ecombinan s ains o Eu opean o igin (e.g.,
M, Rec, and T). The low ep esen a ion o PPV-An in cu en PPV popula ions is likely he consequence o a popula ion
eplacemen phenomenon possibly linked o a highe i ness o he ecombinan s ains de i ing om i . These esul s
highligh he PPV-An s ain as a key playe in PPV e olu iona y his o y and consolida e PPV as one o he p omising
models o s udy hos -adap i e e olu ion p ocesses and phylogeog aphy among he mos damaging i uses o ag i-
cul u al sys ems.
Keywo ds Sha ka disease, Plum pox i us, S ain dynamics, Phylogeog aphic in e ence, Recombina ion, Ances o
†F. Palmisano and S. Kawakubo con ibu ed equally o his wo k.
*Co espondence:
V. Pan aleo
i an onio.pan aleo@cn .i
A. Mina a
angelan onio.mina a@cn .i
Full lis o au ho in o ma ion is a ailable a he end o he a icle
Page 2 o 16
Palmisanoe al. Vi ology Jou nal (2025) 22:319
Backg ound
Po y i us plumpoxi (plum pox i us, PPV) is a membe
o he genus Po y i us, amily Po y i idae. Like o he
po y i uses, PPV has a single-s anded posi i e-sense
genomic RNA o abou 10kb which encodes a single
la ge open eading ame (ORF). T ansla ion o his
ORF gene a es a polyp o ein p ecu so o ∼350 kDa
ha is in u n p ocessed, gi ing ise o p ocessing in e -
media es and o 10 inal p o ein p oduc s [1]. Po y i -
ids also encode a second, small ORF named PIPO, ia
a ibosomal ameshi ing wi hin he P3 cis on. PPV
is he agen o he Sha ka disease, he mos de as a ing
disease o s one ui ees wo ldwide [2–7]. Acco d-
ingly, PPV is conside ed as ei he a qua an ine pa ho-
gen o a egula ed non-qua an ine pa hogen in a wide
ange o coun ies [4, 5, 8].
PPV is ansmi ed by se e al species o aphids [9],
which acqui e he i us when p obing in ec ed plan s
and hen ans e i in a non-pe sis en manne o
heal hy plan s. Once a ee is in ec ed, i can exhibi
a ious symp oms including discolo ed lea a abesques
and ings, lea ugosi y (w inkling), and ui de o ma-
ions such as ingspo s o nec o ic spo s. Se e e ui
d ops may occu in he mos suscep ible a ie ies. In
addi ion, PPV is ansmi ed h ough all ege a i e
p opaga ion echniques, such as g a ing, making he
ade o P unus spp. p opaga ion ma e ial esponsible
o i s long- ange p opaga ion, including in e con inen-
al sp ead.
The genomes o a b oad ange o PPV isola es ha e been
comple ely sequenced [10–16] and PPV has been s udied
om bo h phylogene ic and e olu iona y pe spec i es.
Resea che s ha e used molecula echniques o analyze
he gene ic di e si y o PPV isola es collec ed om di -
e en geog aphic a eas and hos species. These s ud-
ies p o ided insigh s in o he e olu iona y ela ionships
among di e en s ains o he i us and helped o ace
he sp ead o PPV popula ions [17–21]. So a , 10 s ains
o PPV ha e been ecognized and named. The h ee majo
s ains, which show he b oade geog aphic dis ibu-
ion, a e PPV-D (Dide on; [22]), PPV-M (Ma cus, [22]
and PPV-Rec (Recombinan ). The o he iden i ied s ains
show mo e limi ed geog aphic dis ibu ions, includ-
ing he PPV-EA (El Ama ; [10, 23]) iden i ied in Egyp , a
ew o he s ains wi h a localized and spo adic p esence
such as PPV-W (Winona; [12]), PPV-T (Tu key; [13]),
and PPV-An (Ances o ; [24]). A leas 3 di e en s ains
a e able o na u ally in ec swee and sou che y PPV-C
(Che y, [11]), PPV-CR (Che y Russia, [25]), PPV-CV
(Che y Volga, [14]), while se e al phylog oups o che y-
adap ed isola es (namely SC, TAT and Y) ha may ep-
esen u he s ains ha e been ecen ly desc ibed in he
Sou h-Eas e n pa o Russia [15].
Unde s anding he phylogene ic ela ionships be ween
PPV s ains is c ucial no only o add essing he e o-
lu iona y his o y o PPV bu also o de eloping and
implemen ing e ec i e con ol s a egies, such as
deploying esis an cul i a s o designing diagnos ic
ools o ea ly s ain-speci ic de ec ion and sp ead p e-
en ion. Mo eo e , s udying he e olu iona y dynamics
o PPV can p o ide insigh s in o how he i us e ol es in
esponse o selec i e p essu es, such as hos esis ance
mechanisms o changes in ec o popula ions [26]. The
weal h o PPV gene ic in o ma ion a ailable om pub-
lic eposi o ies and associa ed me ada a makes PPV an
in e es ing model o phylogeog aphic and e olu iona y
s udies [20, 27, 28]. I is widely accep ed ha ecombina-
ion has been a majo d i ing o ce in PPV e olu ion. Fo
ins ance, he PPV-Rec s ain o igina ed om ecombi-
na ion be ween PPV-M and PPV-D, wi h a b eakpoin in
he NIb gene 3’ egion [29]; due o i s ypical PPV-M coa
p o ein, i was long misiden i ied as PPV-M [7]. Simi-
la ly, a Canadian PPV-W isola e was iden i ied as a com-
plex ecombinan in ol ing PPV-W, PPV-M, and PPV-D
[30]. Mo eo e , al hough he P1 p o ein is he leas con-
se ed among po y i al p o eins and is belie ed o be
in ol ed in po y i uses hos adap a ion [31], he 5’ pa
o he genome shows high homology among he PPV-
D, -M, -T, and -Rec s ains bu no wi h o he s ains,
sugges ing hese ou s ains a e linked by (an) ances-
al ecombina ion e en (s). The p ecise ecombina ion
his o y linking hese s ains is complex o un a el, and
wo en a i e ecombina ion b eakpoin s ha e been p o-
posed in he HC-P o and he P3 genes [17]. Two sce-
na ios o he e olu iona y his o y o PPV ha e been
p oposed, in which ei he PPV-D o PPV-M (and PPV-
T) would be (a) ecombinan s ain(s), while he o he
would ha e con ibu ed as one o he pa en s in ol ed in
he co esponding ecombina ion e en [17].
In bo h scena ios, he exis ence o an ances al, non-
ecombined o m o PPV-D o PPV-M was pos ula ed.
Pa ial genome sequencing o PPV isola es collec ed
du ing ield su eys pe o med in Albania in 2011 dem-
ons a ed he p esence o se e al PPV isola es [32](F.
Palmisano and S. Dallo , unpublished) and allowed he
iden i ica ion o an unusual isola e named AL-11pl, which
was cha ac e ized by a di e gen 5’genomic egion while
he es o i s genome is mo e ypical o al eady known
isola es o he PPV-M and -T s ains. These ea u es i
wi h he p ope ies hypo hesized by Glasa & Cand esse
[17] o a pu a i e ances al s ain which could ha e con-
ibu ed, in di e en ecombina ion e en s wi h PPV-D,
o he eme gence o he -M and -T s ains [7]. The disco -
e y and he cha ac e iza ion o he AL-11pl isola e hus
p o ide suppo o one o he wo al e na i e e olu ion-
a y scena ios p oposed by Glasa & Cand esse [17] and
Page 3 o 16
Palmisanoe al. Vi ology Jou nal (2025) 22:319
lead o he naming o he co esponding s ain as PPV-An
(Ances al o Ma cus s ain) [24]. Al hough an ini ial anal-
ysis ound ain e idence linking PPV-T o PPV-An [19],
a global phylogene ic analysis con i med a s ic associa-
ion be ween PPV-An and he PPV-T clade [18]. Fo he
i s ime, hese la e au ho s [18] p oposed a ecombi-
na ion e en s-based global PPV phylogeny on a da ase o
206 isola es, and es ima ed he age o he mos common
ecen ances o a abou 800yea s be o e common e a
h ough sub- ee compa ison me hod. Recen ad ances
in he s udy o he phylogeny and molecula e olu ion
o po y i uses [33] ha e in eg a ed Bayesian phylody-
namic analysis and da ing app oaches, as demons a ed
in esea ch on PVY [34] and TuMV [35, 36]. Genome
cha ac e iza ion and Bayesian phylogene ic in e ence o
geo e e enced isola es o hese i uses ha e acili a ed he
econs uc ion o hei sp ead and e olu iona y his o y
[37, 38].
In he p esen s udy, we used Maximum Likelihood
phylogeny and Bayesian app oaches o he analysis o
molecula e olu ion and iming o PPV s ains di e en-
ia ion [39–42]. Fo his, we used he comple e AL-11pl
sequence and hose o an addi ional 539 PPV isola es o
en a i ely da e he appea ance o he PPV-An-de i ed
lineage. F om hese elabo a ions, i was possible o p o-
pose a hypo he ical scena io abou he geog aphical o i-
gin and hos ances al s a e o PPV and o he u he
spa io- empo al sp ead o i s s ains.
Me hods
Reco e y ompublic da abases o PPV sequences
associa ed wi h hei me ada a and econs uc ion o hei
phylogeny
A o al o 609 ull leng h PPV genomes we e down-
loaded on 07 No embe 2023 om he NCBI Vi us da a-
base (h ps:// www. ncbi. nlm. nih. go / labs/ i us/ ssi/#/).
Sequences showing ameshi s o in e up ions o he
polyp o ein we e excluded om u he analysis. Me a-
da a (name o isola e, sampling yea , hos and coun y)
associa ed wi h he emaining sequences we e down-
loaded and manually cu a ed. A o al o 539 sequences
o which comple e me ada a we e a ailable, we e kep
o subsequen analyses (Addi ional Table1). A mul i-
ple alignmen o hese ull-leng h genomic sequences
was ob ained using MAFFT [43]. The aligned comple e
genomic sequences we e u he manually checked in
Geneious .2024.0.7 o iden i y he ends o he la ge ORF
encoding he polyp o ein and he co ec genome e -
mini. Pai wise iden i y pe cen ages we e calcula ed om
he aligned posi ions.
In addi ion o he phylogene ic analysis un on he ull-
leng h genome alignmen , h ee dis inc genomic egions
we e selec ed, a oiding he b eakpoin s o known ecom-
bina ion e en s al eady desc ibed and con i med he e
(see below and Fig.2D) [7, 17]: (i) a 5’ e minal egion
(n 36–1400, he numbe ing used h oughou ollows he
consensus alignmen using all 539 isola es); (ii) an in e -
nal 5’ agmen (n 1500–2600); (iii) a 4 kilobases cen-
al egion (n 3500–7500). The bes i ing subs i u ion
models o he ull-leng h genome and a ious pa ial
sequences alignmen s we e in es iga ed using MEGA X
[44] and ModelFinde implemen ed in IQ- ee ( e sion
2.3.6 [45];). The subs i u ion model sugges ed by his la -
e p og am was inally applied o he es . Maximum
likelihood (ML) phylogenies we e hen in e ed using he
IQ-T ee so wa e. Boo s ap alues we e calcula ed using
1,000 eplica es. T ees we e inally isualized in he In e -
ac i e T ee o Li e (iTOL) [46].
Recombina ion analysis o genomic sequences
To econs uc he e olu iona y his o y o PPV, and o
unde s and he possible o igin o he genomic agmen s
and he ole played by ecombina ion in such e olu ion
among he s ains, a sea ch o po en ial ecombina ion
b eakpoin s and he iden i ica ion o he pu a i e pa en s
was pe o med using he RDP4 p og am ( e sion 4.10.1,
de aul p og am se ings) [47] and he abo e desc ibed
ull-leng h genomes mul iple sequence alignmen . Only
Table 1 Pe cen ages o iden i y wi h PPV-An (isola e AL-11pl) in di e en genome egions. Values a e de i ed om pai wise iden i y
ma ices ob ained om a MAFFT mul iple alignmen o ull-leng h genomes o 539 PPV isola es. Me ada a (yea , hos and coun y o
isola ion) o each epo ed isola e a e p o ided
Region (nucleo ide
posi ion) MAFFT alignmen Iden i y (%) S ain Me ada a o s ains/sequence
1–1584 max 81.7 T ON745776|P13_ANK|Tu key|P unus_ce asus|2017
min 73.2 CV MF447179|Ta _2|Russia|P unus_ce asus|2015
1585- 2758 max 96.3 T MF346246|AnKuAp8|Tu key|P unus_a meniaca|2014
min 76.8 TAT OK562686|TAT_85|Russia|P unus_ce asus|2018
2759–7532 max 96.4 M LC494682|Y3|Japan|P unus_mume|2016
min 77.6 W HQ670746|LV_141pl|La ia|P unus_domes ica|2010
Page 4 o 16
Palmisanoe al. Vi ology Jou nal (2025) 22:319
e en s de ec ed by a leas 4 me hods and wi h co ec ed
p- alues < 10–6 we e conside ed as eliable. The iden i-
ied b eakpoin s we e u he manually examined, and
BLASTN sea ches we e used o e i y he pa en /dono
s ain assignmen and hei espec i e sequence homol-
ogy le els.
Ances al s a es econs uc ion h oughBayesian
phylogeny
The exis ence o a empo al signal o PPV e olu ion
in he 539 ull-leng h genomic sequences da ase was
assessed using he p og ams TempEs ( e sion 1.5.3; [48])
and T eeTime ( e sion 0.11.4; [49]).
Fo Bayesian phylogene ic analysis, wo da ase s we e
used. The i s co esponds o he cen al genomic egion
(n 3500–7500) aligned o he 539 isola es ep esen ing
all s ains in ou o iginal da ase . This egion was shown
o be ee o ecombina ion e en s o he isola es used
and will be he ea e e e ed o as he 4K egion. The
second one co esponds o he 5’ genome agmen (n
36–1406) om he same PPV ull-leng h genome align-
men . The package bModelTes [50] was used o assess
he si e model and associa ed subs i u ion model. Ma -
ginal likelihood es ima ion (MLE) h ough Pa h sam-
pling/S epping-s one sampling (PS/SS) analysis [51] was
un on BEAST o compa e he i o he elaxed clock
model wi h a s ic clock model in he abo e-men ioned
sequence da ase s. Based on he a ailable me ada a (see
pa ag aph abo e), o each PPV sequence a ailable dis-
c e e loca ion (coun y o o igin) and hos o o igin
s a es we e assigned. A symme ic subs i u ion model
was applied o each disc e e ai (hos and coun y)
and social ne wo ks we e in e ed by Bayesian s ochas-
ic sea ch a iable selec ion (BSSVS). Ances al s a es
we e econs uc ed o all he conside ed pa i ions. The
Ma ko Chain Mon e Ca lo (MCMC) me hod in he
BEAST package ( 10.5.0; [52]), along wi h BEAGLE, was
in e ed wi h he bes i subs i u ion model (GTR + G + I)
sugges ed by bModelTes o he wo independen da a-
se s desc ibed abo e. In he case o he 4K egion, 13
uns o a o al o 1300 million MCMC s eps we e ca ied
ou , while in he case o he 5’ egion 3 uns including 800
million s eps, we e me ged using LogCombine ( 10.5.0).
The T ace so wa e ( 1.7.2) was used o con i m ha
all es ima ed pa ame e s yielded e ec i e sampling sizes
(ESS) g ea e han 200 and ha 10% o he o al chain
leng h had been bu ned-in o educe he in luence o he
ini ial alue. The inal Bayesian maximum clade c edibil-
i y (MCC) ee was gene a ed by T eeAnno a o ( 10.5.0)
and isualized in Fig ee ( 1.4.4). A g aphical elabo a ion
o he inal MCC ees was ob ained in RS udio using he
packages gg ee and eeio [53, 54]. The esul ing Bayes-
ian di usion s a es in space and ime we e calcula ed
wi h he SPREAD applica ion ( e sion 1.0.7; [55]). The
ances al ai o he hos was econs uc ed a he oo
and he oldes nodes o he same MCC ees.
To u he con i m he molecula clock signals ob ained
h ough he p e iously desc ibed BEAST analysis, a syn-
ch onous BEAST elabo a ion o he wo da ase s (sim-
ula ing he sampling da es as done all a he same ime,
i.e. 2017.5) was un wi h he same pa ame e s desc ibed
abo e. Then, da e- andomiza ion es was pe o med
wi h 10 eplica es whose sampling da e we e andomly
pe mu a ed in he TipDa ing Beas package. We assumed
he p esence o empo al s uc u e when he 95% c ed-
ible in e al o he a e es ima e om he o iginal da ase
did no o e lap wi h he 95% c edible in e al o any o
he a e es ima es om he da e- andomized eplica es.
Finally, o unde s and i he p e alence o PPV-D
sequences p esen in he da ase s could ha e biased he
Bayesian calcula ion o subs i u ion a es and TRMCAs,
an addi ional analysis was pe o med wi h he e y same
pa ame e s desc ibed abo e o he 5’and 4k egions,
selec ing only PPV-D non ecombinan and PPV-Rec
s ains, o a o al o 291 sequences.
Resul s
Rele an molecula andse ological ea u es o PPV‑An
The AL-11pl isola e (GenBank HF674399), which ypi ies
he PPV-An s ain and will be he ea e indica ed as An,
was iden i ied in a domes ic plum ee in Eas e n Alba-
nia. Se ologically, PPV-An eac ed o he PPV-M-speci ic
monoclonal an ibody (MAb) AL (no shown). Rema k-
ably, i also es ed posi i e wi h he PPV-D-speci ic MAb
4DG5, an unusual beha iou p e iously epo ed o
some PPV-T isola es [56]. The comple e genome o he
PPV-An isola e is 9,786 nucleo ides long, excluding he
3’ e minal polyA ail, and has a GC con en o 43.8%.
The genomic o ganiza ion is ypical o membe s o genus
Po y i us, and iden ical o ha o o he PPV isola es. A
s a codon (AUG) is p esen a posi ions 147–149, and
an ambe s op codon a posi ions 9567–9569, esul ing
in a single open eading ame (ORF) o 9420 n /3140
amino acids. In addi ion, he PIPO ORF [57] pu a i ely
encoding a 12kDa p o ein was iden i ied in he P3 coding
egion as a + 2 ameshi sequence s a ing a nucleo ide
posi ion 2906.
Sequence alignmen s show a conse a ion o he nine
polyp o ein clea age si es as compa ed o PPV-M iso-
la es, wi h he excep ion o mu a ions obse ed in he
NIa-VPg/NIa-P o clea age si e (EEVGHE/S in PPV-An
and DEVDHE/S in PPV-M isola es) and NIa-P o/NIb si e
(EFVHNQ/S s. EFVYNQ/S). All conse ed mo i s ypi-
cal o po y i uses we e also iden i ied a hei expec ed
loca ions, including he KITC [58], PTK [59] and DAG
mo i s associa ed wi h aphid ansmission.
Page 5 o 16
Palmisanoe al. Vi ology Jou nal (2025) 22:319
Pe cen ages o nucleo ide pai wise iden i y calcula ed
be ween PPV-An and all o he PPV s ains – exp essed
as hei maximal and minimal alues in a ious genomic
egions selec ed om he ull-leng h genome alignmen
a e gi en in Table 1. The comple e genome sequence
compa isons indica e ha he closes s ain o PPV-An is
PPV-T, wi h an o e all nucleo ide iden i y o 93.5%. Ne -
e heless, while PPV-An is closely ela ed o isola es o
he T and M s ains a he whole genome le el, i shows
a much lowe iden i y le el (74–77%) wi h hese s ains
o he 5’ non-coding egion (5’NCR) and he P1 gene
(81.7%, om s a codon up o n 1584). This dissimila -
i y pa e n ex ends o he HC-P o gene (n 1585–2758)
in he case o he M s ain (82% n sequence iden i y)
bu i is no longe obse ed in he case o he T s ain
(96.3% iden i y; Table1). Nucleo ide iden i y le els wi h
bo h T and M s ains o all he 3’ downs eam pa s o
he genome a e highe han 95%. This unusual iden i y
pa e n obse ed o he a ious egions o he PPV-An,
s ongly poin s o i s possible implica ion in ecombina-
ion e en s.
Analysis o ecombina ion e en s in ol ing PPV‑An
aspa en al ances o
The mul iple alignmen o 539 ull-leng h genomic
sequences o PPV isola es was analysed o ecombina-
ion e en s using RDP4. A numbe o pu a i e ecombi-
na ion e en s we e iden i ied by he p og am, bu mos
o hem in ol ed jus one o a ew isola es and we e
de ec ed by only h ee o less o he me hods, wi h s a-
is ically non-signi ican p- alues. Howe e , wo ecom-
bina ion e en s which in ol e s ains sha ing a la ge
genomic po ion wi h PPV-An we e iden i ied by a s ong
signa u e and a e discussed in de ail he e (Addi ional
Table 2). A i s s a is ically signi ican ecombina ion
e en (i.e. lowes p- alue 1.77 × 10–73) iden i ies he ag-
men a n posi ions 1561–2735 in he alignmen (wi h
99% con idence in e als [CI 99%] o he b eakpoin s a
903–1598 and 2678–2765) and a ec s 130 PPV-M iso-
la es (Fig.1A; e en #4, Addi ional Table2). This e en
in ol es a pu a i e majo pa en belonging o he PPV-T
s ain (GenBank MF346274), and a mino pa en belong-
ing o he PPV-D s ain (GenBank KR006730). Gi en he
RDP4 ou pu , s ain T isola es a e in e p e ed by he p o-
g am as non- ecombinan majo dono s o a backbone
which ecei ed he inse ion o he he e ologous ag-
men om a s ain D mino pa en , hus leading o he
cu en PPV-M isola es.
The o he ele an ecombina ion e en iden i ied, wi h
a b eakpoin a n 2691 o he PPV-An sequence (CI 99%:
2665–2816), in ol es a PPV-M isola e as he majo pa -
en p o iding he en i e 3’ genome pa ( om n 2691
up o 3’ end) and an unknown mino pa en p o iding
he 5’genome po ion (Fig. 1A; e en #3, Addi ional
Table2). This ecombina ion e en is consis en ly sup-
po ed by six me hods (i.e. RDP, GENECONV, Boo Scan,
MaxChi, Chimae a, SiScan) wi h he lowes p- alue o
2.23 × 10–115. The b eakpoin o his p edic ed ecom-
bina ion e en is in e y close p oximi y o he one (a
posi ion 2814) hypo hesized by Glasa & Cand esse [17]
o a ecombina ion e en be ween PPV-D and a pu a i e
ances al isola e leading o he eme gence o he PPV-M
s ain. Howe e , when he 3’ end genomic egion o PPV-
An (n 3500–9700) was used as a que y o a BLASTN
in e oga ion o GenBank, he closes isola e o PPV-An
in ha genomic egion was PPV-T K PnPl345 (Gen-
Bank MF346272), sha ing 96.4% o n iden i y. Fo he
5’ e minal egion (1–1561), an unknown mino pa en
was pos ula ed in he RDP4 ou pu . A BLASTN analy-
sis indica ed ha he isola e mos simila o PPV-An in
ha egion is P93 ANK (GenBank ON745778), ano he
PPV-T isola e, bu i s iden i y wi h PPV-An is only o
85,3%.
The #3 and #4 (Addi ional Table2) e en s iden i ied
he e a e espec i ely simila o e en s X2 and X3 iden-
i ied by [18], al hough a easoning based on phyloge-
ne ic analysis led hem o conclude ha he T s ain was
de i ed om a PPV-M s ain pa en , and no he e e se
as p edic ed by ou RDP analysis. Howe e , he po en-
ial scena io desc ibed by hese au ho s has wo weak-
nesses. Fi s , conside ing ei he PPV-M o PPV-T as a
non- ecombined s ain pa en o PPV-An ails o p o ide
an explana ion o he low di e gence in he P1 p o ein
be ween hese wo s ains and PPV-D s ain, despi e he
ac ha P1 is e y gene ally he mos di e gen po y i al
p o ein. Second, his scena io is no pa simonious since
i needs o pos ula e he exis ence o a u he unknown
mino pa en p o iding he 5’ e minal po ion o PPV-An
in he ecombina ion e en #3. In con as , he scena io
p oposed by [17] and u he discussed by [7], which
iden i ies PPV-An and PPV-D as ances al non ecom-
binan pa en s and PPV-M and PPV-T as he de i ed
ecombinan p ogeny, is mo e pa simonious (i does no
pos ula e he exis ence o ano he unknown pa en ) and
eadily p o ides an explana ion o he high P1 homology
obse ed be ween PPV-D and M and T s ains (Fig.1B,
Hypo hesis 2).
Maximum likelihood phylogene ic analyses on ull leng h
sequences andpa ial genomic agmen s
The maximum likelihood (ML) phylogene ic analysis
pe o med on he ull-leng h genome da ase showed a
s iking posi ion o PPV-An as a long b anch linked o
he PPV-T clade (Addi ional Fig.1). This analysis shows
e y clea ly a sepa a ion be ween he clades o isola es
belonging o he Eu opean and Medi e anean s ains
Page 6 o 16
Palmisanoe al. Vi ology Jou nal (2025) 22:319
(PPV-M, -T, -An, -D and -EA) om hose co esponding
o PPV-W and he che y-adap ed s ains. As an al e na-
i e o he exclusion o he la ge numbe o ecombinan
isola es (PPV-M, T and Rec s ains) om he da ase
o pe o m an e alua ion o he phylogene ic signals
( he e o e losing use ul in o ma ion), we decided o use
only genome po ions known o be ee o ecombina-
ion b eakpoin s. The bes subs i u ion model ob ained
wi h ModelFinde o he ull-leng h genomes and he
h ee pa ial genomic da ase s was GTR + F + I + G4, and
i was ho oughly applied o he u he phylogene ic
calcula ions.
A phylogene ic analysis was pe o med using a
5’genomic agmen (nucleo ide posi ions 36 o 1400),
which ends be o e he bo de o he i s ecombina-
ion b eakpoin sugges ed by [17] and essen ially o e -
laps wi h he 5’bo de o he ecombina ion e en #4
discussed abo e. The esul ing ML cladog am showed
A
B
PPV-D
PPV-T
PPV-M
VV VVV
PPV-Unknow
PPV-M
PPV-An
VV VVV
P1 HC-P o P3 CI
6K16K2
VPg P o NIb CP
(A) n
VPg
Hypo esis 1, PPV-An is a ecombinan
C
PPV-An
PPV-D
PPV-T
VV VVV
PPV-An
PPV-D
PPV-M
VV VVV
Hypo esis 2, PPV-An is NOT a ecombinan
RDP e en 3
RDP e en 4
O igin o PPV-T
O igin o PPV-M
Fig. 1 Schema ic ep esen a ion o he RDP4 ou pu o some ecombina ion e en s and hypo hesis o ecombina ion in ol ing PPV-An s ain.
A Schema ic ep esen a ion o PPV genomic o ganiza ion, whe e he open box ep esen s he ansla ed ORFs and he unc ional polyp o ein
agmen s a e named. B In he hypo hesis 1 ame, he ep esen a ion only pic u es wo independen ecombina ion e en s as iden i ied
by RDP4 in e ms o pa en al PPV s ains and ecombined egions and is no ela ed o any phylogene ic assump ion. He e, PPV-An is conside ed
as ecombinan . E en #3) depic s he PPV-An isola e, AL-11pl, as o igina ed by ecombina ion be ween an unknown mino pa en p o iding
he 5’-end agmen (n posi ions 1–2691) and a majo pa en belonging o M s ain and p o iding he es o he genome. E en #4) depic s
he o igin o he M s ain isola es h ough he inse ion o a agmen (n posi ions 1561–2735) om a D isola e in he backbone o a T isola e
(see Addi ional Table 2). C In he hypo hesis 2 ame, he scena io in which An is conside ed non- ecombinan and a pa en o M and T s ains,
when ecombining wi h PPV-D, is illus a ed. The di e en colo codes o he s ains de i e om he di e en oles as pa en o ecombinan hey
co e in he di e en e en s
Page 7 o 16
Palmisanoe al. Vi ology Jou nal (2025) 22:319
a igh clus e ing o all PPV-D, PPV-M, and PPV-T
isola es (61.2% boo s ap), which we e sepa a ed om
a small clade con aining he PPV-EA and PPV-An
isola es (59.2% boo s ap) (Fig. 2A). In he ML ee
econs uc ed using he egion be ween he i s and
second iden i ied ecombina ion b eakpoin s (n posi-
ions 1500–2600, acco ding o [17] and which spans
he C- e minal hal o HC-P o and he beginning o he
P3 gene (Fig.2B), PPV-An now o ms a loose clus e
(41.4% boo s ap) wi h he PPV-T isola es.
Las ly when a la ge, non- ecombined in e nal genome
agmen (n 3500–7500, e e ed as he 4K egion)
was used o ML phylogeny, PPV is spli in o i e phylo-
gene ic g oups (Fig.2C) wi h PPV-An clea ly lanking,
as a single long b anch, he PPV-M and PPV-T clade
(65.4% boo s ap).
In his ee, a single isola e (SK-111pl, belonging o
M s ain, GenBank HF585099) has an anomalous posi-
ion as basal ip a he node which o igina ed all he An-
de i ed s ains. This posi ion could mos likely be due o
he peculia ecombina ion his o y o his isola e, which
conce ns, acco ding he RDP4 analysis, a ecombina ion
e en in ol ing a small egion ( om n 3264 o 4010;
Addi ional Table2, e en #5) a he beginning o he 4K
egion. O e all, he incong uen opologies in he di e -
en ees when i comes o he PPV-D, Rec, M, T isola es
e sus he An one, suppo he ecombina ion analysis
and con i ms ha ecombina ion e en s ha e con ib-
u ed in a majo way o PPV e olu iona y his o y o he
Eu o-Medi e anean (Eu o-Med) PPV s ains. Figu e2D
esumes he cu en iew o ecombina ion e en s along
he PPV genome which cha ac e ize he single s ains.
0.1
Region 36-1400
0.1
0.1
A
Region 1500-2600
B
Region 3500-7500
C
WChe y
EA
An
D, M, T, Rec
W
Che y
EA
D, M, Rec
T
D, Rec
M, T
W
Che
y
An
An
D
P1 HC-P o P3 CI
6K1 6K2
VPg P o NIb CP
(A)
n
VPg
PPV-Che y
PPV-W
PPV-EA
PPV-M
PPV-An
PPV-T
PPV-Rec
PPV-D
Fig. 2 Maximum likelihood phylogene ic ees econs uc ed using di e en genome egions o 539 PPV isola es. S ains o phylog oups a e
indica ed as PPV-T (T), PPV-An (An), PPV-M (M), PPV-EA (EA), PPV-W (W), PPV-C (C), PPV-Y (Y), PPV-CV (CV), PPV-SC (SC), PPV (TAT), PPV-CR (CR),
PPV-D (D) and PPV-Rec (Rec). Colo ed spo s (when p esen ) co espond o he P unus species om which any isola e was ob ained (see he legend
o Addi ional Fig. 1). All he s ains belonging o he same clade and sha ing simila sequence iden i y in he genomic po ion unde e alua ion,
a e g ouped by he same colo backg ound. The dis ance ba is also shown unde each ee. A T ee o he 5’ egion (n posi ion 36–1400). B T ee
o an in e nal agmen co esponding o n posi ions 1500–2600. C T ee o he 4 K genomic agmen (n posi ion 3500–7500). D Colo -coded
schema ic ep esen a ion o he genome o he main PPV s ains. Di e en colo s ep esen he en a i e s ain o igin o he pu a i e ecombined
agmen s. The genome o ganiza ion o PPV is shown a he op, whe e he open box ep esen s he ansla ed ORFs and he unc ional polyp o ein
agmen s a e named. The boo s ap alues a he main nodes a e no epo ed in he pic u es o imp o e isualiza ion, since mos o hem a e >
= 60
Page 8 o 16
Palmisanoe al. Vi ology Jou nal (2025) 22:319
PPV‑An‑de i ed s ains di e si ica ion andBayesian da ing
in heglobal PPV e olu ion andsp ead
The p esence o a empo al signal was i s in es iga ed
on he global PPV phylogeny o de e mine whe he ana-
lyzing he e olu ion o exis ing lineages could p o ide
a eliable es ima e o he ime o a common ances o
be ween An and he o he PPV lineages. The TempEs
so wa e [48] un on he ull-leng h PPV genomes, eco -
e ed a TMRCA in e cep a 3604yea s ago, while he
oo - o- ip eg ession using T eeTime pu he TMRCA
in e ed om he same sequence alignmen a 4499yea s
ago. The quan i a i e me ics ob ained wi h TempTes a e
epo ed in Addi ional Table3.
This p elimina y da ing analysis o a e ie al o a em-
po al signal could incu a bias because o he inclusion o
ecombinan sequences in he da ase . In ac , acco ding
o ou e idence o PPV ecombina ion his o y (Fig.2D)
only PPV-D, PPV-An, PPV-EA, mos o PPV-W and
he che y-adap ed s ains a e conside ed non- ecom-
binan s, so ha a la ge p opo ion o he isola es included
in he da ase a e ecombinan s. To o e come his p ob-
lem, Bayesian da a ion and phylogeog aphic di usion
analysis we e sepa a ely pe o med on wo non- ecom-
binan genome egions desc ibed abo e: he 4K cen al
egion and he sho e 5’ agmen . Fi s , o e alua e he
empo al s uc u e in he da ase , he da e- andomiza-
ion es was pe o med. The 95% c edible in e als o
in e ed subs i u ion a e o bo h da ase s did no o e -
lap wi h any o he da e- andomized eplica es, indica ing
he p esence o empo al signal in ou da ase (Addi-
ional Fig.3, A and B). Mo eo e , he p esence (s ic
clock model) and absence (unco ela ed elaxed clock
model) o a molecula clock signal we e also es ed using
Bayesian s a is ics. MLE calcula ed o he wo hypo h-
esized models h ough PS/SS analysis yielded Bayes ac-
o (BF) alues indica ing a decisi e suppo in a o o
he p esence o a molecula clock o bo h da ase s (i.e.,
4K BF = 6405.9; 5’ egion BF = 6244.3). The bModelTes
sea ch o he bes subs i u ion model o be applied in he
Bayesian ees calcula ion sugges ed he GTR + G + I one.
The a e age a e o nucleo ide subs i u ion es ima ed
o he cen al 4K egion o all a ailable PPV isola es, cal-
cula ed a e BEAST elabo a ion, was 6.91 × 10–5 subs ./
si e/yea (wi h he 95% highes pos e io densi y (HPD)
in e al o 4.67 × 10–5—9.09 × 10–5). In he Bayesian MCC
ee calcula ed o he 4K egion (Fig.3 and Addi ional
Table4 A) he posi ion o he TMRCA was de e mined a
4546yea s ago (95% HPD 3141–6099) when he g oup o
Eu o-Med s ains sepa a ed om he b anch leading o
PPV-W and he che y-adap ed s ains. The sepa a ion o
PPV-EA om he Eu o-Med g oup could ha e happened
3655yea s ago (95% HPD 2531–4944) while o i s pa ,
he sepa a ion o he clade o che y-adap ed s ains om
he W s ain is da ed back a 3087yea s ago (95% HPD
2111–4136). The ini ial di e si ica ion o he Eu o-Med
g oup is p edic ed o ha e s a ed 1733yea s ago (95%
HPD 1200–2350). The sepa a ion o he M and T s ains
om PPV-An is sugges ed o ha e occu ed much mo e
ecen ly, be ween 289 and 279 yea s ago, espec i ely
(95% HPD 200–394 and 185–400). Using his da ase ,
he en a i e da ing o he sepa a ion o he D and Rec
s ains a 251yea s ago (95% HPD 178–338) could also
be hypo hesized.
Using he 5’ agmen encoding he P1 gene ( ag-
men 36–1406 n ), he BEAST elabo a ion e ie ed a
subs ./s/y a e o 2.55 × 10–4 o he global alignmen o
539 isola es (95% HPD 2.025 × 10–4—3.084 × 10–4), which
is abou 3.6 imes as e han ha ob ained o he 4K
egions and is likely accoun ed o he lowe conse a ion
o he P1. In Fig.4 (and Addi ional Table4B), he oo o
he ee da ed a 1211yea s ago (95% HPD 962–1482)
and he main di e si ica ion o he che y-adap ed s ains
a 707yea s ago (95% HPD 545–872).
Rega dless he di e ence in subs i u ion a es, he
Bayesian analysis on bo h da ase s indica ed ha he
sepa a ion o he PPV-An s ain is an e io o ha o he
D s ain, and hus o all he subsequen node bi u ca ion
e en s in he Eu o-Med g oup.
Fu he mo e, he Bayesian elabo a ions done on bo h
he same egions, bu only on a da ase es ic ed o
D isola es (including Rec, o a o al o 291 sequences),
ob ained e y simila esul s, as subs i u ion a es (5’
egion: 2.029E-4; 4 K egion: 8.688E-5) and TMRCA
da a ion (5’ egion: 1409.404yea s be o e p esen ; 4K
egion: 3541.299 yea s be o e p esen ), hus indica -
ing ha he la ge p edominance o he D isola es in he
da ase could subs an ially ha e in luenced he comple e
da ase ou pu alues.
Pa allel o hese da ing in e ences, he wo da ase s
we e used in synch onous BEAST elabo a ions assuming
(See igu e on nex page.)
Fig. 3 Time-scaled maximum-clade-c edibili y ee in e ed om he 4 K genome egion o 539 selec ed PPV isola es. Isola es in he ee a e
collapsed by s ain and b anch leng hs a e scaled by dis ance om TMRCA acco ding o ime (yea s om p esen ), as shown by he e e ence
ba a bo om. A he majo nodes, a pie cha shows he hos se p obabili y wi h P unus hos species colo ed acco ding o he legend inse . The
pink ba a each node ep esen s he 95% con idence in e al (HPD 95%). Speci ic alues o each node da e and pe cen age o hos se p e e ence
a e epo ed in Addi ional Table 4A
Page 9 o 16
Palmisanoe al. Vi ology Jou nal (2025) 22:319
Che y
T
EA
M
Rec
D
W
An
P. ce asus
P. domes ica
P. dulcis
P. u icosa
P. japonica
P. mume
P. pe sica
P. pe sica a . nuci e a
P. a meniaca
P. ce asi e a
P. omen osa
P unus
Fig. 3 (See legend on p e ious page.)
Page 16 o 16
Palmisanoe al. Vi ology Jou nal (2025) 22:319
43. Ka oh K, S andley DM. MAFFT mul iple sequence alignmen so wa e
e sion 7: imp o emen s in pe o mance and usabili y. Mol Biol E ol.
2013;30:772–80.
44. Kuma S, S eche G, Li M, Knyaz C, Tamu a K. MEGA X: molecula e o-
lu iona y gene ics analysis ac oss compu ing pla o ms. Mol Biol E ol.
2018;35:1547–9.
45. Minh BQ, Schmid HA, Che nomo O, Sch emp D, Woodhams MD, Von
Haesele A, e al. IQ- ee 2: new models and e icien me hods o phylo-
gene ic in e ence in he genomic e a. Mol Biol E ol. 2020;37:1530–4.
46. Le unic I, Bo k P. In e ac i e ee o li e (iTOL) 5: an online ool o phylo-
gene ic ee display and anno a ion. Nucleic Acids Res. 2021;49:W293–6.
47. Ma in DP, Mu ell B, Golden M, Khoosal A, Muhi e B. RDP4: De ec ion
and analysis o ecombina ion pa e ns in i us genomes. Vi us E ol.
2015;1: e 003.
48. Rambau A, Lam TT, Max Ca alho L, Pybus OG. Explo ing he empo al
s uc u e o he e och onous sequences using TempEs ( o me ly Pa h-O-
Gen). Vi us E ol. 2016;2: ew007.
49. Sagulenko P, Pulle V, Nehe RA. T eeTime: Maximum-likelihood phylody-
namic analysis. Vi us E ol. 2018;4: ex042.
50. Bouckae RR, D ummond AJ. bModelTes : Bayesian phylogene ic si e
model a e aging and model compa ison. BMC E ol Biol. 2017;17:42.
51. Baele G, Lemey P, Bed o d T, Rambau A, Sucha d MA, Alekseyenko AV.
Imp o ing he accu acy o demog aphic and molecula clock model
compa ison while accommoda ing phylogene ic unce ain y. Mol Biol
E ol. 2012;29:2157–67.
52. Sucha d MA, Lemey P, Baele G, Ay es DL, D ummond AJ, Rambau A.
Bayesian phylogene ic and phylodynamic da a in eg a ion using BEAST
1.10. Vi us E ol. 2018;4: ey016.
53. Wang L-G, Lam TT-Y, Xu S, Dai Z, Zhou L, Feng T, e al. T eeio: an R package
o phylogene ic ee inpu and ou pu wi h ichly anno a ed and associ-
a ed da a. Mol Biol E ol. 2020;37:599–603.
54. Yu G, Smi h DK, Zhu H, Guan Y, Lam TT. gg ee : an package o isualiza-
ion and anno a ion o phylogene ic ees wi h hei co a ia es and o he
associa ed da a. McIne ny G, (ed). Me hods Ecol E ol. 2017;8:28–36.
55. Bielejec F, Rambau A, Sucha d MA, Lemey P. SPREAD: spa ial phy-
logene ic econs uc ion o e olu iona y dynamics. Bioin o ma ics.
2011;27:2910–2.
56. Cand esse T, Camb a M, Dallo S, Lanneau M, Asensio M, Go is MT, e al.
Compa ison o monoclonal an ibodies and polyme ase chain eac ion
assays o he yping o isola es belonging o he D and M se o ypes o
plum pox po y i us. Phy opa hology®. 1998;88:198–204.
57. Chung BY-W, Mille WA, A kins JF, Fi h AE. An o e lapping essen ial gene
in he Po y i idae. P oc Na l Acad Sci USA. 2008;105:5897–902.
58. U cuqui-Inchima S, Haenni A-L, Be na di F. Po y i us p o eins: a weal h o
unc ions. Vi us Res. 2001;74:157–75.
59. Wang Y, Gal-On A, Hue H, Kadou y D, Raccah B, Peng YH. Mu a ions in
he HC-P o gene o zucchini yellow mosaic po y i us: e ec s on aphid
ansmission and binding o pu i ied i ions. J Gen Vi ol. 1998;79:897–904.
60. Gü can K, Tebe S, Akbulu M, Çağlayan K. Gene ic di e si y and a long
e olu iona y his o y o plum pox i us s ain ec in Tu key. Eu J Plan
Pa hol. 2021;161:453–61.
61. Camb a M, Asensio M, Go is MT, Pé ez E, Cama asa E, Ga ciá JA, e al.
De ec ion o plum pox po y i us using monoclonal an ibodies o s uc-
u al and non-s uc u al p o eins1. EPPO Bull. 1994;24:569–77.
62. Boscia D, Ze amdini H, Camb a M, Po e e O, Go is MT, My a A, e al.
P oduc ion and cha ac e iza ion o a monoclonal an ibody speci ic o he
M se o ype o plum pox po y i us. Eu J Plan Pa hol. 1997;103:477–80.
63. Shan H, Pasin F, Valli A, Cas illo C, Rajulu C, Ca bonell A, e al. The Po y i i-
dae P1a leade p o ease con ibu es o hos ange speci ici y. Vi ology.
2015;476:264–70.
64. Nigam D, LaTou e e K, Souza PF, Ga cia-Ruiz H. Genome-wide a ia ion
in po y i uses. F on Plan Sci. 2019;10:1439.
65. Sanjuán R, Agudelo-Rome o P, Elena SF. Uppe -limi mu a ion a e es ima-
ion o a plan RNA i us. Biol Le . 2009;5:394–6.
66. Jenkins GM, Rambau A, Pybus OG, Holmes EC. Ra es o molecula
e olu ion in RNA i uses: a quan i a i e phylogene ic analysis. J Mol E ol.
2002;54:156–65.
67. Gibbs AJ, Ohshima K, Phillips MJ, Gibbs MJ. The p ehis o y o po y i uses:
hei ini ial adia ion was du ing he dawn o ag icul u e. PLoS ONE.
2008;3:e2523.
68. Mohammadi M, Hosseini A, Nas ollanejad S. In silico iden i ica ion o wo
no el i uses on I anian pis achio. I an J Plan Pa hol. 2021;57:81–5.
69. Glasa M, Palko ics L, Komínek P, Labonne G, Pi ne o á S, Kúdela O, e al.
Geog aphically and empo ally dis an na u al ecombinan isola es o
Plum pox i us (PPV) a e gene ically e y simila and o m a unique PPV
subg oup. J Gen Vi ol. 2004;85:2671–81.
70. She ele a A, Glasa M, Kud ya se a A, I ano P, Chi ko S. Gene ic di e -
si y, hos ange and ansmissibili y o CR isola es o Plum pox i us. J Gen
Plan Pa hol. 2019;85:39–43.
71. Dallo S, Ka yche R, Dolgikh S, Thébaud G, Jacquo E, Dec oocq V. Fi s
epo o Plum pox i us s ain W in Kazakhs an, on P unus domes ica.
Plan Dis. 2019;103:2702.
72. A anaso D. Mosaic o s one ui s. 1935 . A ailable om: h ps:// www.
cabid igi a llib a y. o g/ doi/ ull/ 10. 5555/ 19350 501302. Ci ed 2025 Feb 27.
73. Mou y B, Desbiez C. Hos ange e olu ion o po y i uses: a global phylo-
gene ic analysis. Vi uses. 2020;12:111.
74. Tsa mpopoulos I, Ma ais A, Fau e C, Theil S, Cand esse T. A new po y i us
om hedge mus a d (Sisymb ium o icinale (L.) Scop.) sheds ligh on he
e olu iona y his o y o u nip mosaic i us. A ch Vi ol. 2023;168: 14.
Publishe ’s No e
Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in pub-
lished maps and ins i u ional a ilia ions.
