RESEARCH Open Access
Assessmen o gene ic di e si y in No dic
imo hy (Phleum p a ense L.)
Pi jo Tanhuanpää
1*
, Ma ia E kkilä
2
, Ruslan Kalenda
2
, Alan Howa d Schulman
1,3
and Ou i Manninen
4
Abs ac
Backg ound: Timo hy (Phleum p a ense L.), a cool-season hexaploid pe ennial, is he mos impo an o age g ass
species in No dic coun ies. Ea lie analyses o gene ic di e si y in a collec ion o 96 genebank accessions o imo hy
wi h SSR ma ke s demons a ed high le els o di e si y bu could no esol e popula ion s uc u e. The e o e, we
examined a subse o 51 accessions wi h REMAP ma ke s, which a e based on e o ansposons, and compa ed he
di e si y esul s wi h hose ob ained wi h SSR ma ke s.
Resul s: Using ou p ime combina ions, 533 REMAP ma ke s we e analyzed, compa ed wi h 464 polymo phic alleles in
he 13 SSR loci p e iously. The a e age ma ke index, which desc ibes in o ma ion ob ained pe expe imen (pe p ime
combina ion o locus) was o e six imes highe wi h REMAPs. Mos o he a ia ion ound was wi hin accessions, wi h
somewha less, 89 %, o REMAPs, han o SSR, wi h 93 %.
Conclusions: SSRs e ealed di e ences in he le el o di e si y sligh ly be e han REMAPs bu nei he ma ke ype
could e eal any clea clus e ing o accessions based on coun ies, ege a ion zones, o di e en cul i a ypes. In ou
s udy, eliable e alua ion o SSR allele dosages was no possible, so each allele had o be handled as a dominan
ma ke . SSR and REMAP, which epo om di e en mechanisms o gene a ing gene ic di e si y and om di e en
genomic egions, oge he indica e a lack o popula ion s uc u e. Taken oge he , his likely e lec s he ou c ossing
and hexaploid na u e o imo hy a he han ailu es o ei he ma ke sys em.
Keywo ds: Gene ic di e si y, Gene ic s uc u e, Phleum p a ense L, REMAP, Re o ansposon ma ke , SSR, Mic osa elli e,
Timo hy
Backg ound
Timo hy (Phleum p a ense L.), a cool-season pe ennial,
is he mos impo an o age g ass species in No dic
coun ies. Gene ic di e si y has been p e iously assessed
[1] in a collec ion o 96 imo hy accessions, o which 88
we e o No dic o igin. Simple sequence epea (SSR)
ma ke s e ealed No dic imo hy accessions o be e y
polymo phic, ha ing signi ican di e ences in he le els
o di e si y be ween coun ies, ege a ion zones, and di -
e en cul i a ypes. Howe e , mos o he a ia ion
(94 %) exis ed wi hin accessions, and no clea clus e ing
o accessions based on any g ouping was obse ed. This
lack o esolu ion may ei he e lec he ou c ossing and
hexaploid na u e o imo hy o ha SSR ma ke s a e no
sui able o esol ing popula ion s uc u e in imo hy.
A wide ange o DNA ma ke s a e a ailable o di-
e si y s udies, which all ha e hei ad an ages and
disad an ages. SSRs a e ampli ied om single loci, bu
a e mul iallelic and highly polymo phic. Al hough
hey a e inhe i ed codominan ly, sepa a ion o di e -
en geno ypes may no be possible in a polyploid spe-
cies such as imo hy. The e o e, each allele has o be
ea ed as a dominan ma ke [1]; consequen ly, he
ma ke s ha a e ampli ied om he same SSR locus
a e no independen o each o he , and consequen-
ially in o ma ion is los . In he REMAP ( e o ans-
poson-mic osa elli e ampli ied polymo phism) ma ke s
[2, 3] assay, he di e si y is gene a ed by he in eg a-
ion o e o ansposons, which mo e in he genome
by a copy-and-pas e mechanism bu a e ixed in pos-
i ion upon inse ion [4]. They a e ubiqui ous and
abundan in plan genomes, whe e hey a e dispe sed
on all ch omosomes [5]. REMAP ma ke s a e amp-
li ied using a p ime designed o a conse ed
* Co espondence: [email p o ec ed]
1
G een Technology, Na u al Resou ces Ins i u e Finland (Luke), Mylly ie 1,
FI-31600 Jokioinen, Finland
Full lis o au ho in o ma ion is a ailable a he end o he a icle
© 2016 Tanhuanpää e al. Open Access This a icle is dis ibu ed unde he e ms o he C ea i e Commons A ibu ion 4.0
In e na ional License (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion, and
ep oduc ion in any medium, p o ided 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 license, and indica e i changes we e made. The C ea i e Commons Public Domain Dedica ion wai e
(h p://c ea i ecommons.o g/publicdomain/ze o/1.0/) applies o he da a made a ailable in his a icle, unless o he wise s a ed.
Tanhuanpää e al. He edi as (2016) 153:5
DOI 10.1186/s41065-016-0009-x
e o ansposon egion and ano he ancho ed o a
simple sequence epea . P oduc s om mul iple loci
a e p oduced in one PCR eac ion, each wi h only
wo allele al e na i es, a dominan one (ampli ica ion)
and a ecessi e one (non-ampli ica ion). Because he
mechanisms ha ac i a e e o ansposons [6] and
he eby gene a e inse ional polymo phisms a e ully
di e en han ha gene a ing SSR allelic a ia ion
(polyme ase slippage) [7], he wo ma ke sys ems
assay di e en componen s o gene ic di e si y. Fo
po a o [8], al al a [9], and g ape ine [10], e o ans-
poson and SSR ma ke s in combina ion we e shown
o be highly disc imina o y and e ec i e.
In he p e ious s udy [1], a collec ion o 96 imo hy
accessions was analyzed using 13 SSRs, hus desc ibing
di e si y only a his numbe o loci. On he o he hand,
hese 13 SSR loci ha bo ed as many as 499 alleles. In he
p esen s udy, we used REMAP ma ke s o s udying di-
e si y in a subse o 51 accessions and compa ed he e-
sul s wi h hose ob ained wi h SSR ma ke s. We wan ed
o de e mine i ano he ype o ma ke , which would e-
po om many mo e loca ions in he genome and as-
sess di e en genomic egions whe e di e si y is
gene a ed by a di e en mechanism, could desc ibe di-
e si y mo e e icien ly and also e eal popula ion s uc-
u e, pa icula ly o a polyploid species. Especially he
au onomous na u e o e o ansposon di e si y gene -
a ion and display, which is independen o he syn enic
o ganiza ion o polyploids, appea ed sui ed o clonal
polyploid species such as imo hy. We expec ed ha he
e o ansposon ma ke s should he eby be mo e likely
o ind gene ic s uc u e in imo hy, should i exis .
Me hods
Plan ma e ial
In he p e ious s udy [1], SSR ma ke s we e analyzed in
a collec ion o 96 imo hy accessions. Fi y-one o hese
we e selec ed o he p esen s udy o be sc eened also
wi h REMAP ma ke s (Table 1). Fi een o wen y an-
domly selec ed indi iduals pe accession we e in es i-
ga ed, in o al 945 indi iduals. The numbe o
indi iduals analyzed om each accession in he wo
s udies was no exac ly he same because 20 indi iduals
had o be omi ed due o hei poo ampli ica ion in RE-
MAP analysis.
The 51 accessions we e mos ly wild (30, loca ions in
Fig. 1 in [1]); se en each we e classi ied as land aces,
cul i a s, and o unknown cul i a ypes. Accessions
we e de i ed om all Scandina ian coun ies (Denma k,
8; Finland, 10; Iceland, 2; No way, 10; Sweden, 13). In
addi ion, eigh gene bank accessions (so-called exo ics)
o igina ing om non-Scandina ian coun ies we e in-
cluded in he s udy.
Ma ke analyses
DNA was ex ac ed using he me hod o Tinke e al.
[11] wi h some modi ica ions as desc ibed in Tanhuan-
pää and Manninen [1]. Using he iPBS (in e - p ime
binding si e) me hod, e o ansposon segmen s we e
isola ed om he imo hy genome, sequenced, and long
e minal epea s (LTRs) iden i ied [12]. LTR p ime s
we e designed o ma ch conse ed mo i s a o nea
hei e mini, acco ding o he me hods o Kalenda
e al. [13]. Fo REMAP ma ke ampli ica ion, ou di e -
en e o ansposon p ime s (TIM1 - 4) o g asses we e
used. These we e combined wi h 19 mic osa elli e-based
p ime s (ISSR + numbe ) ha con ain epea uni s (com-
posed o wo o h ee bases); he 3′ends o he p ime s
we e ancho ed by a single nucleo ide. Because analyzing
ma ke s by gel elec opho esis is e y labo ious, he
e o ansposon p ime s we e labelled wi h a luo escen
dye, FAM (5-ca boxy luo escein), HEX (hexachlo o-6-
ca boxy luo escein), o TET (6-ca boxy e achlo o luo -
escein) o enable esolu ion and isualiza ion o ampli i-
ca ion p oduc s wi h a MegaBACE
TM
500 Sequence
(GE Heal hca e, Buckinghamshi e, UK).
Fi y-nine REMAP p ime combina ions we e i s
es ed in a small se o indi iduals o hei unc ionali y
and e iciency o p oduce polymo phic bands. The ou
bes p ime combina ions we e chosen o inal analyses
(TIM1 wi h ISSR1, 15 and 20, and TIM2 wi h ISSR5).
These p ime s, oge he wi h hei sequences and p op-
e ies, a e shown in Table 2. The REMAP ma ke s we e
ampli ied in a eac ion olume o 10 μl, using 0.25 U o
FIREPol® DNA polyme ase I (Solis BioDyne OU, Ta u,
Es onia), bu e B wi h 2.5 mM MgCl
2
as supplied by he
enzyme manu ac u e , 200 μmol/L each dNTP, 10 ng o
DNA, and 500 nmol/L each p ime . The PCR p og am
was un on a PTC-220 DNA Engine Dyad
TM
Pel ie
The mal Cycle (MJ Resea ch, Wal ham, MA, USA) and
consis ed o an ini ial dena u a ion s ep o 2 min a 94 °
C; 32 cycles o 30 s a 94 °C, 30 s a 60 °C and 2 min a
72 °C; a inal ex ension s ep o 10 min a 72 °C. A e
PCR, he ampli ied p oduc s wi h di e en labels we e
combined o MegaBACE uns. SSRs we e de eloped o
imo hy [14], and analyses we e un as desc ibed p e i-
ously [1].
Da a analyses
Each REMAP agmen ep esen s a sepa a e locus, and
he p esence and absence o he agmen was sco ed in
a bina y code (1/0). Likewise, each SSR allele was ea ed
as a sepa a e locus and sco ed in a bina y code, e en
hough SSRs a e codominan ma ke s. This was because
we ound he e alua ion o allele dosages e y un eliable
in hexaploid imo hy. Di e si y indices o ma ke s, in-
cluding polymo phic in o ma ion con en (PIC), gene di-
e si y, and majo allele equency, we e calcula ed wi h
Tanhuanpää e al. He edi as (2016) 153:5 Page 2 o 10
Table 1 Fi y-one Phleum p a ense ssp. p a ense accessions analyzed in he s udy
Numbe code Accession no. Genebank Name Coun y Cul i a ype
1
Veg. zone
2
3 NGB10830 No dgen VA88119 Denma k W 1
4 NGB10831 No dgen HF88266 Denma k W 1
5 NGB15461 No dgen Vildbje g AC0103 Denma k W 1
6 NGB16650 No dgen Ejsing Denma k W 1
7 NGB1672 No dgen BILBO Denma k CV
8 NGB1675 No dgen POTA Denma k CV
9 NGB4053 No dgen SR SALTUM MH0202 Denma k W 1
10 NGB4548 No dgen NR FARUP MH0202 Denma k W 1
11 NGB132 No dgen LIPINLAHTI ME0901 SEP A Finland L 4
13 NGB14394 No dgen KÄRKÖLÄ HM0102 Finland W 3
16 NGB14404 No dgen PAATTINEN MH0201 Finland L 2
18 NGB14417 No dgen MEDVASTÖ MH0101 Finland W 2
20 NGB747 No dgen NUVVUS AK0401 Finland W 6
24 NGB1095 No dgen LAITASAARI ME0201 Finland L 4
27 NGB1111 No dgen MÄLÄSKÄ ME0101 Finland L 4
30 NGB1119 No dgen KATERMA ME0401 Finland L 4
32 NGB2791 No dgen NORRGÅRD AP0101 Finland L 3
35 NGB4066 No dgen TAMMISTO Finland CV
36 NGB4140 No dgen KORPA Iceland L
37 NGB4141 No dgen ADDA Iceland CV
42 NGB7592 No dgen SKJØLSVIK 01-5-46-5 No way W 3
45 NGB10785 No dgen SANDBU 01-6-49-4 No way W 5
47 NGB17194 No dgen I jo d 1-1-2-2 No way W 5
48 NGB17198 No dgen Ka asjok 1-1-3-2 No way W 5
49 NGB2169 No dgen BODIN No way CV
51 NGB2180 No dgen GRINDSTAD No way CV
53 NGB2918 No dgen HUSETER 01-9-70-1 No way W 2
57 NGB4226 No dgen HATLESTAD 01-7-56-3 No way W 5
59 NGB4231 No dgen GJERDÅKER 01-7-58-1 No way W 5
62 NGB7548 No dgen NAMSVATN 01-5-40-1 No way W 5
64 NGB722 No dgen KUOSSENJARKA JP0404 Sweden W 5
65 NGB728 No dgen PJESKER PH0405 Sweden W 4
66 NGB11428 No dgen JONATHAN Sweden CV
69 NGB14224 No dgen SÖNDRARP IB0101 Sweden W 2
71 NGB731 No dgen RÖRMYRBERG JP0204 Sweden W 4
73 NGB16975 No dgen NORRA KYLSÄTER FO0103 Sweden W 2
76 NGB16981 No dgen BRÄCKETORP FO0501 Sweden W 2
78 NGB1306 No dgen BRATTÅKER GB0101 Sweden W 4
81 NGB1327 No dgen HAMMARN PR0401 Sweden W 4
83 NGB1331 No dgen VÄSTANSJÖ SH0102 Sweden W 4
85 NGB1537 No dgen ESKELHEM TL0104 Sweden W 2
86 NGB2530 No dgen RÄMNE GJ0301 Sweden W 2
87 NGB4349 No dgen BENESTAD JK1506 Sweden W 1
Tanhuanpää e al. He edi as (2016) 153:5 Page 3 o 10
he p og am Powe ma ke 3.0 [15]. A ma ke index
(MI) o each REMAP p ime combina ion and each
SSR locus was de e mined by mul iplying he numbe o
polymo phic ma ke s gene a ed (EMF = E ec i e mul i-
plex a io) by a e age PIC alue [16]. I illus a es he
amoun o in o ma ion ob ained pe expe imen (pe
p ime combina ion o locus).
Gene ic di e si y in each accession was desc ibed wi h
i e di e en di e si y indices: 1) he numbe o all
ma ke s obse ed (A
A
), co ec ed o a sample size o
n = 15 wi h 1000 esamplings wi hou eplacemen ; 2)
he mean numbe o all ma ke s obse ed in each indi idual
(A
I
); 3) he mean numbe o pai wise di e ences (PWD)
(Euclidean dis ances) be ween indi iduals, which was
coun ed wi h he p og am ARLEQUIN e sion 2.000 [17];
4) Shannon’sdi e si yindexI[18]; 5) he pe cen age o
polymo phic loci. The las wo we e calcula ed using
he p og am GenAlex 6.4 [19, 20]. Co ela ions be ween
di e si y indices based on REMAP and SSR ma ke s, and
di e ences in he le el o di e si y be ween di e en
Table 1 Fi y-one Phleum p a ense ssp. p a ense accessions analyzed in he s udy (Con inued)
89 PI381926 GRIN F ance P
90 PI406317 GRIN Russia P
91 IHAR151908 IHAR Ge many P
92 PI210426 GRIN G eece P
93 PI325461 GRIN Russia P
94 PI204480 GRIN Tu key P
95 14G2400116 RICP Czech Republic P
96 RCAT040682 RCAT Hunga y W
1
CV ad anced cul i a , L adi ional cul i a , land ace, Ppending, unknown cul i a ype, Wwild popula ion, weedy
2
ege a ion zones, acco ding o [21]
Fig. 1 P incipal coo dina e analysis based on Nei’s gene ic dis ances be ween accessions based on: a533 REMAP ma ke s; b464 SSR ma ke s; c997
REMAP and SSR ma ke s
Tanhuanpää e al. He edi as (2016) 153:5 Page 4 o 10
g oups such as coun ies, ege a ion zones [21], o cul i a
ypes we e de e mined by ANOVA P oc GLM (SAS
En e p ise Guide 4.3).
The p og am GenAlex 6.4 [19, 20] was used o pe -
o m analysis o molecula a iance (AMOVA) [22]
which pa i ions o al gene ic a ia ion o wi hin- and
among-accession a iance componen s. The signi icance
o he esul s was es ed by pe mu ing he da a 999
imes. P incipal coo dina e analyses (PCA) based on
Nei’s gene ic dis ances [23] be ween accessions, and a
Man el es [24], which was used o compa e Nei’s dis-
ances based on REMAP o SSR da a, we e ca ied ou
wi h he so wa e GenAlex.
Resul s
Di e si y a ma ke loci
Fou REMAP p ime combina ions we e used o s udy-
ing di e si y o he 51 accessions. Because no all ag-
men s could be ead as ma ke peaks, selec ions we e
made on he basis o he size and shape o he peaks.
The numbe s o sco ed polymo phic ma ke s p oduced
by di e en p ime combina ions we e as ollows: TIM2
+ ISSR5, 91; TIM1 + ISSR20, 84; TIM1 + ISSR1, 209;
TIM1 + ISSR15, 149. A o al o 533 REMAP ma ke s
we e analyzed, anging in size om 80 o 650 bp. A o al
o 464 polymo phic alleles in he 13 SSR loci we e amp-
li ied om he 51 accessions, he numbe a ying om
13 o 71 pe accession [1]. The a e age di e si y indices
o REMAP ma ke s we e highe han hose o SSR
ma ke s (Table 3) leading o a six- old highe MI o
REMAPs.
Gene ic di e si y wi hin accessions
The obse ed numbe o REMAP ma ke s pe accession
a ied om 195 (PL204480) o 352 (NGB1672) (Table 4),
and he numbe o SSR alleles om 95 (NGB10785) o
194 (NGB1111). The e was only one p i a e REMAP
ma ke (in accession PL325461), bu 43 p i a e SSR al-
leles we e ound [1]. Di e si y indices o accessions s ud-
ied wi h REMAP o SSR ma ke s, espec i ely, a ied as
ollows: A
I
om 47.5 (PL204480) o 84.8 (NGB1672)
and om 28.4 (NGB10831) o 35.2 (NGB7592); PWD
om 53.1 (PL204480) o 100.2 (NGB1672) and om
28.9 (NGB10785) o 44.9 (NGB7592); I om 0.159
(RCAT040682) o 0.280 (NGB1672), wi h mean o 0.203
± 0.029, and om 0.109 (NGB722) o 0.156 (NGB1111),
wi h mean o 0.138 ± 0.014; he pe cen age o poly-
mo phic loci om 35.8 % (PL204480) o 64.9 %
(NGB1672), wi h a mean o 49.0 ± 7.3 %, and om
19.8 % (NGB10785) o 41.4 % (NGB1095), wi h a mean
o 34.4 ± 5.1 % (Table 4). The A
I
alues based on SSRs
changed sligh ly om he p e ious esul s [1] due o ex-
clusion o 20 indi iduals (see Me hods).
The s eng h o co ela ion be ween di e si y indices
based on REMAP o SSR ma ke s a ied depending on
he index. No co ela ion exis ed in he le el o A
I
. PWD
and Ico ela ed weakly a = 0.27 (P=0.059)and =
0.25 (P=0.073), espec i ely. The numbe o ma ke s
pe accession (A
A
) co ela ed mode a ely a = 0.37 (P=
0.0075) and he pe cen age o polymo phic loci s ongly
wi h = 0.44 (P= 0.0012). Nei’s gene ic dis ances be-
ween accessions based on REMAP and SSR da a co e-
la ed s ongly ( = 0.67, P< 0.001) wi h each o he .
When s udying le els o di e si y be ween coun ies,
ege a ion zones, o di e en cul i a ypes, we ound no
signi ican di e ences in A
A
and PWD based on REMAP
ma ke s (Table 5). On he o he hand, s a is ically sig-
ni ican (P< 0.05) di e ences in A
A
and PWD be ween
di e en ege a ion zones and in A
A
be ween di e en
Table 2 REMAP p ime s ha we e used in he analysis o imo hy di e si y, hei sequences and p ope ies
Name Sequence n Tm (°C) CG % Linguis ic
complexi y (%)
TIM1 GGTGCCGGCATCGATCCTTTCA 22 62.4 59.1 88
TIM2 ACGAGTGAGGACAAAGTGCGCAGA 24 61.9 54.2 79
ISSR1 ACCACCACCACCACCACCC 19 63.2 68.4 24
ISSR5 AGCAGCAGCAGCAGCAGCG 19 64.4 68.4 30
ISSR15 GTGGTGGTGGTGGTGGTGGTGA 22 64.2 63.6 28
ISSR20 TGCTGCTGCTGCTGCTGCC 19 64.6 68.4 30
n nucleo ides, Tm mel ing empe a u e, CG % pe cen age o C and G bases
Table 3 Compa ison o di e si y measu es o REMAP and SSR
ma ke s in he analysis o 51 imo hy accessions
REMAP SSR
No. o p ime combina ions o loci 4 13
To al no. o ma ke s 533 464
No. o ma ke s pe p ime
combina ion o locus = EMF
1
133.3 35.7
PIC, a e age 0.131 0.086
Ma ke s wi h PIC > 0.1 258 = 48 % 148 = 32 %
Ma ke s wi h MAF < 0.1 365 = 68 % 371 = 80 %
A e age gene di e si y 0.152 0.098
Ma ke index (MI) = EMF x PIC 17.4 3.1
1
e ec i e mul iplex a io
Tanhuanpää e al. He edi as (2016) 153:5 Page 5 o 10

Table 4 REMAP and SSR di e si y in 51 imo hy accessions
REMAP ( o al no. o ma ke s = 533) SSR ( o al no. o ma ke s
1
= 464)
Accession No. o ind. No. ma ke s A
A
2
A
I
3
PWD
4
I
5
% polymo phic
loci
No. ma ke s A
A
2
A
I
3
PWD
4
I
5
% polymo phic
loci
NGB10830 19 213 200.5 55.8 61.0 0.170 38.6 124 116.7 28.8 30.7 0.114 26.7
NGB10831 18 221 207.3 50.1 56.2 0.161 40.5 148 137.9 28.4 32.5 0.126 31.3
NGB15461 19 323 302.2 73.1 85.7 0.244 59.3 188 171.5 33.3 44.0 0.155 40.5
NGB16650 18 281 264.9 60.6 71.8 0.206 52.0 142 132.9 31.3 36.2 0.127 30.6
NGB1672 19 352 328.7 84.8 100.2 0.280 64.9 162 149.7 31.7 40.1 0.137 34.7
NGB1675 20 297 277.8 78.4 87.2 0.244 54.8 119 110.1 30.0 32.5 0.111 25.6
NGB4053 18 239 226.9 58.8 65.3 0.185 44.7 145 135.2 30.7 36.2 0.135 31.3
NGB4548 19 236 219.2 54.8 61.3 0.177 43.3 154 139.4 30.1 35.1 0.137 33.2
NGB132 19 313 293.0 75.6 84.7 0.242 57.2 186 167.5 31.1 39.1 0.144 39.7
NGB14394 19 276 255.6 59.3 71.2 0.204 50.8 175 162.2 32.7 42.6 0.148 37.7
NGB14404 20 295 266.0 65.7 74.5 0.214 54.6 176 158.4 33.6 40.5 0.144 37.7
NGB14417 18 213 201.4 60.7 59.4 0.166 38.8 121 115.2 28.4 32.5 0.115 25.9
NGB747 20 310 285.9 75.2 81.5 0.235 56.8 144 131.4 30.3 34.7 0.126 30.8
NGB1095 20 330 300.6 72.9 85.0 0.245 61.2 192 170.2 34.2 43.1 0.153 41.4
NGB1111 19 308 289.2 64.3 78.1 0.229 56.7 194 172.7 33.8 44.3 0.156 41.2
NGB1119 19 226 210.1 52.8 58.6 0.170 41.5 189 170.9 31.9 40.5 0.149 40.7
NGB2791 19 315 294.5 67.7 80.7 0.238 58.2 186 169.8 33.3 42.5 0.152 40.1
NGB4066 19 259 239.8 53.6 65.4 0.188 47.5 180 164.5 32.7 38.6 0.149 38.4
NGB4140 18 263 248.3 55.0 70.5 0.201 49.0 193 176.2 33.4 43.9 0.160 40.7
NGB4141 19 260 246.6 66.5 71.6 0.206 48.0 117 111.2 31.3 34.6 0.118 25.0
NGB7592 16 244 239.2 64.5 68.6 0.192 44.7 182 168.4 35.2 44.9 0.155 37.1
NGB10785 18 202 192.0 67.0 59.9 0.165 36.6 95 86.8 32.7 28.9 0.092 19.8
NGB17194 20 286 260.7 62.1 72.7 0.213 52.7 166 149.5 31.4 37.2 0.134 35.8
NGB17198 18 336 313.7 70.1 85.7 0.245 61.9 179 166.4 32.4 42.5 0.150 38.6
NGB2169 19 290 269.0 62.9 76.5 0.219 53.8 167 153.4 30.3 38.8 0.140 35.8
NGB2180 18 304 284.6 67.7 77.4 0.226 56.1 185 169.2 31.2 39.5 0.146 39.7
NGB2918 20 332 305.3 73.4 88.2 0.252 61.7 164 150.2 32.1 39.0 0.140 35.3
NGB4226 17 236 228.4 64.4 67.9 0.191 43.3 158 151.6 34.9 42.2 0.142 34.1
NGB4231 19 230 216.4 56.3 63.0 0.183 42.2 156 143.6 31.3 38.4 0.137 33.6
NGB7548 17 272 260.1 61.4 70.0 0.201 49.7 150 140.8 29.8 35.4 0.131 31.9
NGB722 19 256 241.9 77.4 71.7 0.202 45.8 115 108.8 29.5 32.7 0.109 24.4
NGB728 18 265 246.3 56.7 65.5 0.196 49.5 182 171.2 33.9 43.4 0.155 39.2
NGB11428 16 231 226.2 58.9 64.4 0.180 42.0 139 138 34.2 39.8 0.135 29.7
NGB14224 19 293 273.6 65.7 79.2 0.224 54.0 175 158.4 33.4 42.2 0.146 36.6
NGB731 20 311 288.5 79.8 84.6 0.241 57.0 182 163.2 31.9 41.1 0.150 39.2
NGB16975 19 258 237.0 56.2 66.4 0.189 47.5 176 161.8 33.5 42.2 0.150 37.7
NGB16981 15 245 245.0 59.2 69.6 0.193 44.7 153 149.8 33.6 41.7 0.138 32.1
NGB1306 18 298 280.5 64.0 76.0 0.224 55.0 184 173.6 32.9 41.4 0.152 39.7
NGB1327 19 284 267.1 70.6 79.8 0.225 52.2 162 151.4 33.6 41.2 0.144 34.9
NGB1331 19 223 208.5 57.1 60.7 0.172 40.3 170 158.4 32.8 42.1 0.144 36.4
NGB1537 16 296 290.1 68.2 82.4 0.231 54.6 139 131.7 30.5 36.1 0.125 28.9
NGB2530 19 235 218.0 57.9 61.8 0.177 42.8 165 152.8 31.9 39.6 0.141 35.6
Tanhuanpää e al. He edi as (2016) 153:5 Page 6 o 10
cul i a ypes we e ound wi h SSR ma ke s (Table 5). In
he p e ious s udy wi h 96 accessions analyzed wi h SSR
ma ke s, we ound signi ican di e ences (P< 0.05) in
le els o di e si y in all g oups [1]. When he o al num-
be o ma ke s was s udied on an indi idual a he han
accession le el (A
I
), signi ican di e ences o each
g ouping and wi h bo h ma ke ypes we e disco e ed
(Table 5). Howe e , hese di e ences explained only a
mino ac ion o a ia ion be ween indi iduals (1 o
5 %).
Gene ic di e gence be ween accessions and g oups
AMOVA was pe o med in o de o di ide he o al gen-
e ic a ia ion in o h ee componen s: a ia ion wi hin
accessions, among accessions, and among coun ies.
Mos o he a ia ion in he s udied ma e ial was ound
wi hin accessions: 89 % when based on REMAP
ma ke s, 93 % when based on SSR ma ke s, and 91 %
when based on bo h ma ke ypes (Table 6).
No gene ic di e gence was obse ed be ween ege a-
ion zones o cul i a ypes ei he using SSR o REMAP
ma ke s o bo h (AMOVA, P< 0.05), which migh be
due o he small numbe s o membe s in di e en clas-
ses. Howe e , he same esul was ob ained wi h SSR
ma ke s when 96 accessions we e s udied [1]. In PCA
analysis as well, no clus e ing o accessions based on
coun ies, ege a ion zones, o cul i a ypes was seen
(Fig. 1). The i s wo axes espec i ely explained 44.1 %,
45.8 %, o 41.1 % o he a ia ion when REMAPs, SSRs,
o bo h ma ke ypes we e used in he analysis.
Discussion
P e iously, SSR ma ke s e ealed imo hy o be e y di-
e se bo h on he indi idual and accession le el when
Table 4 REMAP and SSR di e si y in 51 imo hy accessions (Con inued)
NGB4349 20 267 244.0 62.7 70.7 0.203 49.3 171 153.4 32.4 40.6 0.141 36.9
PL381926 19 240 222.2 67.4 66.5 0.187 43.9 131 122.3 31.1 34.6 0.125 28.0
PL406317 19 243 226.8 56.5 64.1 0.183 44.5 165 151.7 31.8 38.5 0.145 35.6
IHAR151908 17 259 249.0 60.3 66.4 0.192 48.0 150 137.4 31.1 34.8 0.126 30.6
PL210426 17 245 237.0 65.5 69.4 0.194 44.8 146 138.9 31.8 38.3 0.131 30.6
PL325461 17 233 223.4 55.5 61.2 0.175 43.2 170 157.8 30.8 39.8 0.138 34.3
PL204480 19 195 182.4 47.5 53.1 0.151 35.8 158 144.3 31.6 37.6 0.133 33.6
14G2400116 19 234 220.2 57.1 65.8 0.185 42.8 186 170.5 34.2 44.0 0.153 39.9
RCAT040682 20 209 191.5 50.8 55.9 0.159 38.6 157 143.5 30.5 39.1 0.141 33.8
1
each SSR allele ea ed as a sepa a e ma ke
2
co ec ed numbe o all ma ke s in each accession
3
mean numbe o all ma ke s obse ed in each indi idual
4
mean numbe o pai wise di e ences (Euclidean dis ances) be ween indi iduals in each accession
5
Shannon’s di e si y index
Table 5 ANOVA ables indica ing F- alues, signi icance le els P, and R
2
o compa isons be ween di e en g oups o hei le els o
REMAP and SSR di e si y
REMAP To al no. o ma ke s (A
A)
No. o pai wise di e ences (PWD) Numbe o ma ke s pe indi idual
Di e si y index d F PR
2
FPR
2
FPR
2
G ouping
Accession 50 6.11 <0.001 0.25
Coun y 5 1.73 0.147 0.16 1.49 0.210 0.14 5.08 <0.001 0.03
Vege a ion zone 5 0.74 0.602 0.11 0.74 0.600 0.11 5.08 <0.001 0.04
Cul i a ype 2 2.03 0.144 0.09 1.96 0.153 0.09 3.98 0.019 0.01
SSR To al no. o ma ke s (A
A)
No. o pai wise di e ences (PWD) Numbe o ma ke s pe indi idual
Di e si y index d F PR
2
FPR
2
FPR
2
G ouping
Accession 50 3.18 <0.001 0.15
Coun y 5 1.15 0.348 0.11 1.53 0.200 0.15 5.28 <0.001 0.03
Vege a ion zone 5 3.90 0.008 0.40 3.49 0.014 0.38 6.67 <0.001 0.05
Cul i a ype 2 4.70 0.015 0.19 1.71 0.194 0.08 4.98 0.007 0.01
Tanhuanpää e al. He edi as (2016) 153:5 Page 7 o 10
s udied in a collec ion o 96 accessions. Because i was
impossible wi h SSRs o esol e any popula ion o geo-
g aphical s uc u e [1], we he e ha e applied a e y di -
e en kind o neu al ma ke , REMAPs, which a e based
on displaying e o ansposon inse ions.
Bo h REMAPs and SSRs we e highly polymo phic.
Va ia ion was obse ed mos ly wi hin accessions bu
wi h sligh ly smalle p opo ion o REMAPs (89 % s.
93 %). This di e ence may be due o he biology o how
SSR and e o ansposon polymo phisms a e gene a ed.
SSRs a e gene a ed by eplica ion slippage [7], a p ocess
expec ed o be independen o he en i onmen . In con-
as , e o ansposons a e known o be ac i a ed by
bo h bio ic and abio ic s esses [6], condi ions which
may well be g ea e in some popula ions compa ed wi h
o he s. Popula ion-le el s ess would he eby lowe he
p opo ion o polymo phism on he indi idual le el and
inc ease i on popula ion o geog aphic le els.
Di e si y indices in accessions we e lowe o SSR han
o REMAP ma ke s. This is likely because SSR ma ke s
(i.e., alleles) a e no independen o each o he ; he e is a
heo e ical maximum numbe o ma ke s ha can exis
in one indi idual. I all SSR loci would ampli y om all
h ee genomes o Phleum, he maximum numbe o
ma ke s would be 78 (13 loci, 6 alleles in each). How-
e e , he e is e idence ha imo hy is an allopolyploid
[25]. Allopolyploidy is consis en wi h ou ea lie esul s
[1], wi h some SSR loci ound only in one genome
whe eas o he s we e p esen in all h ee. The e o e, he
eal maximum numbe o SSR alleles in any one indi id-
ual lies somewhe e be ween 26 and 78. In he p esen
s udy, he obse ed maximum was 45.
Polyploids ep esen abou 50 % o lowe ing plan s
[26]. In polyploids, he p oblem o lack o independence
be ween SSR loci is pa icula ly a p oblem, bu gi en a
e y high numbe o loci de eloped om he genome se-
quences o majo c ops such as co on o whea ,
ch omosome-speci ic ma ke s can be eco e ed [27]. Fo
ag icul u al species wi hou e e ence genomes such as
imo hy o o many wild species [28], selec ion o
ma ke s wi h diploid inhe i ance can educe he usable
loci o e y low numbe s.
In con as o SSRs, no limi exis s o he maximum
amoun o REMAPs in one indi idual because e o-
ansposon inse ions a e independen o each o he .
Mo eo e , di e en e o ansposon amilies, such as in
he hexaploid whea genomes [29], show di e en e olu-
iona y his o ies, enabling disc imina ion be ween home-
ologues. Re o ansposon ma ke s ha e been deployed
e ec i ely o e en he highly polyploid suga cane [30].
Al hough codominan REMAPs also exis , codominance
does no es ic he possibili y o co-exis ence o
ma ke s in one indi idual. The maximum amoun o
REMAPs obse ed in one indi idual in he p esen s udy
was 121. Co ela ions be ween di e si y indices based on
REMAP o SSR ma ke s we e mos ly low o mode a e
because he wo ma ke sys ems epo om di e en
genomic egions whe e polymo phisms a e gene a ed by
di e en p ocesses. On he o he hand, e en hough
SSRs could be ea ed as codominan ma ke s, i has
been sugges ed ha la ge simila i ies be ween di e si y
indices wi h dominan ma ke s bu somewha lowe be-
ween dominan ma ke s and SSRs a e due o insu i-
cien numbe s o analyzed SSR loci [31].
When using ma ke s o measu ing dis ances, PWD
be ween indi iduals co ela ed weakly ( = 0.26) bu gen-
e ic dis ances be ween accessions s ongly ( = 0.67) be-
ween he wo ma ke ypes. PWD is based on he
Table 6 Analysis o molecula a iance in 51 imo hy accessions based on 533 REMAP ma ke s, 464 SSRs o bo h
REMAP SSR REMAP and SSR
Sou ce d SS MS Va iance
componen s
%
o al
SS MS Va iance
componen s
%
o al
SS MS Va iance
componen s
%
o al
Among coun ies 5 896.84 179.37 0.42 1 574.09 114.82 0.21 1 1471.68 294.34 0.63 1
Among accessions/
coun ies
45 5181.97 115.15 4.21 10 3719.53 82.66 2.45 6 8901.50 197.81 6.66 8
Wi hin accessions 894 33312.58 37.26 37.26 89 33271.32 37.22 37.22 93 66583.90 74.48 74.48 91
To al 944 39391.39 41.89 100 37564.94 39.88 100 76957.08 81.77 100
S a Value Value Value
PhiRT 0.010 0.005 0.008
PhiPR 0.101 0.062 0.082
PhiPT 0.110 0.067 0.089
P obabili y, P( and ≥da a), o PhiRT, PhiPR and PhiPT = 0.001, and is based on pe mu a ion ac oss he ull da a se
PhiRT = AC / (WA + AA + AC) = AC / TOT
PhiPR = AA / (WA + AA)
PhiPT = (AA + AC) / (WA + AA + AC) = (AA + AC) / TOT
Key: AC = es . a . among coun ies, AA = es . a . among accessions, WA = es . a . wi hin accessions
Tanhuanpää e al. He edi as (2016) 153:5 Page 8 o 10
Euclidean dis ances be ween indi iduals whe eas dis-
ances be ween accessions a e based on ma ke equen-
cies. The same so o esul –poo o nonexis en
indi idual-by-indi idual co ela ions bu mode a e co -
ela ion be ween accessions –was ob ained when ampli-
ied agmen leng h polymo phisms (AFLPs), which a e
compa able o REMAPs by being a mul ilocus and dom-
inan ma ke ype, and SSRs we e compa ed [32]. In po-
a o, a low co ela ion o SSR and REMAP ma ke s ( =
0.17) in he Man el’s ma ix co espondence es was
ound [8].
Compa ing he wo ma ke ypes, REMAP ma ke s
we e mo e cos -e icien . The PCRs o ou REMAP p i-
me combina ions we e made sepa a ely, and p oduc s
om wo di e en combina ions wi h di e en luo es-
cen labels we e combined o MegaBACE uns. As a
consequence, o he whole di e si y analysis s udy (945
samples), 40 PCRs on 96-well mic o i e pla es we e
made and analyzed in 20 Megabace uns. A o al o 533
polymo phic ma ke s was p oduced. On he o he hand,
he 13 SSR loci we e mul iplexed in o 5 PCR eac ions
and analyzed in 5 MegaBACE uns, equi ing in o al 50
PCR pla es and 50 MegaBACE uns. In addi ion, some
planning and op imiza ion was equi ed in o de o
mul iplex he PCR eac ions o he a ious SSR loci. A
o al o 464 SSR ma ke s (i.e., alleles) was ampli ied. Ac-
co dingly, mo e REMAP ma ke s (i.e., loci) we e p o-
duced wi h less labo , money, and ime. The MI was
o e six- old highe wi h REMAPs, which is no due
only o he need o in e p e SSR alleles as sepa a e
ma ke s bu is also ypical o ma ke s wi h an e ec i e
mul iplex a io, and has been de ec ed also when AFLPs
ha e been compa ed wi h SSRs [16].
Knowledge o gene ic a ia ion and ela ionships be-
ween indi iduals and accessions is essen ial when con-
se ing and using gene ic esou ces. E alua ion o
gene ic di e si y equi es analysis o mul iple ma ke s as
e icien ly as possible. Choosing a sui able ma ke ype,
se e al aspec s ha e o be aken in o accoun , no only
expec ed he e ozygosi y and ma ke index, bu also ech-
nical di icul y, ease o geno yping, cos , and a ailabili y.
Technically, he e we e no di e ences be ween REMAPs
and SSRs and we encoun e ed analysis di icul ies wi h
bo h ma ke ypes. All SSR peaks con ained some deg ee
o s u e , which complica ed he iden i ica ion o alleles.
On he o he hand, in e p e a ion o REMAP ma ke s
was e y slow because se e al ma ke s we e ampli ied in
one PCR eac ion and he e was a wide a ia ion in peak
heigh s. All peaks could no be analyzed and selec ions
had o be made acco ding o peak heigh s and e-
quency. Di icul ies in sco ing hinde ed he use o au o-
ma ed analysis p og ams o bo h ma ke ypes.
Rega ding a ailabili y, he e a e uni e sal e o ans-
poson p ime s ha can be used in any species, and
p ime s speci ic o G aminae also ha e a as ange o
use. Mo eo e , SSRs ha e no been de eloped o e e y
species, and ans o m a es om one species o ano he
depends on he gene ic dis ance o he axa [33]. These
gene al conclusions ega ding he u ili y o SSR and
e o ansposon ma ke s alone and in combina ion a e
consis en wi h hose o ou di e se dico species, dis-
an om he monoco imo hy [8–10].
Conclusions
When di e si y in a polyploid species is examined, whe e
he codominan na u e o SSRs is o no use, dominan
REMAP ma ke s, as analyzed by size on a sequence ,
we e mo e cos -e icien . REMAPs also desc ibed di e -
si y om a la ge segmen o he genome compa ed o
he same numbe o SSR alleles. On he o he hand,
SSRs de ec ed di e ences in he le el o di e si y in di -
e en g oups be e han REMAPs. Fu he mo e, p i a e
SSR alleles we e ound, making SSRs be e o accession
iden i ica ion. P i a e alleles, howe e , can be de eloped
om e o ansposon ma ke s using he RBIP ( e o-
ansposon-based inse ion polymo phism) and ISBP
(inse ion si e-based polymo phism) me hods, which a e
locus-speci ic [34]. Gene ic dis ances be ween accessions
we e simila wi h REMAP o SSR ma ke s, bu nei he
ma ke ype could e eal any clea di e gence be ween
ege a ion zones, cul i a ypes o coun ies in he poly-
ploid, e y polymo phic and he e ozygous imo hy spe-
cies. SSR and REMAP polymo phisms de i e om e y
di e en mechanisms. Va ia ions in SSR numbe s a in-
di idual loci de i e om polyme ase slippage du ing
eplica ion. In con as , e o ansposon inse ions,
which can be s ess-d i en, gene a e he p iming si es
o e o ansposon-based ma ke me hods. Gi en he
as ly di e en numbe s o mic osa elli e and e o ans-
poson loci que ied by he ma ke sys ems used, which
epo om e y di e en genomic egions, he ac ha
hey oge he show a lack o s uc u e likely e lec s he
ou c ossing and hexaploid na u e o imo hy a he han
ailu es o ei he ma ke sys em. Bo h e o ansposons
and SSRs, howe e , a e neu al ma ke s; pa e ns o a i-
a ion in he gene space o imo hy, such as h ough sin-
gle nucleo ide polymo phism (SNP) geno yping, emain
o be explo ed. These would allow he possibili y o
e alua e allele dosages, he eby inc easing he in o ma-
ion embodied in each locus. SNP ma ke s ha e been
used in suga cane, a complex au opolyploid species, o
es ima e ploidy le el and also he dosage o SNPs [35].
The a ailabili y o SNP ma ke s has inc eased wi h he
in en ion o he geno yping by sequencing s a egy
(GBS) [36] and a ecen s udy p esen s i s use o e alu-
a e allele equencies in popula ions in an ou b eeding
species, pe ennial yeg ass [37]. Such echniques could
be applied o imo hy as well o s udy he s uc u e o
Tanhuanpää e al. He edi as (2016) 153:5 Page 9 o 10