免费文献传递   相关文献

Tracing the Origin of the Bimodal Karyotypes of the Tribe Lilieae (Liliaceae) Based on Comparative Karyotype Analyses

基于比较核型分析方法追溯百合族 (百合科) 二型性核型的起源



全 文 :基于比较核型分析方法追溯百合族 (百合科)
二型性核型的起源∗
殷根深1ꎬ2ꎬ 杨志云1ꎬ 蒋镇宇3ꎬ 龚  洵1∗∗
(1 中国科学院昆明植物研究所资源植物与生物技术所级重点实验室ꎬ 云南 昆明  650201ꎻ
2 中国科学院大学ꎬ 北京  100049ꎻ 3 国立成功大学生命科学系ꎬ 台湾 台南  701)
摘要: 百合族具有非常一致的二型性核型ꎬ 由 4条长染色体以及 20条短染色体组成ꎮ 目前有两个假说解
释其二型性核型的来源ꎬ 着丝粒横裂和多倍化ꎮ 但是ꎬ 具体是哪一种机制起主要作用仍然不清楚ꎮ 根据文
献以及自己的实验结果ꎬ 我们整理并重新分析了百合亚科和美德兰亚科所有属的核型资料ꎮ 比较核型分析
结果表明ꎬ 来自单条染色体的特征、 染色体臂数、 核型不对称性以及染色体的相对长度诸方面的证据都支
持着丝粒横列是百合族核型进化的主要机制ꎬ 但不能排除其它的机制也在起着作用ꎬ 如臂间倒位和易位ꎮ
臂间倒位和易位可能在郁金香族的核型进化中起着主要的作用ꎮ 另外ꎬ 本研究还报道了三个种的核型ꎬ 粗
茎贝母 (Fritillaria crassicaulis)、 准格尔郁金香 (Tulipa suaveolens) 和尖果洼瓣花 (Gagea oxycarpa)ꎮ
关键词: 二型核型ꎻ 着丝粒横裂ꎻ 比较核型分析ꎻ 核型进化ꎻ 百合族
中图分类号: Q 941ꎬ Q 942        文献标识码: A            文章编号: 2095-0845(2014)06-737-10
Tracing the Origin of the Bimodal Karyotypes of the
Tribe Lilieae (Liliaceae) Based on Comparative
Karyotype Analyses
YIN Gen ̄Shen1ꎬ2ꎬ YANG Zhi ̄Yun1ꎬ CHIANG Tzen ̄Yuh3ꎬ GONG Xun1∗∗
( 1 Key Laboratory of Economic Plants and Biotechnologyꎬ Kunming Institute of Botanyꎬ Chinese Academy of Sciencesꎬ
Kunming 650201ꎬ Chinaꎻ 2 University of Chinese Academy of Sciencesꎬ Beijing 100049ꎬ Chinaꎻ
3 Department of Life Sciencesꎬ National Cheng ̄Kung Universityꎬ Tainan 701ꎬ Taiwanꎬ China)
Abstract: The karyotypes of the tribe Lilieae are rather constant bimodalꎬ which is composed of 4 longer and 20
shorter chromosomes. Howeverꎬ it remains unclear which mechanisms ( centric fission or polyploidisation) might
contribute to the origin of biomodality in this tribe. Hereꎬ we collected the published data for all genera of the Li ̄
lioideae and Medeoloideaeꎬ and re ̄analyzed the karyotype data. The evidence of individual chromosome featuresꎬ
fundamental number (FN)ꎬ karyotype asymmetry and chromosome relative length indicated that centric fission is the
main mechanism underlying karyotypic evolution in the Lilieae. Howeverꎬ centric fission is not expected to occur in
isolationꎬ and different mechanisms of karyotype changeꎬ such as pericentric inversion and segment translocationꎬ
are not mutually exclusive. Pericentric inversion and segment translocation may have played major roles in the karyo ̄
type evolution of the Tulipeae. Additionallyꎬ the karyotype analyses are carried out for the first time in three taxa:
Fritillaria crassicaulisꎬ Tulipa suaveolensꎬ and Gagea oxycarpa.
Key words: Bimodal karyotypeꎻ Centric fissionꎻ Comparative karyotype analysisꎻ Karyotype evolutionꎻ Lilieae
植 物 分 类 与 资 源 学 报  2014ꎬ 36 (6): 737~746
Plant Diversity and Resources                                    DOI: 10.7677 / ynzwyj201414018

∗∗
Funding: The National Natural Science Foundation of China (31170633)
Author for correspondenceꎻ E ̄mail: gongxun@mail􀆰 kib􀆰 ac􀆰 cn
Received date: 2014-02-12ꎬ Accepted date: 2014-04-09
作者简介: 殷根深 (1985-) 男ꎬ 博士研究生ꎬ 主要从事染色体进化研究ꎮ E ̄mail: yingenshen@126􀆰 com
  The tribe Lilieae (Takhtajanꎬ 2009ꎻ Tamuraꎬ
1998)ꎬ belonging to the family Liliaceaeꎬ contains
four genera: Lilium L. (including Nomocharis Fran ̄
ch.)ꎬ Fritillaria L.ꎬ Notholirion Wallich ex Boissier
and Cardiocrinum (Endlicher) Lindley. Some spe ̄
cies of the Lilieae are the most beautiful ornamentals
of the plant kingdom. Many species of Liliumꎬ Fritil ̄
lariaꎬ and Cardiocrinumꎬ are cultivated as potted
plants or cut flowers for their bright colour and fra ̄
grance (Woodcock and Stearnꎬ 1950). Several spe ̄
cies of tribe are important Chinese herbs. For exam ̄
pleꎬ the bulbs of Fritillaria cirrhosa D. Donꎬ F􀆰 uni ̄
bracteata P􀆰 K. Hsiao & K􀆰 C. Hsia in S􀆰 C. Chen &
K􀆰 C. Hsiaꎬ F􀆰 przewalskii Maximowicz in Trautvet ̄
terꎬ and F􀆰 delavayi Franchet are used as a tradition ̄
al Chinese medicine named “Chuan bei mu” for the
treatment of cough and chest congestionꎻ Notholirion
bulbuliferum ( Lingelsh.) Stearn is known as “ Tai
bai mi”ꎬ which is used to treat stomach bloating
(Chinese Pharmacopoeia Commissionꎬ 2010).
Lilieae is belonged to the subfamily Lilioideaeꎬ
which is sister to another tribeꎬ Tulipeae. Tribe Tuli ̄
peae contains four generaꎬ Gagea Salisb. (including
Lloydia Salisb. et Reichb.)ꎬ Tulipa L.ꎬ Amana Hon ̄
daꎬ and Erythronium L. (Peruzzi et al.ꎬ 2009). The
species of Lilioideae have been of great interest to
botanists for their cytological diversity in sizeꎬ num ̄
ber and structure (Caveꎬ 1970ꎻ Gao et al.ꎬ 2012ꎻ
Peruzzi et al.ꎬ 2009ꎻ Satoꎬ 1942ꎻ Tamuraꎬ 1995).
Variations in chromosome morphology between Lilie ̄
ae and Tulipeae were huge: the tribe Lilieae shares
a rather constant and distinctive bimodal karyotype
consisting of 4 longer and 20 shorter chromosomes
(Stewart 1947ꎻ Tamuraꎬ 1998)ꎬ while the karyo ̄
types of the Tulipeae exhibit a continuous range of
chromosome sizes.
Bimodal karyotypes are unique in consisting of
two sharply distinct size classes of chromosomesꎬ long
and short (McKain et al.ꎬ 2012ꎻ Pires et al.ꎬ 2006ꎻ
Stebbinsꎬ 1971). It represents a very specialized karyo ̄
typic form occuring in the monocotyledons (e􀆰 g. Ben ̄
nett et al.ꎬ 1992ꎻ Gitai et al.ꎬ 2005ꎻ Goldblatt and
Takeiꎬ 1997ꎻ Stedjeꎬ 1989)ꎬ and may have different
origins (Stebbinsꎬ 1971ꎻ Bennett et al.ꎬ 1992ꎻ Vosaꎬ
2005). As for Lilieaeꎬ it has been suggested that the
bimodal karyotype arose through multiple reciprocal
centric fissionsꎬ most likely involving the five smaller
chromosome pairs of a Medeoloideae—like ancestor
(Peruzzi et al.ꎬ 2009ꎻ Tamuraꎬ 1995)ꎬ or through
polyploidyꎬ most likely involving parental species of
different chromosome sizes ( Peruzzi et al.ꎬ 2009).
Howeverꎬ the test of the polyploidisation hypothesis
has been hampered by the lack of candidate parental
species. The chromosome evolution of Tulipeae may
have involved an equal change in the amount of DNA
for each chromosome arm (Peruzzi et al.ꎬ 2009).
Centric fission involves the division of a single
chromosome into two smaller telocentric chromo ̄
somes (Stebbinsꎬ 1971ꎻ Schubert and Lysakꎬ 2011ꎻ
Sharma and Senꎬ 2002). In this processꎬ the total
number of chromosome arms (NF) and the amount
of genomic material (C ̄value) stay the sameꎬ where ̄
as the number of chromosomes and karyotype sym ̄
metry increases (Stebbinsꎬ 1971ꎻ Jonesꎬ 1998ꎻ Par ̄
do ̄Manuel de Villena and Sapienzaꎬ 2001ꎻ Schubert
and Lysakꎬ 2011ꎻ Souza et al.ꎬ 2011). Thereforeꎬ
we can determine whether centric fission has been
chiefly responsible for chromosomal changes by com ̄
paring karyotypes of related species.
To dateꎬ chromosome data are available for rep ̄
resentative species of all genera from the Lilioideae
and Medeoloideaeꎬ the phylogenetic relationships a ̄
mong these families were fully resolved. Thereforeꎬ it
is now possible to infer the hypotheses regarding the
origin of chromosomal bimodality in the Lilieae. We
defined the Medeoloideae to be the outgroupꎬ combi ̄
ning our data with previously published dataꎬ and
used comparative karyotype methods for the following
goals: (1) conducting the cytogenetic characterisati ̄
on for the first time in three species of the subfamily
Lilioideaeꎻ (2) testing the centric fission hypothesis
of bimodal karyotype in the Lilieaeꎻ and (3) reac ̄
hing a preliminary conclusion for the continuous
range of chromosome sizes observed in the Tulipeae.
837                                  植 物 分 类 与 资 源 学 报                            第 36卷
1  Materials and methods
1􀆰 1  Sample selection and cytological studies
Karyotype data for representative species of all
genera of Lilioideae (155 accessions) and Medeoloid ̄
eae (8 accessions) were compiled from the published
literature (127 accessions) and our data (36 acces ̄
sions) (Supplementary dataꎬ http: / / journal􀆰 kib􀆰 ac􀆰
cn / UserFiles / File / YGS􀆰 xlsx). Only literature data
containing ideogramsꎬ and / or karyotype measurements
and / or good metaphase plates were included in ana ̄
lyses to ensure the accuracy of the measurements.
The experimental procedures for our data are as
follows: actively growing root tips were cut for chro ̄
mosome observation from the transplanted plants.
The root tips were pretreated with 0􀆰 1% colchicines
at room temperature for 4 h and then fixed in Carnoy
I (3 ∶ 1 95% ethanol: glacial acetic acid) at room
temperature for 12 h. Fixed root tips were stored at
-20 ℃ with 70% alcohol until requiredꎬ then mac ̄
erated in 1 mol􀅰L-1 hydrochloric acid at 60 ℃ for 8
minꎬ stained in carbolfuchsin for 2 h and squashed
in a drop of 45% formic acid on a microscope slide.
1􀆰 2  Karyotype analysis
Asymmetry is a good morphology for the expres ̄
sion of karyotype in plants ( Peruzzi and Erogluꎬ
2013). An asymmetrical karyotype is characterised
by the the predominance of t / st chromosomes (intra ̄
chromosomal asymmetry) and highly heterogeneous
chromosome sizes ( interchromosomal asymmetry ).
Increasing asymmetry can occur either through the
shift of centromere postion from median / submedian
to terminal / subterminalꎬ or through the accumula ̄
tion of differences in relative size betweeen the chro ̄
mosomes of the complement (Paszkoꎬ 2006ꎻ Peruzzi
and Erogluꎬ 2013ꎻ Stebbinsꎬ 1971ꎻ Zuo and Yuanꎬ
2011). According to the review of Peruzzi and Ero ̄
glu (2013)ꎬ MCA index (mean centromeric asymme ̄
try) (Peruzzi and Erogluꎬ 2013) and CVCL (Coef ̄
ficent of variation of the chromosome length) (Lava ̄
nia and Srivastavaꎬ 1992ꎻ Paszkoꎬ 2006ꎻ Watanabe
et al.ꎬ 1999) were employed to describe the intrach ̄
romosomal asymmetry and interchromosomal asym ̄
metryꎬ respectively. These two values increase with
increasing asymmetry (Peruzzi and Erogluꎬ 2013).
Chromosomes were classified as metacentric
(m)ꎬ submetacentric ( sm)ꎬ subtelocentric ( st )
and telocentric (t) (Levan et al.ꎬ 1964). The fun ̄
damental number (FNꎬ or total number of chromo ̄
some arms) is an important index utilized to infer
karyotype evolution (Cox et al.ꎬ 1998). If the FN is
conservative but not in chromosome numbersꎬ it
strongly suggests a Robertsonian relationship is gen ̄
erated by the fission or fusion (Cox et al.ꎬ 1998ꎻ
Pardo ̄Manuel de Villena and Sapienzaꎬ 2001ꎻ Steb ̄
binsꎬ 1971). This measure was calculated through
the number of m / sm chromosomes with two arms and
st / t chromosomes with one arm ( Ricardo et al.ꎬ
2006ꎻ Perazzo et al.ꎬ 2011).
It is worth noting that all indices were calcula ̄
ted from the chromosome setꎬ thereforeꎬ the features
of individual chromosome cannot be shown. In order
to compare chromosome size between related taxaꎬ
Plummer et al. (2003) proposed a novel method for
comparing the chromosomes of related species using
individual chromosomes. This method was accepted
and modified by Peruzzi et al. (2009) using C ̄value
(deduced from absolute chromosome sizes) and R ̄
value (l / s) instead of New Relative Length and R ̄
value (l / s) as the X and Y axesꎬ respectively. How ̄
everꎬ the chromosomes of the same species may vary
greatly in their absolute sizes due to technical rea ̄
sons and / or differential degrees of chromosome com ̄
pactness. Thereforeꎬ relative length is more reliable
for comparisons as it standardizes the total length
(Parraguez et al.ꎬ 2009). In an attempt to know the
features of individual chromosomeꎬ two most impor ̄
tant features of chromosome ( Singhꎬ 2010ꎻ Steb ̄
binsꎬ 1971)ꎬ the relative length and R value (s / l)
are placed on the X and Y axes in the present study.
“Relative length” indicates the percentage of the
length of each chromosome to the total haploid
length. The “R value” is a ratio obtained by divid ̄
ing the length of a short arm by the length of a long
arm. The correlations of relative length and R value
9376期      YIN Gen ̄Shen et al.: Tracing the Origin of the Bimodal Karyotypes of the Tribe Lilieae (Liliaceae) 􀆺     
were disposed by Bivariate Correlations Analysis.
Pearson’s correlation coefficient was used to assess
the correlations between the different indices. Only
correlations significant at the 1% level or stronger
are discussed. Statistical analysis was carried out
using SPSS 10.
2  Results and discussion
2􀆰 1  First reported karyotypes
The chromosome numbers and karyotypes of
three species are reported here for the first time:
F􀆰 crassicaulis S􀆰 C. Chen in S􀆰 C. Chen & K􀆰 C.
Hsiaꎬ T􀆰 suaveolens Rothꎬ and G􀆰 oxycarpa (Franch.)
Zarrei & Wilkin ( Table 1ꎬ Fig􀆰 1). Because the
three species grow in very inaccessible areasꎬ it is
difficult to obtain live specimensꎬ and no karyotypic
data for these species have been previously reported.
Like other species of the Lilieaeꎬ F􀆰 crassicaulis pos ̄
sesses uniformity of the bimodal karyotype. Analysis
of somatic metaphaese showed that the karyotype for ̄
mula of it is 2n = 24 = 2m+2sm+8st+12t. The total
haploid length of chromosomes (TCL) was 154􀆰 28
μm. The analysis of the karyotype asymmetry indices
showed values of 21􀆰 21 and 65􀆰 74 for CVCL and
MCAꎬ respectively ( Table 1). Polyploidy has long
been recognized as an important role in driving plant
diversification and speciation (Stebbinsꎬ 1971ꎻ Grantꎬ
1981ꎻ Levinꎬ 2002). In this studyꎬ we found both
diploid and triploid chromosome counts in T􀆰 suaveo ̄
lens. This pattern was also observed in T􀆰 aleppensis
Boissier ex Regelꎬ T􀆰 lannata Regelꎬ T􀆰 hoogiana
Fedtschenko (Van Raanmsdonk and De Vriesꎬ 1995).
The basic chromosome sets of the triploid cytotype
in T􀆰 suaveolens show no essential differences from
those of the diploid cytotype in size and morphology.
For exampleꎬ the TCL was 95􀆰 16 μm for triploid
cytotype and 93􀆰 62 μm for diploid cytotype. Karyo ̄
type asymmetry also indicated there is little variation
between triploid and diploid cytotypeꎬ the values
were 20􀆰 27ꎬ 51􀆰 65 in the former and 19􀆰 51ꎬ 53􀆰 01
in the latter for CVCL and MCAꎬ respectively (Table
1ꎬ Fig􀆰 1Eꎬ Gꎬ Fꎬ H). This resultꎬ together with
the fact that the two cytotypes grow togetherꎬ indi ̄
cates that the triploid is likely an autopolyploid of re ̄
cent origin. To prove this hypothesisꎬ a C ̄banding
experiment is currently being conducted. Meanwhileꎬ
we found G􀆰 oxycarpa is also a triploid ( Table 1ꎬ
Fig􀆰 1Eꎬ Gꎬ Fꎬ H). Analysis of somatic metaphaese
showed that the karyotype formula of G􀆰 oxycarpa is
3n= 36= 6m+6sm+9st+15t. The total length of chro ̄
mosomes was 34􀆰 46 μm. The analysis of the karyo ̄
type asymmetry indices showed values of 39􀆰 65 and
34􀆰 46 for CVCL and MCAꎬ respectively ( Table 1).
As suggested by Esteban et al. (2009)ꎬ triploid is
very common in Gagea (up to 18%) and polyploidy
may play an important role in the evolution of this
genus (Peruzziꎬ 2012). Many individual polyploid
genotypes have novel traits that are not present in
their diploid progenitors and which are able to toler ̄
ate an extreme climate (Levinꎬ 1983ꎻ Ramsey and
Schemskeꎬ 2002). That is maybe why G􀆰 oxycarpa
lives in the alpine region up to 4 800 m (Liangꎬ
1980).
2􀆰 2  The features of individual chromosome
The karyotypes of the Lilieae are rather constant
bimodal ( Tamuraꎬ 1998)ꎬ as can be readily ob ̄
served in the scatter plot ( Fig􀆰 2A). The chromo ̄
somes of the Lilieae fell into two distinct groups: one
Table 1  Source of material and karyotype sturctures of the three species studied
Taxa Locality m sm st t CVCL MCA TCL / μm 2n
Fritillaria crassicauli S􀆰 C. Chen in S􀆰 C.
Chen & K􀆰 C. Hsia Lijiangꎬ Yunnan 1 1 4 6 21􀆰 21 65􀆰 74 154􀆰 28 24
Tulipa suaveolens Roth Tachengꎬ Xinjiang 0 3 9 0 20􀆰 27 51􀆰 65 95􀆰 16 36
T􀆰 suaveolens Roth Tachengꎬ Xinjiang 0 3 9 0 19􀆰 51 53􀆰 01 93􀆰 62 24
Gagea oxycarpa (Franch.) Zarrei & Wilkin Lijiangꎬ Yunnan 2 2 3 5 39􀆰 65 62􀆰 62 34􀆰 46 36
Abbreviations: m=medianꎻ sm=submedianꎻ st = subterminalꎻ t = terminalꎻ CVCL = Coefficent of variation of the chromosome lengthꎻ MCA =mean
centromeric asymmetryꎻ TCL= total chromosome length
047                                  植 物 分 类 与 资 源 学 报                            第 36卷
is composed of long metacentric / submetacentric chro ̄
mosomesꎬ and the other is composed of short subte ̄
locentric / telocentric chromosomes. Howeverꎬ the kary ̄
otypes of the Tulipeaeꎬ and the Medeoloideae exhibit
a continuous range of chromosome lengths and R val ̄
ues (Fig􀆰 2Bꎬ C).
Fig􀆰 1  Karyograms of the three species
Aꎬ C. Fritillaria crassicaulisꎻ Bꎬ D. Tulipa suaveolensꎻ Eꎬ G. T􀆰 suaveolensꎻ Fꎬ H. Gagea oxycarpa. Scale bars= 10 μm
1476期      YIN Gen ̄Shen et al.: Tracing the Origin of the Bimodal Karyotypes of the Tribe Lilieae (Liliaceae) 􀆺     
Fig􀆰 2  Analysis of karyotype morphology in Lilioideae and
Medeoloideae. R value and relative length were plotted
for each chromosome studied
A. Longer chromosomes and shorter chromosomes of tribe Lilieaeꎻ
B. Chromosomes of tribe Tulipeaeꎻ C. Chromosomes of Medeoloideae
    Possible outliers were determined and removed
based on the Pauta criterionꎬ and the model was re ̄
calculated after the excision of the outliers of the
original dataset. In totalꎬ 6 chromosomes were not
used in the statistical calculations. A good positive
correlation was detected between relative length and
R value across the Lilioidaceae (r=0􀆰 53ꎬ removing 2
outliers)ꎬ indicating that higher R values are associ ̄
ated with increasingly metacentric chromosomes. This
goodpositive correlation was also found for various
subgroups within the Lilioideaeꎬ including the Lilie ̄
ae (r=0􀆰 77ꎬ removing 2 outliers)ꎬ Lilium (r=0􀆰 79)ꎬ
Nomocharis ( r= 0􀆰 80)ꎬ Fritillaria ( r= 0􀆰 71ꎬ remo ̄
ving 2 outliers)ꎬ Cardiocrinum ( r = 0􀆰 85)ꎬ and
Notholirion ( r = 0􀆰 75). Howeverꎬ the correlation
was negative for the Tulipeae ( r = -0􀆰 40ꎬ removing
2 outliers)ꎬ Tulipa ( r= -0􀆰 35)ꎬ and Erythronium
( r= -0􀆰 68ꎬ removing 2 outliers) . These results sug ̄
gest that the chromosomes of the Lilieaeꎬ and the
Tulipeae have taken different evolutionary paths.
Thereforeꎬ it is inappropriate to combine these taxa
as a whole and draw a conclusion for the karyotype
evolution of the entire Lilioidaceae.
It is noteworthy that the correlation was also
positive for the Medeoloideae ( r = 0􀆰 84ꎬ removing 2
outliers)ꎬ Medeola ( r= 0􀆰 94ꎬ removing 2 outliers)ꎬ
and Clintonia ( r = 0􀆰 83). In the Lilieaeꎬ a strong
positive correlation was also observed with the longer
m / sm chromosomes (chromosome pairs I and IIꎬ r=
0􀆰 61)ꎻ howeverꎬ the shorter st / t chromosomes (chro ̄
mosome pairs III to XII) had a low negative correla ̄
tion ( r= -0􀆰 13). These results were consistent with
the conclusion of Peruzzi et al. (2009)ꎬ in which
metacentric to submetacentric and subtelocentric to
telocentric chromosomes undergo different patterns of
karyotype evolution.
The longer chromosomes (I and II) of the Li ̄
lieae and the chromosomes of the Medeoloideae pres ̄
ented an obvious similarity in correlationꎬ which may
provide clues for tracing the karyotype evolutionary
lines of the Lilieae. It is reasonable to treat the cor ̄
relation between the relative length and R value as a
character. If the character has only two statesꎬ the
state which is in the outgroup is primitive (Heren ̄
deen and Millerꎬ 2000ꎻ Lipscombꎬ 1998). Thusꎬ it
is easily to distinguish positive correlation as primi ̄
tive character states because these states were detec ̄
ted within outgroupꎬ Medeoloideae ( r= 0􀆰 84ꎬ remo ̄
247                                  植 物 分 类 与 资 源 学 报                            第 36卷
ving 2 outliers) . The shorter chromosomes ( III to
XII) of Lilieaeꎬ possessed a negative correlation and
were regarded as evolutionarily advanced. Thereforeꎬ
some evolution has occurred in the shorter chromo ̄
somes of Lilieae to a more extent. Similarlyꎬ other
changes occurred as well as in tribe Tulipeae and
these changes should be detected by other methods.
2􀆰 3  Outliers
It is important to realise that not all outliers are
illegitimate contaminants of the data (Osborne and
Overbayꎬ 2004). If outliers do not arise from data
errors or faulty distributional assumptionsꎬ they can
also be clues for inquiry. The outliers in Erythronium
are the two biggest chromosomal pairs of E􀆰 sibiricum
(Fisch. & C􀆰 A. Mey.) Krylov. Compared with the
karyotype of the other species in the tribe Tulipeae
which consists primarily of 12 telocentric / sub ̄telo ̄
centric chromosome pairs (Fig􀆰 3C)ꎬ the Erythroni ̄
um karyotype consists of 2 metacentric and 8 telo ̄
centric / subtelocentric chromosome pairs (Fig􀆰 3D).
Based on the results of fluorescence in situ hybridi ̄
sation using 5S rDNA and telomere sequences as
probesꎬ one possible explanation for these outliers is
Robertsonian fusion involving the loss of 4 telocentric
chromosome pairs to gain 2 metacentric chromosomes
(Yin et al.ꎬ unpublished data). Robertsonian rear ̄
rangements ( chromosome fission and fusion) have
also been observed in Fritillaria ( Courꎬ 1978ꎻ Li
and Shangꎬ 1989ꎻ Zaharofꎬ 1989) and Cardiocri ̄
num (Chauhan and Brandhamꎬ 1985ꎻ Zhu et al.ꎬ
2002). In this workꎬ the similar phenomenon ob ̄
served in the first chromosome pair ( Fig􀆰 3B) of
F􀆰 usuriensis Maxim.ꎬ manifested an unusually long
metacentric chromosome ( the relative length of the
first chromosome pair is 15􀆰 53ꎬ but the mean value
is 12􀆰 58 in Fritillaria) and the missing of one sub ̄
telocentric / telocentric chromosome pair. Howeverꎬ a
different explanation is required for the outliers in
M􀆰 virginiana L. These outliers are the sixth and sev ̄
enth chromosome pairs (Fig􀆰 3F)ꎬ which should be
subtelocentric / telocentric according to the regression
model (not shown) instead of metacentric (Fig􀆰 3E).
Reciprocal translocations and / or pericentric inver ̄
sion may be important in M􀆰 virginianaꎻ these mech ̄
anisms keep the chromosome number stable while in ̄
troducing variations among karyotypes (Schubert and
Riegerꎬ 1985).
Fig􀆰 3  Comparisons between the species contained the outlier
chromosomes and the mean haploid idiograms
of the corresponding tribe / family
A. Mean haploid idiograms of Lilieaeꎻ B. Ideograms of somatic meta ̄
phase of Fritillaria usuriensisꎻ C. Mean haploid idiograms of Tulipeaeꎻ
D. Ideograms of somatic metaphase of Erytronium sibiricumꎻ
E. Mean haploid idiograms of Medeoloideaeꎻ F. Mean
haploid idiograms of Medeola virginiana
2􀆰 4  Fundamental number
Presentlyꎬ there are two feasible hypotheses
( i􀆰 e. centric fission and polyploidisation) accoun ̄
ting for the bimodal karyotype evolution of the Lilie ̄
ae. As suggested by Tobias (1953)ꎬ counting the
number of chromosome arms helps to distinguish be ̄
tween fission events and other karyotype changes.
The total number of chromosome arms remains basi ̄
cally the same during the centric fission processꎬ
whereas the total count may either increase or de ̄
crease due to other karyotypic changes (Cox et al.ꎬ
1998ꎻ Pardo ̄Manuel de Villena and Sapienzaꎬ
2001ꎻ Stebbinsꎬ 1971). Accordinglyꎬ if centric fis ̄
sion alone is responsible for karyotype evolution in
3476期      YIN Gen ̄Shen et al.: Tracing the Origin of the Bimodal Karyotypes of the Tribe Lilieae (Liliaceae) 􀆺     
the Lilioideaeꎬ the number of chromosome arms
should remain unchanged as the chromosome number
increases from x = 7 to x = 12. For exampleꎬ as the
mean arm number of the Medeoloideae is 13􀆰 13
(Table 2)ꎬ those species of the Lilioideae with 2n=
24 would also be expected to have a mean arm num ̄
ber approaching 13􀆰 13. The actual mean NF of the
Lilieae is 14􀆰 21ꎬ whereas the NF of the Tulipeae is
17􀆰 11 (Table 2). The mean NF of the Lilieae is clos ̄
er to that of the Medeoloideaeꎬ indicating that the
karyotypes of the Lilieae may have resulted from cen ̄
tric fissionꎬ but not exclusivelyꎬ while other mecha ̄
nisms were responsible for the large differences of
mean NF between the Medeoloideae and the Tulipeae.
2􀆰 5  The relative chromosomes length
Similarlyꎬ relative chromosome length also sup ̄
ports the hypothesis of centric fission. If centric fis ̄
sion occurred in the two metacentric / submetacentric
chromosome pairs of a Medeoloideae—like ancestorꎬ
the proportion of the ten subtelocentric / telocentric
chromosome pairs in the chromosome set of the Lilie ̄
ae should be similar to the five pairs of probable
chromosomes of the Medeoloideae. The mean propor ̄
tion of the shorter chromosome pairs in the Lilieae is
76􀆰 39ꎬ while this value is 77􀆰 89 for the five meta ̄
centric / submetacentric chromosome pairs in the
Medeoloideae (Table 2). This value was not calcu ̄
lated in the Tulipeae due to the impossibility of iden ̄
tifying longer and shorter homologous chromosomes.
2􀆰 6  Intrachromosome asymmetry
The centric fission of one metacentric or sub ̄
metacentric chromosome results in an increase in
chromosome number. This process in turn leads to an
increase in the intrachromosome asymmetry of the
karyotype. In the Medeoloideae controlꎬ MCA is 28􀆰 84.
Converselyꎬ in the Lilioideaeꎬ intrachromosome a ̄
symmetry is clearly on the rise: in the Lilieaeꎬ MCA
is 65􀆰 63. An analogous situation occurred in the Tu ̄
lipeaꎬ the value was 54􀆰 82 for MCA .
3  Conclusions
On the basis of these resultsꎬ the origin of the
extreme asymmetry and bimodality of the Lilieae is
best explained by centric fission as mentioned by
Tamura (1995). Howeverꎬ centric fission is not ex ̄
pected to occur in isolationꎬ and the different mech ̄
anisms of karyotype change are not mutually exclu ̄
sive. The slightly different NF values and relative
lengths of the probable chromosomes (Table 2)ꎬ a ̄
long with the subtelocentric rather than telocentric
chromosomesꎬ suggested that other mechanisms con ̄
tributed as well. Evidence for these mechanisms has
been reported in the variation of C ̄banding patterns
(Smyth et al.ꎬ 1989)ꎬ fluorochrome staining (Am ̄
brozova et al.ꎬ 2011ꎻ Muratovic et al.ꎬ 2010) in
Lilium and Fritillaria. The karyotypic evolution of
the Tulipeae remains elusive. Because the extent of
the changes to the karyotypes has led to no apparent
similarity between the Tulipeae and the Mede ̄
oloideaeꎬ it is difficult to determine the evolutionary
history. Pericentric inversion and segment transloca ̄
tion may have played major rolesꎬ as these changes
Table 2  Mean karyotype values using data of Lilioideae and Medeoloideae
Chromosomes Relative length NF CVCL MCA
whole 100    14􀆰 21 23􀆰 12 65􀆰 63
Lilieae longer 23􀆰 61 3􀆰 97 8􀆰 60 20􀆰 11
shorter 76􀆰 39 10􀆰 24 13􀆰 72 73􀆰 54
Tulipeae whole 100    17􀆰 11 27􀆰 94 54􀆰 82
whole 100    13􀆰 13 25􀆰 23 28􀆰 84
Medeoloideae VI ̄VII 22􀆰 11 3􀆰 25 5􀆰 83 23􀆰 15
I ̄V 77􀆰 89 9􀆰 88 22􀆰 66 41􀆰 65
Abbreviations: NF=Nombre Fondamentalꎻ CVCL =Coefficent of variation of the chromosome lengthꎻ MCA =mean centromeric asymmetry
447                                  植 物 分 类 与 资 源 学 报                            第 36卷
keep the chromosome number constant while allo ̄
wing variations among the karyotypes (Schubert and
Riegerꎬ 1985).
This research is focused on karyomorphological
characteristics. Clearlyꎬ efforts need to be undertak ̄
en to obtain data of molecular cytogeneticsꎬ by the
fluorescence in situ hybridisation technique and / or
chromosomal mappingꎬ to provide more direct evi ̄
dence. These experiments are in progress.
Acknowledgments: We thank the anonymous reviewers for
their valuable comments and suggestionsꎻ Dr. Zhao Yujuan
for constructive suggestions on improving the manuscriptꎻ Yao
Yao (Kunming University of Science and Technology) for his
assistance with field sampling.
References:
梁松筠 (Liang SJ)ꎬ 1980. 中国植物志 第 14 卷 [M]. 北京: 科
学出版社ꎬ 81—84
中国药典编辑委员会 (Chinese Pharmacopoeia Commission)ꎬ 2010.
中华人民共和国药典 [M]. 北京: 中国医药科技出版社
Ambrozova Kꎬ Mandakova Tꎬ Bures P et al.ꎬ 2011. Diverse retro ̄
transposon families and an AT ̄rich satellite DNA revealed in gi ̄
ant genomes of Fritillaria lilies [ J] . Annals of Botanyꎬ 107
(2): 255—268
Bennett STꎬ Kenton AYꎬ Bennett MDꎬ 1992. Genomic in situ hybrid ̄
ization reveals the allopolyploid nature of Milium montianum
(Gramineae) [J] . Chromosomaꎬ 101 (7): 420—424
Cave MSꎬ 1970. Chromosomes of the California Liliaceae [ J] . Uni ̄
versity of California Publications in Botanyꎬ 57: 1—48
Chauhan Kꎬ Brandham Pꎬ 1985. The significance of Robertsonian fu ̄
sion and monosomy in Cardiocrinum (Liliaceae) [ J] . Kew Bul ̄
letinꎬ 40 (3): 567—571
Cour Lꎬ 1978. Two types of constitutive heterochromatin in the chro ̄
mosomes of some Fritillaria species [J] . Chromosomaꎬ 67 (1):
67—75
Cox AVꎬ Abdelnour GJꎬ Bennett MD et al.ꎬ 1998. Genome size and
karyotype evolution in the slipper orchids (Cypripedioideae: Or ̄
chidaceae) [ J] . American Journal of Botanyꎬ 85 (5): 681—
687
Esteban MGꎬ Chaparro JLLꎬ Pardo FMVꎬ 2009. Revisión bibliográfica
de las aportaciones sobre recuentos cromosómicos en el género Ga ̄
gea salibs ( Liliaceae) [ J] . Folia Botanica Extremadurensisꎬ 4
(1): 45—57
Gao YDꎬ Zhou SDꎬ He XJ et al.ꎬ 2012. Chromosome diversity and e ̄
volution in tribe Lilieae ( Liliaceae) with emphasis on Chinese
species [J] . Journal of Plant Researchꎬ 125 (1): 55—69
Gitai Jꎬ Horres Rꎬ Benko ̄Iseppon AMꎬ 2005. Chromosomal features
and evolution of Bromeliaceae [J] . Plant Systematics and Evolu ̄
tionꎬ 253 (1 ̄4): 65—80
Goldblatt Pꎬ Takei Mꎬ 1997. Chromosome cytology of Iridaceae—Pat ̄
terns of variationꎬ determination of ancestral base numbersꎬ and
modes of karyotype change [J] . Annals of The Missouri Botanical
Gardenꎬ 84 (2): 285—304
Grant Vꎬ 1981. Plant Speciation [M]. New York: Columbia Univer ̄
sity Press
Herendeen PSꎬ Miller RBꎬ 2000. Utility of wood anatomical charac ̄
ters in cladistic analyses [J] . Iawa Journalꎬ 21 (3): 247—276
Jones Kꎬ 1998. Robertsonian fusion and centric fission in karyotype evolu ̄
tion of higher plants [J]. Botanical Reviewꎬ 64 (3): 273—289
Lavania Uꎬ Srivastava Sꎬ 1992. A simple parameter of dispersion in ̄
dex that serves as an adjunct to karyotype asymmetry [ J] . Jour ̄
nal of Biosciencesꎬ 17 (2): 179—182
Levan Aꎬ Fredga Kꎬ Sandberg AAꎬ 1964. Nomenclature for centro ̄
meric position on chromosomes [J] . Hereditasꎬ 52 (2): 201—
220
Levin DAꎬ 1983. Polyploidy and novelty in flowering plants [ J] .
American Naturalistꎬ 122 (1): 1—25
Levin DAꎬ 2002. The Role of Chromosomal Change in Plant Evolution
[M]. New York: Oxford University Press
Lipscomb Dꎬ 1998. Basics of cladistic analysis [EB / OL]. Washing ̄
ton DC: George Washington University
Li RJꎬ Shang ZYꎬ 1989. The chromosome observation on 5 species of
rare plants of China [J] . Journal of Wuhan Botanical Researchꎬ
7 (3): 217—221
Mckain MRꎬ Wickett Nꎬ Zhang Y et al.ꎬ 2012. Phylogenomic analy ̄
sis of transcriptome data elucidates co ̄occurrence of a pal ̄
eopolyploid event and the origin of bimodal karyotypes in Aga ̄
voideae (Asparagaceae) [ J] . American Journal of Botanyꎬ 99
(2): 397—406
Muratovic Eꎬ Robin Oꎬ Bogunic F et al.ꎬ 2010. Karyotype evolution
and speciation of European lilies from Lilium sect. Liriotypus
[J] . Taxonꎬ 59 (1): 165—175
Osborne JWꎬ Overbay Aꎬ 2004. The power of outliers ( and why re ̄
searchers should always check for them) [ J / OL]. Practical As ̄
sessmentꎬ Research and Evaluationꎬ 9: 1—12
Pardo ̄Manuel de Villena Fꎬ Sapienza Cꎬ 2001. Recombination is pro ̄
portional to the number of chromosome arms in mammals [ J] .
Mammalian Genomeꎬ 12 (4): 318—322
Parraguez Mꎬ Gajardo Gꎬ Beardmore JAꎬ 2009. The new world ar ̄
temia species A􀆰 franciscana and A􀆰 persimilis are highly differenti ̄
ated for chromosome size and heterochromatin content [J] . Here ̄
ditasꎬ 146 (2): 93—103
Paszko Bꎬ 2006. A critical review and a new proposal of karyotype a ̄
symmetry indices [ J] . Plant Systematics and Evolutionꎬ 258
(1): 39—48
Perazzo Gꎬ Noleto RBꎬ Vicari MR et al.ꎬ 2011. Chromosomal studies
5476期      YIN Gen ̄Shen et al.: Tracing the Origin of the Bimodal Karyotypes of the Tribe Lilieae (Liliaceae) 􀆺     
in Crenicichla lepidota and Australoheros facetus (Cichlidaeꎬ Per ̄
ciformes) from extreme Southern Brazil [J] . Reviews in Fish Bi ̄
ology and Fisheriesꎬ 21 (3): 509—515
Peruzzi Lꎬ 2012. Chromosome diversity and evolution in the genus Ga ̄
gea (Liliaceae) [J] . Bocconeaꎬ 24: 147—158
Peruzzi Lꎬ Eroglu Hꎬ 2013. Karyotype asymmetry: againꎬ how to
measure and what to measure? [J] . Comparative Cytogeneticsꎬ 7
(1): 1—9
Peruzzi Lꎬ Leitch IJꎬ Caparelli KFꎬ 2009. Chromosome diversity and
evolution in Liliaceae [ J] . Annals of Botanyꎬ 103 (3): 459—
475
Pires JCꎬ Maureira IJꎬ Givnish TJ et al.ꎬ 2006. Phylogenyꎬ genome
sizeꎬ and chromosome evolution of Asparagales [ J] . Alisoꎬ 22:
287—304
Plummer JAꎬ Shan Fꎬ Galwey N et al.ꎬ 2003. New methods for com ̄
parison of chromosomes within and between species [J] . Caryo ̄
logiaꎬ 56 (2): 227—231
Ramsey Jꎬ Schemske DWꎬ 2002. Neopolyploidy in flowering plants
[J] . Annual Review of Ecology and Systematicsꎬ 33: 589—639
Ricardo VMꎬ Artoni RFꎬ Moreira ̄Filho O et al.ꎬ 2006. Basic and mo ̄
lecular cytogenetics in fresh ̄water Cichlidae (Osteichthyesꎬ Per ̄
ciformes) . Karyotypic conservationism and divergence [ J ] .
Caryologiaꎬ 59 (3): 260—266
Sato Dꎬ 1942. Karyotype alteration and phylogeny in Liliaceae and al ̄
lied families [J] . Journal of Japanese Botanyꎬ 12: 57—161
Schubert Iꎬ Lysak MAꎬ 2011. Interpretation of karyotype evolution
should consider chromosome structural constraints [J] . Trends in
Geneticsꎬ 27 (6): 207—216
Schubert Iꎬ Rieger Rꎬ 1985. A new mechanism for altering chromo ̄
some number during karyotype evolution [ J] . Theoretical and
Applied Geneticsꎬ 70 (2): 213—221
Sharma Aꎬ Sen Sꎬ 2002. Chromosome Botany [M]. Enfield: Science
publishers
Singh RJꎬ 2010. Plant Cytogenetics [M]. Boca Raton: CRC press
Smyth Dꎬ Kongsuwan Kꎬ Wisudharomn Sꎬ 1989. A survey of C ̄band
patterns in chromosomes of Lilium (Liliaceae) [ J] . Plant Sys ̄
tematics and Evolutionꎬ 163 (1): 53—69
Souza Eꎬ Ribeiro Lꎬ Feldberg E et al.ꎬ 2011. Comparative cytogenet ̄
ics of two of the smallest Amazonian fishes: Fluviphylax simplex
Costaꎬ 1996 and Fluviphylax zonatus Costaꎬ 1996 (Cyprinodon ̄
tiformesꎬ Poeciliidae) [ J] . Comparative Cytogeneticsꎬ 5 ( 5):
411—422
Stebbins GLꎬ 1971. Chromosomal Evolution in Higher Plants [M].
London: Edward Arnold
Stedje Bꎬ 1989. Chromosome evolution within the Ornithogalum
tenuifolium complex (Hyacinthaceae)ꎬ with special emphasis on
the evolution of bimodal karyotypes [ J] . Plant Systematics and
Evolutionꎬ 166 (1 ̄2): 79—89
Stewart RNꎬ 1947. The morphology of somatic chromosomes in Lilium
[J] . American Journal of Botanyꎬ 34 (1): 9—26
Takhtajan Aꎬ 2009. Flowering Plants [M]. New York: Springer
Tamura MNꎬ 1995. A karyological review of the orders Asparagales
and Liliales (Monocotyledonae) [ J] . Feddes Repertoriumꎬ 106
(1 ̄2): 83—111
Tamura MNꎬ 1998. Liliaceae [A]. In: Kubitzki K (ed.)ꎬ The Fam ̄
ilies and Genera of Vascular Plants [ M]. Berlin: Springerꎬ
343—353
Tobias Pꎬ 1953. Trends in the evolution of mammalian chromosomes
[J] . South African Journal of Scienceꎬ 50: 134—140
Van Raamsdonk Lꎬ De Vries Tꎬ 1995. Species relationships and tax ̄
onomy in Tulipa subg. Tulipa (Liliaceae) [J] . Plant Systematics
and Evolutionꎬ 195 (1 ̄2): 13—44
Vosa CGꎬ 2005. On chromosome uniformityꎬ bimodality and evolution
in the tribe Aloineae ( Asphodelaceae) [ J] . Caryologiaꎬ 58
(1): 83—85
Watanabe Kꎬ Yahara Tꎬ Denda T et al.ꎬ 1999. Chromosomal evolu ̄
tion in the genus Brachyscome (Asteraceaeꎬ Astereae): statisti ̄
cal tests regarding correlation between changes in karyotype and
habit using phylogenetic information [ J] . Journal of Plant Re ̄
searchꎬ 112 (2): 145—161
Woodcock HBDꎬ Stearn WTꎬ 1950. Lilies of the World [M]. Lon ̄
don: Country Life
Zaharof Eꎬ 1989. Karyotype variation of Fritillaria graeca and F􀆰 davisii
from Greece [J] . Nordic Journal of Botanyꎬ 9 (4): 367—373
Zhu XNꎬ Jin XFꎬ Gao Zꎬ 2002. Karyotype analysis on three species of
Liliaceae [J] . Journal Zhejiang Forest Science & Technologyꎬ 22
(2): 22—25
Zuo Lꎬ Yuan Qꎬ 2011. The difference between the heterogeneity of
the centromeric index and intrachromosomal asymmetry [ J ] .
Plant Systematics and Evolutionꎬ 297 (1 ̄2): 141—145
647                                  植 物 分 类 与 资 源 学 报                            第 36卷