Gene duplication and an accelerated evolutionary rate in 11S globulin genes are associated with higher protein synthesis in dicots as compared to monocots.

Abstract

Seed storage proteins are a major source of dietary protein, and the content of such proteins determines both the quantity and quality of crop yield. Significantly, examination of the protein content in the seeds of crop plants shows a distinct difference between monocots and dicots. Thus, it is expected that there are different evolutionary patterns in the genes underlying protein synthesis in the seeds of these two groups of plants.
Gene duplication, evolutionary rate and positive selection of a major gene family of seed storage proteins (the 11S globulin genes), were compared in dicots and monocots. The results, obtained from five species in each group, show more gene duplications, a higher evolutionary rate and positive selections of this gene family in dicots, which are rich in 11S globulins, but not in the monocots.
Our findings provide evidence to support the suggestion that gene duplication and an accelerated evolutionary rate may be associated with higher protein synthesis in dicots as compared to monocots.

Full text

Page 1

RESEARCH ARTICLEOpen Access
Gene duplication and an accelerated evolutionary
rate in 11S globulin genes are associated with
higher protein synthesis in dicots as compared to
monocots
Chun Li1,2†, Meng Li1†, Jim M Dunwell3and Yuan-Ming Zhang1*
Abstract
Background: Seed storage proteins are a major source of dietary protein, and the content of such proteins
determines both the quantity and quality of crop yield. Significantly, examination of the protein content in the
seeds of crop plants shows a distinct difference between monocots and dicots. Thus, it is expected that there are
different evolutionary patterns in the genes underlying protein synthesis in the seeds of these two groups of
plants.
Results: Gene duplication, evolutionary rate and positive selection of a major gene family of seed storage proteins
(the 11S globulin genes), were compared in dicots and monocots. The results, obtained from five species in each
group, show more gene duplications, a higher evolutionary rate and positive selections of this gene family in
dicots, which are rich in 11S globulins, but not in the monocots.
Conclusion: Our findings provide evidence to support the suggestion that gene duplication and an accelerated
evolutionary rate may be associated with higher protein synthesis in dicots as compared to monocots.
Keywords: 11S globulin, dicot, evolutionary rate, gene duplication, legumins, monocot, positive selection
Background
The plant seed is not only an organ of propagation and
dispersal but also the major plant tissue harvested and
used either directly as part of the human diet or as feed
for animals. At the present time there is concern over
long term food security and the impact of the move
towards meat-based diets that will lead to a significant
increase in the demand for plant protein for animal feed
[1]. The amount of protein present in plant seeds varies
from ~10% of the dry weight in most monocot (e.g.
O. sativa, S. bicolor, S. italica, Z. mays and B. distachyon)
to more than 30% in most dicots (e.g. G. max, R. commu-
nis, C. sativus and A. thaliana), and forms a major source
of dietary protein [2-7]. To determine whether
differences in evolutionary patterns may explain the phe-
notypic differences observed, a comparative investigation
of evolutionary divergence in genes underlying protein
synthesis in these two groups of plants is thus warranted.
Seed storage proteins can be classified into four groups:
albumins, globulins, prolamins and glutelins [8]. Albu-
mins and globulins comprise the storage proteins of
dicots, whereas prolamins and glutelins are the major
proteins in monocots [4,9,10]. 2S albumins, a major class
of dicot seed storage proteins, have been most widely stu-
died in the Cruciferae, notably B. napus and A. thaliana
[9,10]. Prolamins, the major endosperm storage proteins
of all cereal grains, with the exceptions of oats and rice,
can be classified into many subgroups, e.g. sulphur-rich
(S-rich), sulphur-poor (S-poor) and high molecular
weight (HMW) prolamins [4]. The globulins, the most
widely distributed group of storage proteins, are part of
the cupin superfamily [11] and are evolved from bacterial
enzymes. The globulins are present not only in dicots but
* Correspondence: soyzhang@njau.edu.cn
† Contributed equally
1State Key Laboratory of Crop Genetics and Germplasm Enhancement,
College of Agriculture, Nanjing Agricultural University, Nanjing 210095, P R
China
Full list of author information is available at the end of the article
Li et al. BMC Evolutionary Biology 2012, 12:15
http://www.biomedcentral.com/1471-2148/12/15
© 2012 Li et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.

Page 2

also in monocots [9] and can be divided into 7S vicilin-
type and 11S legumin-type globulins according to their
sedimentation coefficients. It should be noted that the
genes encoding the 11S-type globulins in monocots are
the same gene family, 11S globulin family, as those in
dicots; whereas the genes encoding the 7S-type globulins
in monocots are not evolutionarily related to those in
dicots [4]. Thus, we focus on the genes encoding the
11S-type globulins in this study.
Many efforts have been made to describe the gene
families encoding seed storage proteins. For albumins,
they are encoded by multi-gene families in many dicots
(e.g., A. thaliana and B. napus); and evolutionary
research into the gene families suggests that the albumin
genes were duplicated prior to the Brassiceae-Sysim-
brieae split, and gene duplication has played a role in
their evolution [12,13]. For prolamins, they are the major
seed storage proteins in most grass species (e.g., Z. mays
and S. bicolor); and studies have suggested that the prola-
min gene families have undergone many rounds of gene
duplication [14,15]. For globulins, eight, four, eleven and
fourteen gene have been identified and classified in
G. max, A. thaliana, R. communis and O. sativa, respec-
tively [16-24].
From the research described above, it can be concluded
that the seed storage protein gene families have expanded
in a lineage-specific manner through gene duplication. As
a major process in the evolution, gene duplications can
provide raw material for evolution by producing new
copies. The human globin gene family is a representative
example: several globin genes have arisen from a single
ancestral precursor, thus making individual genes avail-
able to take on specialized roles, with some genes becom-
ing active during embryonic and fetal development, and
others becoming active in the adult organism [25]. Gene
duplications may also affect phenotype by altering gene
dosage: the amount of protein synthesized is often pro-
portional to the number of gene copies present, so extra
genes can lead to excess proteins. This applies to many
kinds of genes, such as rRNAs, tRNA and histones
[26,27]. A critical question thus can be asked: does gene
duplication contribute to the higher levels of protein
synthesis in dicots than in monocots? On the other hand,
gene duplication usually brings variation in evolutionary
rate [26], and such variation has been predicted to be
associated with phenotypic differences. For example,
Hunt et al. [28] investigated the evolution of genes asso-
ciated with phenotypically plastic castes, sexes, and devel-
opmental stages of the fire ant Solenopsis invicta, and
argued that an elevated rate is a precursor to the evolu-
tion of phenotypic differences. Thus, we are also inter-
ested in another question: does an accelerated
evolutionary rate play a role in the evolution of storage
protein content?
To shed light on the two questions above, we investi-
gated the process of molecular evolution of the 11S globu-
lin gene family, which is widely distributed in dicot and
monocot species, by comparing the differing evolutionary
patterns in the two groups. Our analyses suggested that
gene duplication and an accelerated evolutionary rate in
11S globulin genes may be associated with higher protein
synthesis in dicots than in monocots.
Results
Sequences retrieval and phylogenetic analysis
We collect the sequences of 11S globulin genes through a
COG method. This procedure is based a simple notion
that, if any proteins from distant genomes are more simi-
lar to each other than to any other proteins from the
same genomes, they are most likely to belong to an
orthologous family. In such a family, there are two kinds
of relationship between a pair of sequences, namely sym-
metrical and asymmetrical BeTs (the Best Hits). If the
symmetrical and asymmetrical BeTs are linked respec-
tively by solid and broken lines, the orthologous family
would forms a network; and thus all other members in
the network can be identified when one member was
investigated [29]. The 11S globulin genes from the five
dicots and the five monocots form a COG containing 56
sequences (Figure 1). Of these genes, four genes encoding
12S globulins [20] come from A. thaliana; six genes,
Gy1-Gy5 and Gy7 [16], come from G. max; twelve genes,
all the functional glutelin genes, come from O. sativa
[24]; eleven genes described in [7] come from R. commu-
nis; and six, seven, six, one, one and two genes come
from C. sativus, P. trichocarpa, B. distachyon, Z. mays,
S. bicolor and S. italica, respectively.
In order to analyze the phylogenetic relationship of the
11S globulin genes from the above ten species, the NJ and
Bayesian methods were used to reconstruct the phyloge-
netic tree. Similar results were achieved (data not shown),
and the NJ tree is shown in Figure 2, in which there are
three subfamily clades: dicot subfamily 1, dicot subfamily
2 and monocot subfamily. The dicot subfamily 1 contains
all the genes from A. thaliana and G. max, two genes
from C. sativus, two genes from P. trichocarpa and
RcLEG1 of R. communis; the dicot subfamily 2 contains
the other genes from C. sativus, P. trichocarpa and R. com-
munis; and the monocot subfamily contains all the genes
from the five monocot species. In the three subfamilies,
with the exception of rice, genes from the same species
form monophyletic groups; and the phylogenetic relation-
ship of the monophyletic groups is largely concordant
with the species tree described by [30].
Gene duplication in 11S globulin gene family
In the 11S globulin gene family, there are four or more
genes from C. sativus, P. trichocarpa, R. communis,
Li et al. BMC Evolutionary Biology 2012, 12:15
http://www.biomedcentral.com/1471-2148/12/15
Page 2 of 9

Page 3

A. thaliana, G. max, B. distachyon and O. sativa, sug-
gesting that the 11S globulin genes in these species have
undergone two or more rounds of duplication. Among
these duplications, tandem duplication is a major type,
and occurred in all the species above, e.g. Rc30005.
m001289-Rc30005.m001290 and At1g03880-At1g03890
and Os02g16820-Os02g16830. Furthermore, most of
the 11S globulin genes from the dicot species are
located on the chromosome segments that share con-
served synteny with each other [18], suggesting that
segment duplication or whole genome duplication
(WGD) is the major origin of duplicate genes.
Evolutionary rate of 11S globulin gene family
To determine whether there were different evolutionary
patterns across different subfamilies of the 11S globulin
genes in monocots and dicots, the ω values for these
genes were calculated by a branch-specific model.
In the branch-specific model, three cases were consid-
ered in this study. One is a two-ratio model that
Figure 1 The COG of 11S globulin gene family. Solid lines show symmetrical BeTs (the Best Hits) and broken lines show asymmetrical BeTs.
Genes from the same species are adjacent. Gene ID is indicated and the prefix “Rc” denotes IDs from Ricinus communis. Among these IDs,
At1g03880, At1g03890, At4g28520 and At5g44120 are known to encode CRB, CRU2, CRC and CRA1, respectively; Glyma03g32030,
Glyma03g32020, Glyma19g34780, Glyma10g04280, Glyma13g18450 and Glyma19g34770 to encode Gy1-5 and Gy7, respectively; Rc29600.
m000561, Rc29600.m000564, Rc30005.m001289, Rc30005.m001290, Rc29611.m000223, Rc29200.m000169, Rc29629.m001355, Rc29709.m001187,
Rc29716.m000305, Rc29200.m000167 and Rc30005.m001288 to encode RcLEG1-1 to RcLEG1-5 and RcLEG2-1 to RcLEG2-6, respectively; and
Os01g55690, Os10g26060, Os03g31360, Os02g15169, Os02g15178, Os02g15150, Os02g16820, Os02g16830, Os02g14600, Os02g15070,
Os02g25640 and Os02g15090 to encode GluA-1, GluA-2, GluA-3, GluB-1a, GluB-1b, GluB-2, GluB-5, GluB-4, GluB-7, GluB-6, GluC-1 and GluD,
respectively.
Li et al. BMC Evolutionary Biology 2012, 12:15
http://www.biomedcentral.com/1471-2148/12/15
Page 3 of 9

Page 4

suggests there are distinct monocot (ω0) and dicot sub-
families (ω1), one is a three-ratio model that suggests
there is one monocot (ω0) and two dicot subfamilies 1
(ω1) and 2 (ω2), and one is a six-ratio model that sug-
gests there are subclades under the branches A-E (ω1~
ω5) and the other (ω0). All the above models were
favored over the one-ratio model by the likelihood ratio
test (P < 0.05, Table 1) and in the two-, three- and six-
ratio models, the ω estimate for the dicot subfamily is
considerably higher than that for the monocot subfam-
ily. These results suggest that the dicot 11S globulin
genes are under reduced evolutionary constraints, and
thus evolve at a higher evolutionary rate. We should
point out that in the six-ratio model, the genes from
S. italica, Z. mays and S. bicolor were not taken into
account, because they may have a different evolutionary
pattern due to the fact that 11S globulins are minor
components [4,31] and there are only one or two 11S
Figure 2 Phylogenetic relationships of sequences within the 11S globulin gene family by neighbor joining (NJ) method with bootstrap
support above 50% shown at the nodes. Letter A-E indicates the branches used in analysis of evolutionary rate and positive selection.
Li et al. BMC Evolutionary Biology 2012, 12:15
http://www.biomedcentral.com/1471-2148/12/15
Page 4 of 9

Page 5

globulin gene copies in these species; and that branches
A-E were chosen because five branches can be grouped
for the phylogenetic tree.
Positive selection in the 11S globulin genes
To test for positive selection in the 11S globulin gene
family, branches A, B, C, D and E were independently
defined as the foreground branch in the branch-site model
(Figure 2). When branches A, B, C or D were defined as
the foreground branch, the null hypothesis is rejected (P
<0.01); and the estimated parameters of the alternative
hypotheses indicate that about 5%-11% sites on these
branches were under positive selection with a ω value (ω2)
larger than one. When branch E was brought to the fore-
ground, the null hypothesis cannot be rejected, and thus
no significant positive selection was detected (Table 2).
This result suggests that positive selection mainly occurred
in the 11S globulin genes in dicots.
Discussion and Conclusions
Gene duplication may be associated with the higher 11S
globulin content in dicots
In our duplication analyses, we found four or more 11S
globulin genes in each of the five dicot species analyzed.
It appears that higher number of duplicates is a feature
of the dicot 11S globulin genes, rather than being ran-
domly produced by the hitchhiking effect of genome
duplication, because:
i) The copy number of the dicot 11S globulin gene is
higher than the average copy number of the genome;
for example, in the genome of A. thaliana, 80% of
genes recovered to single copy through gene loss in a
short period after the duplications, resulting an aver-
age of 2.3 copies per family [32], which is lower than
the 4 duplicates in the 11S globulin gene family; in
the genome of G. max, 74.1% and 56.6% genes were
lost following the early and the recent WGD, respec-
tively, resulting an average of about 3 copies per
family [33,34], which is lower than the 6 duplicates in
the globulin gene family;
ii) In each of the five dicots, there are 11S globulin
genes that arose from tandem duplications; and
iii) The genomes of S. italica, Z. mays and S. bicolor
are thought to have undergone several rounds of
duplications [35,36], but they contain only one or
two 11S globulin genes, suggesting that gene losses
are common in the family, and thus implying that
Table 1 Evolutionary rate analysis of 11S globulin family using branch-specific model of PAML
Modelω setting-ln L Estimated parametersLikelihood ratio test
One-ratio
Two-ratio
entire tree: ω0
monocot subfamily: ω0
dicot subfamily 1& 2: ω1
monocot subfamily: ω0
dicot subfamily 1: ω1
dicot subfamily 2: ω2
branch A: ω1
branch B: ω2
branch C: ω3
branch D: ω4
branch E: ω5
other branches: ω0
38214.64
38210.14
ω0= 0.234
ω0= 0.212
ω1= 0.249
ω0= 0.212
ω1= 0.252
ω2= 0.243
ω1= 0.225
ω2= 0.373
ω3= 0.198
ω4= 0.242
ω5= 0.187
ω0= 0.283
two ratio vs. one ratio: P < 0.01
Three-ratio
38209.98 three ratio vs. one ratio: P < 0.01
three ratio vs. two ratio: P > 0.05
Six-ratio
38182.58six ratio vs. one ratio: P < 0.01
six ratio vs. two ratio: P < 0.01
six ratio vs. three ratio: P < 0.01
Table 2 Summary statistics for detecting selection using branch-site models of PAML
Foreground branchNull hypothesis Alternative hypothesis
-ln L
Estimated parameters-ln L
Estimated parameters
Branch A
46859.30
p0= 0.66, p1= 0.28 (p2+p3= 0.06)
ω0= 0.23, ω1= ω2= 1
p0= 0.53, p1= 0.22 (p2+p3= 0.25)
ω0= 0.23, ω1= ω2= 1
p0= 0.53, p1= 0.21 (p2+p3= 0.36)
ω0= 0.23, ω1= ω2= 1
p0= 0.61, p1= 0.26 (p2+p3= 0.13)
ω0= 0.23, ω1= ω2= 1
p0= 0.70, p1= 0.30 (p2+p3< 0.01)
ω0= 0.23, ω1= ω2= 1
46851.6**
p0= 0.69, p1= 0.27 (p2+p3= 0.04)
ω0= 0.23, ω1= 1.00, ω2= 999
p0= 0.63, p1= 0.26 (p2+p3= 0.11)
ω0= 0.23, ω1= 1.00, ω2= 56.08
p0= 0.64, p1= 0.26 (p2+p3= 0.10)
ω0= 0.23, ω1= 1.00, ω2= 526.55
p0= 0.68, p1= 0.27 (p2+p3= 0.05)
ω0= 0.23, ω1= 1.00, ω2= 999
p0= 0.70, p1= 0.30 (p2+p3< 0.01)
ω0= 0.23, ω1= ω2= 1
Branch B
46854.00
46837.66**
Branch C
46852.38
46831.76**
Branch D
46859.11
46842.30**
Branch E
46859.33
46859.33
Note: ** P <0.01, calculated from LRT.
Li et al. BMC Evolutionary Biology 2012, 12:15
http://www.biomedcentral.com/1471-2148/12/15
Page 5 of 9