A Green-Cotyledon/Stay-Green Mutant Exemplifies
the Ancient Whole-Genome Duplications in Soybean
Michiharu Nakano1,6, Tetsuya Yamada2,6, Yu Masuda1,6, Yutaka Sato3, Hideki Kobayashi2, Hiroaki Ueda1,
Ryouhei Morita4,7, Minoru Nishimura4,8, Keisuke Kitamura2 and Makoto Kusaba1,*
Graduate School of Science, Hiroshima University, Kagamiyama, Higashi-Hiroshima, Hiroshima, 739-8526 Japan
Graduate School of Agriculture, Hokkaido University, Kita, Sapporo, Hokkaido, 060-8589 Japan
3
Agrogenomics Research Center, National Institute of Agrobiological Sciences, Kannondai, Tsukuba, Ibaraki, 305-860 Japan
4
Institute of Breeding Science, National Institute of Agrobiological Sciences, Kamimurata, Hitachi-Ohmiya, 219-2293 Japan
5
Core Research for Evolutional Science and Technology, The Japan Science and Technology Agency, Kagamiyama, Higashi-Hiroshima, Hiroshima,
739-8526 Japan
6
These authors contributed equally to this work.
7
Present address: Nishina Center for Accelerator-based Science, RIKEN, Wako, 351-0198 Japan.
8
Present address: Faculty of Agriculture, Niigata University, Ikarashi, Nishi-ku, Niigata, 950-2181 Japan.
2
*Corresponding author: E-mail, [email protected]; Fax, +81-82-424-0738.
(Received January 29, 2014; Accepted August 5, 2014)
The recent whole-genome sequencing of soybean (Glycine
max) revealed that soybean experienced whole-genome duplications 59 million and 13 million years ago, and it has an
octoploid-like genome in spite of its diploid nature. We
analyzed a natural green-cotyledon mutant line, Tenshindaiseitou. The physiological analysis revealed that Tenshindaiseitou shows a non-functional stay-green phenotype in
senescent leaves, which is similar to that of the mutant of
Mendel’s green-cotyledon gene I, the ortholog of SGR in pea.
The identification of gene mutations and genetic segregation
analysis suggested that defects in GmSGR1 and GmSGR2
were responsible for the green-cotyledon/stay-green phenotype of Tenshin-daiseitou, which was confirmed by RNA
interference (RNAi) transgenic soybean experiments using
GmSGR genes. The characterized green-cotyledon double
mutant d1d2 was found to have the same mutations, suggesting that GmSGR1 and GmSGR2 are D1 and D2. Among
the examined d1d2 strains, the d1d2 strain K144a showed a
lower Chl a/b ratio in mature seeds than other strains but
not in senescent leaves, suggesting a seed-specific genetic
factor of the Chl composition in K144a. Analysis of the soybean genome sequence revealed four genomic regions with
microsynteny to the Arabidopsis SGR1 region, which
included the GmSGR1 and GmSGR2 regions. The other two
regions contained GmSGR3a/GmSGR3b and GmSGR4, respectively, which might be pseudogenes or genes with a
function that is unrelated to Chl degradation during seed
maturation and leaf senescence. These GmSGR genes were
thought to be produced by the two whole-genome duplications, and they provide a good example of such wholegenome duplication events in the evolution of the soybean
genome.
Keywords: Green-cotyledon SGR Soybean Stay-green Whole-genome duplication.
Abbreviations: CaMV, Cauliflower mosaic virus; Chr,
chromosome; GFP, green fluorescent protein; LHC,
light-harvesting complex; RNAi, RNA interference; RT–PCR,
reverse transcription–PCR.
Introduction
Soybean (Glycine max) is an important crop that supplies oil
and good-quality proteins. The recent completion of the wholegenome sequencing of soybean revealed its paleopolyploidy
despite the diploid behavior of soybean chromosomes
(Schmutz et al. 2010). Two whole-genome duplications 13 million and 58 million years ago formed the basis of the present
octoploidic features of the soybean genome. Furthermore, the
soybean genome experienced a more ancient whole-genome
triplication >130 million years ago, before the radiation of the
Rosid or Fabid clades (Cannon and Shoemaker 2012).
There are not many reports about recessive mutants in soybean. This may be partly because single-gene mutations are
often compensated by functionally redundant genes (paralogs)
that were generated by the whole-genome duplications (Arase
et al. 2011). Most known single locus mutants exhibit their
phenotypes because of the functional diversification or the
gene dose effect of paralogous genes. Among them, d1d2, a
natural green-cotyledon mutant line with the recessive mutations d1 and d2, is an interesting example of a double mutant of
fully functionally redundant genes (Guiamet et al. 1991). In
soybean, another type of green-cotyledon mutant line has
been identified. In contrast to the Mendelian genetic factors
D1 and D2, this mutation, known as cytG, is maternally inherited (Guiamet et al. 1991). Both d1d2 and cytG are not
only green-cotyledon but also stay-green mutants, which
retain green leaves during senescence.
Stay-green mutants are classified into two groups: functional
stay-green mutants and non-functional stay-green mutants (or
Type C stay-green mutants) (Thomas and Howarth 2000).
While the progression of leaf senescence is impaired in the
functional stay-green mutant, the non-functional stay-green
Plant Cell Physiol. 55(10): 1763–1771 (2014) doi:10.1093/pcp/pcu107, Advance Access publication on 9 August 2014,
available online at www.pcp.oxfordjournals.org
! The Author 2014. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists.
All rights reserved. For permissions, please email: [email protected]
Regular Paper
1
M. Nakano et al. | SGR genes in soybean
mutant has a defect in Chl degradation, and leaf senescence
proceeds by senescence-inducing treatments. So far, several
non-functional stay-green genes have been isolated (Kusaba
et al. 2013). NYC1 and NOL encode Chl b reductases, the enzymes that catalyze the first step of Chl b degradation (Kusaba
et al. 2007, Horie et al. 2009, Sato et al. 2009). In rice nyc1 and nol
mutants, Chl b degradation is severely inhibited during senescence, and the final Chl a/b ratio approaches 1. In other nonfunctional stay-green mutants such as pph/nyc3, pao, Atnap1,
sgr/nye1 and thf1/nyc4, the Chl a/b ratio essentially does not
change during leaf senescence (Pružinská et al. 2003, Park et al.
2007, Ren et al. 2007, Morita et al. 2009, Schelbert et al. 2009,
Nagane et al. 2010, Sakuraba et al. 2012, Huang et al, 2013,
Yamatani et al. 2013). Interestingly, the Chl a/b ratio is about
1 in the senescent green leaves of cytG, while that of d1d2 is
higher (Guiamet et al. 1991).
The SGR ortholog in pea is known as Mendel’s green cotyledon gene; mutants of this gene show the green-cotyledon
phenotype in mature seeds and the non-functional staygreen phenotype in senescent leaves (Armstead et al. 2007,
Y. Sato et al. 2007, Aubry et al. 2008). In this report, we revealed
that D1 and D2 are the soybean SGR orthologs GmSGR1 and
GmSGR2. The requirement to impair both GmSGR1 and
GmSGR2 to express the green-cotyledon/stay-green phenotype
indicates that they are fully functionally redundant genes. In
addition, in the soybean genome, we found two more genomic
regions containing SGR-like genes (GmSGR3/GmSGR4) with
microsynteny to Arabidopsis thaliana SGR1 (NYE1/SGN). It is
possible that the ancestral soybean genome harbored two SGR
genes that were generated by the ancient whole-genome duplication (58 million years ago), and the GmSGR1/GmSGR2 and
GmSGR3/GmSGR4 pairs were generated by the recent genome
duplication (13 million years ago).
Results and Discussion
Fig. 1 Green-cotyledon/stay-green phenotype and physiological
changes in Tenshin-daiseitou during incubation in the dark. (A)
Senescent leaves and seed color of Tachiyutaka (wild-type) and
Tenshin-daiseitou. Pre-senescent and senescent leaves and mature
seeds with (upper row) or without (lower row) seed coats are
shown. (B–D) Changes of Chl content (SPAD value) (B), Fv/Fm (C)
and membrane ion leakage (D) during dark incubation. Solid lines and
dotted lines indicate Tachiyutaka and Tenshin-daiseitou, respectively.
Bars indicate the SDs. n = 4 in (B) and (C); n = 3 in (D).
Physiological characterization of Tenshin-daiseitou
Several stay-green mutants, which retain green in the leaf
during senescence, have been reported in legume (Guiamet
et al. 1991, Armstead et al. 2007, Y. Sato et al. 2007, Zhou
et al. 2011). Although the cotyledon color of mature seeds is
yellow in the wild-type legume, these stay-green mutants have
green cotyledons. Among them, we analyzed Tenshin-daiseitou,
a soybean line with a nuclear-inherited green-cotyledon/staygreen trait (Fig. 1A). When mature leaves were incubated in the
dark at 25 C to induce leaf senescence, the Chl content
decreased drastically by 9 d after dark incubation in the wildtype cultivar Tachiyutaka, but it remained relatively high in
Tenshin-daiseitou (Fig. 1B). However, the Fv/Fm value, an indicator of PSII activity, was not higher in Tenshin-daiseitou than
in Tachiyutaka during the incubation despite green color being
retained (Fig. 1C). Similarly, the membrane ion leakage of
Tenshin-daiseitou, an indicator of cell death, was not lower
than that of Tachiyutaka at 12 d after dark incubation (Fig. 1D).
Additionally, we observed ultrastructural changes in
the chloroplasts during dark-induced leaf senescence
1764
(Supplementary Fig. S1). The chloroplast ultrastructure of
Tenshin-daiseitou is similar to that of Tachiyutaka in the presenescent leaves; however, they are very different in senescent
leaves. In the senescent leaves, poorly stacked grana were
observed in Tachiyutaka, while dense and thick grana were
observed in Tenshin-daiseitou (Supplementary Fig. S1E, F).
Such retention of the grana structure during leaf senescence
is the common feature among non-functional stay-green mutants (Park et al. 2007, Kusaba et al. 2007, Morita et al. 2009,
Sato et al. 2009, Schelbert et al. 2009, Zhou et al. 2011). These
results indicate that Tenshin-daiseitou is a non-functional staygreen mutant.
Next, the protein degradation during the dark incubation was
examined using green gel analysis, which is used to visualize Chl–
protein complexes (Fig. 2). In pre-senescent leaves of
Tachiyutaka and Tenshin-daiseitou, the PSI reaction center, containing only Chl a, and trimers and monomers of light-harvesting
complex II (LHCII), containing Chl a and Chl b, were
observed. While both the PSI reaction center and LHCII were
Plant Cell Physiol. 55(10): 1763–1771 (2014) doi:10.1093/pcp/pcu107
unchanged from that of non-senescent leaves (3.6 ± 0.09).
These phenotypes are distinct from those of nyc1 and nol mutants, in which LHCII is selectively retained and the Chl a/b ratio
is about 1, indicating that the stay-green phenotype of Tenshindaiseitou is not caused by a mutation of NYC1 or NOL. Rather,
the phenotypes of Tenshin-daiseitou are similar to those of sgr/
nye1 or pph/nyc3 mutants, in which both the PSI reaction
center and LHCII are retained during leaf senescence.
Isolation of the stay-green genes in
Tenshin-daiseitou
Fig. 2 Green gel analysis of chloroplast proteins in senescent leaves of
Tenshin-daiseitou. The upper panel shows Chl-binding proteins of
pre-senescent (P) and senescent (S) leaves. The gel was visualized by
Chl bound to proteins. The middle panel shows D1 protein detected
by Western blot analysis. The lower panel shows the Rubisco small
subunit stained with Coomassie Brilliant Blue G250. LHC II, light-harvesting complex II.
degraded in the senescent leaves of Tachiyutaka, these protein
complexes were retained in Tenshin-daiseitou. Western blot analysis revealed that D1, a core component of PSII, was also retained
in the senescent leaves of Tenshin-daiseitou but its retention
level was lower than that of PSI. SDS–PAGE analysis showed
that the Rubisco small subunit was degraded in the senescent
leaves of both Tachiyutaka and Tenshin-daiseitou (Fig. 2), suggesting that not all of the photosynthetic proteins are retained in
Tenshin-daiseitou during leaf senescence.
In HPLC analysis of photosynthetic pigments, two Chls, Chl a
and Chl b, and four carotenoids, neoxanthin, violaxanthin,
lutein and b-carotene, were detected in pre-senescent leaves
(Fig. 3A). In Tachiyutaka, the contents of all the pigments were
drastically reduced in senescent leaves. On the other hand, Chl
a and Chl b were retained at higher levels in Tenshin-daiseitou.
In addition, neoxanthin, violaxanthin and lutein were also retained in Tenshin-daiseitou, which is very probably due to the
inhibited degradation of LHCII, which contains these carotenoids (Kusaba et al. 2007). Spectrophotometric analysis confirmed that both Chl a and Chl b were retained at
remarkably higher levels in the senescent leaves of Tenshindaiseitou than in those of Tachiyutaka, while Tenshin-daiseitou
contains slightly more Chl in the pre-senescent leaves than
Tachiyutaka (Fig. 3B). The Chl a/b ratio in senescent leaves
of Tenshin-daiseitou was 3.5 ± 0.47, and it was essentially
For genetic characterization of the green-cotyledon/stay-green
genes in Tenshin-daiseitou, we obtained F2 seeds from the cross
between Tenshin-daiseitou and Tachiyutaka. Among 182 F2
seeds that we examined, plants from 11 seeds showed the
green-cotyledon phenotype. This segregation ratio is not significantly different from 15 : 1 (2 = 0.032; P = 0.86), suggesting
that a homozygous state of recessive alleles at two loci is
required for the phenotype. The 11 green-cotyledon plants
showed the stay-green phenotype, confirming that the greencotyledon phenotype is due to the pleiotropic effects of
impaired Chl degradation. As mentioned above, the greencotyledon/stay-green gene is not thought to be an orthlog of
either NYC1 or NOL. In pea and alfalfa, genes responsible for the
green-cotyledon mutants have been shown to be orthologs of
SGR (Armstead et al. 2007, Y. Sato et al. 2007, Zhou et al. 2011),
and their physiological characteristics are similar to those of
Tenshin-daiseitou. Considering these lines of information, we
analyzed SGR homologs of soybean in Tenshin-daiseitou. Two
soybean genes that are highly similar to SGR, GmSGR1
(AAW82959.1) and GmSGR2 (AAW82960.1), have been deposited in the database. The amino acid sequences of the proteins encoded by these genes show 66% and 64% identity to the
SGR ortholog in A. thaliana (called AtSGR1 here), and 91%
identity to each other. Like AtSGR1 and other SGR orthologous
proteins, GmSGR1 and GmSGR2 were predicted to be localized
in chloroplasts (WoLF PSORT; http://wolfpsort.org).
The sequence analysis of GmSGR2 in Tenshin-daiseitou revealed a single base pair deletion in the second exon, resulting
in a frameshift and premature termination of translation (Fig.
4A). On the other hand, while most parts of the GmSGR1 genes
were amplified from the Tenshin-daiseitou genome, the entire
GmSGR1 gene was not amplified. Reverse transcription–PCR
(RT–PCR) amplification using primers for the entire coding
sequence revealed that GmSGR1 was transcribed in senescent
leaves, but its size was aberrant. This product has a complex
structure with a truncated fourth exon that is followed by a
duplication of the complete third and fourth exons (Fig. 4B).
This aberrant mRNA may encode a protein lacking an amino
acid sequence corresponding to the truncated region of the
fourth exon, suggesting that the GmSGR1 allele in Tenshindaiseitou is not functional. The genomic fragment corresponding to the full size of the aberrant mRNA was not amplified by
PCR, probably due to a complex genomic structure or a
long insertion. The observation that both SGR homologs are
defective is consistent with the involvement of two
1765
M. Nakano et al. | SGR genes in soybean
A
Tachiyutaka
400
400
presenescent
200
b
L
N
C
V
0
0
5
10
15
20
25
30
400
senescent
200
absorbancce at 441nm (mAU)
absorbancce at 441nm (mAU)
Tenshin-daiseitou
a
presenescent
a
200
L
N
C
0
400
0
5
10
15
20
25
senescent
30
a
200
L
N
b
C
V
0
0
0
5
10
15
20
25
30
retention time (min)
B
**
1.4
chloro
ophyll content (nm
mol/mgFW)
b
V
**
ns
*
0
5
10
15
20
25
30
retention time (min)
ns
ns
1.2
1.0
0.8
0.6
0.4
0.2
0.0
P
S
P
S
Tachiyutaka Tenshin-daiseitou
Fig. 3 Analysis of photosynthetic pigments in senescent leaves of Tenshin-daiseitou. (A) HPLC analysis of photosynthetic pigments. The same
volume of extract was loaded in each injection. AU, absorption unit; N, neoxanthin; V, violaxanthin; L, lutein; b, Chl b; a, Chl a; C, b-otene. (B) Chl
a and Chl b contents of pre-senescent (P) and senescent (S) leaves. The filled and open bars indicate Chl a and Chl b contents, respectively.
Asterisks indicate significant differences between samples (Student’s t-test P-values, *P < 0.05; **P < 0.01; ns, non-significant). Bars indicate the
SDs. n = 4.
recessive mutations in the green-cotyledon phenotype of
Tenshin-daiseitou.
To confirm that the two SGR homologs are the green-cotyledon/stay-green genes, we performed linkage analysis using the
F2 population between Tenshin-daiseitou and Tachiyutaka.
All of the 11 plants that were grown from green-cotyledon
seeds showed no amplification of GmSGR1 and were homozygous for the single-base deletion in GmSGR2, indicating
that they were homozygous for mutant alleles of both
GmSGR1 and GmSGR2. Next, 28 plants grown from yellowcotyledon seeds were analyzed. GmSGR1 was amplified from
1766
all except five of these plants, and all of these five plants carried
the wild-type GmSGR2 allele in a homozygous or heterozygous
manner, indicating that the 28 plants carried at least one
wild-type allele of either GmSGR1 or GmSGR2. These observations indicated complete linkage between the green-cotyledon
phenotype and the impairment of both GmSGR1 and GmSGR2.
For further confirmation that the impairment of GmSGRs
caused the green-cotyledon/stay-green phenotype, we produced the GmSGR2 RNA interference (RNAi) transgenic soybean line. Because the double-stranded RNA-forming region
of GmSGR2 showed a high degree of identity to GmSGR1
Plant Cell Physiol. 55(10): 1763–1771 (2014) doi:10.1093/pcp/pcu107
Fig. 5 Phenotype of GmSGR RNAi transgenic plants. (A) Semi-quantitative RT–PCR analysis of the expression of GmSGR1 and GmSGR2 in
the GmSGR2 RNAi transgenic plants. Total RNA was extracted from
mature seeds. Lane 1: control BASTA-resistant transgenic plant. Lanes
2 and 3: RNAi-1 and RNAi-2 plants. RNAi-1 and RNAi-2 are independent GmSGR2 RNAi transgenic plants. b-Tubulin was used as a reference. (B) Seed color of transgenic seeds from control, RNAi-1 and
RNAi-2 plants with (upper row) and without (lower row) the seed
coat.
Fig. 4 Expression and structures of GmSGR1 and GmSGR2 in
Tachiyutaka and Tenshin-daiseitou. (A) The upper panel shows genomic and RT–PCR amplification of GmSGR2. The lower panel shows
the structure of the GmSGR2 transcript. Tenshin-daiseitou has a single
base pair deletion in the second exon, resulting in a frameshift and
premature termination of translation. (B) The upper panel shows
genomic and RT–PCR amplification of GmSGR1. The genomic fragment of GmSGR1 was not amplified in Tenshin-daiseitou, and an
aberrantly sized product was amplified by RT–PCR using the same
primer pair. The lower panel shows the structure of the GmSGR1
transcript and its deduced amino acid sequence. The aberrant
GmSGR1 transcript in Tenshin-daiseitou has a complex structure: a
truncation of the fourth exon at 625 bases from the initiation codon
followed by a duplication of the complete third and fourth exons. This
structural change results in a frameshift and premature termination of
translation. The numbers in the boxes show the exon numbers, and 40
indicates the truncated fourth exon. The asterisk shows the position
of translation termination.
(97%), GmSGR1 was also expected to be silenced efficiently
(Supplementary Fig. S2). The mRNA levels of both
GmSGR1 and GmSGR2 in the mature seeds of the resultant
RNAi transgenic plants were lower than those in the control
transgenic line (Fig. 5A). These RNAi transgenic lines produced
green-cotyledon seeds, supporting the hypothesis that mutations in GmSGR1 and GmSGR2 caused the green-cotyledon
phenotype (Fig. 5B).
In addition, we examined the genomic structures of
GmSGR1 and GmSGR2 in seven other green-cotyledon lines,
which showed Chl a/b ratios >1 in mature seeds
(Supplementary Table S1). All of these green-cotyledon lines
showed no amplification of the full-length GmSGR1 and have
the same point mutation in GmSGR2 as Tenshin-daiseitou.
Taken together, we concluded that GmSGR1 and GmSGR2
are functionally redundant genes; additionally, defects of both
genes are necessary for the green-cotyledon/stay-green phenotype in soybean. Consistent with this conclusion, GmSGR1 and
GmSGR2 showed very similar expression patterns (Fig. 6). They
are abundantly expressed in senescent leaves and mature seeds,
in which Chl degradation occurs actively. T41, one of the greencotyledon lines that we analyzed, carries the genetically
characterized green-cotyledon mutations, d1 and d2. The
observation that T41 has the same changes in GmSGR1 and
GmSGR2 as Tenshin-daiseitou suggests that D1 and D2 are
GmSGR1 and GmSGR2.
Very recently, Fang et al. (2014) reported that D1 and D2
encode GmSGR2 and GmSGR1, respectively. The lesions in
GmSGR1 and GmSGR2 are the same as those in Tenshin-daiseitou, suggesting that the d1d2 alleles that they used have the
same origin as Tenshin-daiseitou. Additionally, Fang et al.
(2014) reported that D2 (GmSGR1) has a transposon insertion,
which probably inhibited our genomic PCR amplification of
full-length GmSGR1 in Tenshin-daiseitou.
1767
M. Nakano et al. | SGR genes in soybean
Fig. 6 Semi-quantitative RT–PCR analysis of GmSGR1 and GmSGR2
expression in various tissues. Total RNA was extracted from young
leaf, mature leaf, senescent leaf, flower, stem, immature seed-1 (green),
immature seed-2 (yellow) and mature seed (dry).
Seed-specific regulation of the Chl a/b ratio in the
d1d2 strains
In Supplementary Table S1, d1d2 mutants showed a Chl a/b
ratio >2.0, except K144a, which showed a ratio of around 1.5.
Therefore, we analyzed K144a in more detail. The Chl a/b ratios
in mature seeds and senescent leaves are shown in
Supplementary Fig. S3. In mature seeds, K144a showed a
Chl a/b ratio of 1.62 ± 0.10, which was significantly lower
than that of Tenshin-daiseitou (2.35 ± 0.28; P < 0.05)
(Supplementary Fig. S3). Conversely, in senescent leaves, a
significant difference was not observed in the Chl a/b ratio
between K144a and Tenshin-daiseitou. Additionally, K144a
and Tenshin-daiseitou did not show an obvious difference in
the green gel analysis of senescent leaves (Supplementary Fig.
S4). K144a has the same lesion in GmSGR1 and GmSGR2 as
Tenshin-daiseitou (Supplementary Table S1). These observations suggest that a genetic factor of K144a, which is a different
locus from D1 and D2, modulates the Chl a/b ratio specifically
in seeds.
Evolution of GmSGR genes
A homology search using the amino acid sequence encoded by
AtSGR1 identified 12 genomic regions containing genes with
high similarity to AtSGR1 (Supplementary Table S2). Of these
regions, four genomic regions on chromosomes 1 (Chr 1), Chr 5,
Chr 11 and Chr 17 showed microsynteny to the AtSGR1 region
in A. thaliana according to SyMap and PGDD (Fig. 7A). The Chr
1 region contains GmSGR2 and the Chr 11 region contains
GmSGR1, confirming that GmSGR1 and GmSGR2 are orthologs
of AtSGR1. Fang et al. (2014) reported that they are embedded
with approximately 7 Mb duplicated blocks and were generated
1768
Fig. 7 Phylogenetic analysis of SGR genes in legumes. (A) Synteny
maps between AtSGR1 and GmSGR genes. Maps were modified
from the Plant Genome Duplication Database (http://chibba.agtec.
uga.edu/duplication). Genes showing synteny are indicated by
hatched arrows with the gene ID number and are bound with a
line. Filled arrows indicate SGR genes. (B) A phylogenetic tree of
GmSGR and SGR-related genes in legumes. The DNA region in
which GmSGR4 shows a high degree of homology to other GmSGR
genes was used for analysis. Bootstrap values are shown as a percentage from 1,000 replicates for each node (values >60% were indicated).
GmSGR1 (Glyma11g02980.1), GmSGR2 (Glyma01g42390.1), GmSGR3a
(Glyma17g14201.2), GmSGR3b (Glyma17g14210.2), GmSGR4
(Glyma05g03700.1), PvSGR (Phvul.002G153100.1) and VuSGR
(FF383809 and other expressed sequence tags) form the Phaseoloid
clade, and PsSGR (AB303331), MtSGR (HQ849484), LjSGR1 (chr2.
CM0021.2870.r2.m), LjSGR2 (chr4.CM0126.600.r2.m) and TpSGR
(BB916350 and other expressed sequence tags) form the Galegoid
clade. Gm, Glycine max; Ps, Pisum sativum; Mt, Medicago truncatula;
Lj, Lotus japonicas; Pv, Phaseolus vulgaris; Tp, Trifolium pratense; Vu,
Vigna unguiculata.
Plant Cell Physiol. 55(10): 1763–1771 (2014) doi:10.1093/pcp/pcu107
by the most recent whole-genome duplication. In the Chr 17
region, there are two tandemly duplicated genes with high
similarity to AtSGR1 named GmSGR3a and GmSGR3b
(Supplementary Fig. S5). The Chr 5 region contains
Glyma05g03700, which is a predicted SGR-like gene according
to the previous annotation (v1.0) of phytozome, and is named
GmSGR4. Recently, Rong et al. (2013) reported that an SGR-like
gene in rice (SGRL) has some function in Chl degradation. There
are two probable orthologs of SGRL in G. max, but they form a
clade that is distinct from that of the SGR genes and are not
syntenic to AtSGR1 (Rong et al. 2013; Supplementary Table S2;
Supplementary Fig. S6).
GmSGR4 is an apparent pseudogene because the second to
fourth exons are incomplete, and the first exon is missing. On
the other hand, both GmSGR3a and GmSGR3b on Chr 17
appear to retain the basic structure of SGR genes including
the transit peptide. However, GmSGR3b lacks the conserved
cysteine-rich motif (C-X3-C-X-C2FP-X5-P) in the C-terminal
region, and the motif is incomplete in GmSGR3a
(Supplementary Fig. S7). In addition, both GmSGR3a and
GmSGR3b have different amino acid residues instead of the
several highly conserved residues among SGRs. For example,
GmSGR3a and GmSGR3b have a C143L change relative to
GmSGR1 (Supplementary Fig. S7). We analyzed the expression
of GmSGR3a and GmSGR3b in several tissues including senescent leaves and mature seeds (Supplementary Fig. S8A). RT–
PCR analysis using the primer pair that amplifies both
GmSGR3a and GmSGR3b revealed that neither of them was
expressed in the examined tissues. Their expression was not
observed in the d1d2 mutants, Tenshin-daiseitou and K144a
(Supplementary Fig. Sb8B). These observations suggest that
GmSGR3a and GmSGR3b are probably pseudogenes.
To analyze the phylogenetic relationships of the five GmSGR
genes, a phylogenetic tree was constructed using genomic DNA
sequences corresponding to the 254–403, 1,086–1,153 and
1,394–1,469 bp regions from the initiation codon of GmSGR1,
where similarity is retained among all five GmSGR genes (Fig.
7B; Supplementary Fig. S9). Probable SGR orthologs in legume
were included in the analysis. GmSGR1 and GmSGR2, and
GmSGR3a, GmSGR3b and GmSGR4 formed very close clades,
respectively. GmSGR3a and GmSGR3b are tandemly repeated
on Chr 17, suggesting that they duplicated independently of the
whole-genome duplication. Thus, it is conceivable that the
common ancestor of these five GmSGR genes duplicated in
the ancient whole-genome duplication that occurred 58 million
years ago, and the GmSGR1/GmSGR2 and GmSGR3/GmSGR4
pairs were formed in the recent whole-genome duplication that
occurred 13 million years ago. GmSGR1 and GmSGR2 were
grouped in the Phaseoloid clade together with PvSGR and
VuSGR. PsSGR, MtSGR and TpSGR formed the Galegoid clade.
Interestingly, GmSGR3 and GmSGR4 belong to neither the
Phaseoloid clade nor the Galegoid clade, and they were long
genetic distances from other SGR genes in legumes, suggesting
that GmSGR3 and GmSGR4 diverged rapidly in comparison
with other GmSGR genes, probably because of the relaxed selective force caused by loss of SGR function or positive selection
to obtain a new function.
In this study, we revealed that the D1 and D2 genes are
functionally redundant orthologs of SGR in soybean
(GmSGR1 and GmSGR2) using genetic and functional analyses.
Our survey of green-cotyledon strains that have a Chl a/b ratio
that was >1 in mature seeds revealed that all seven strains had
the same mutations in GmSGR1 and GmSGR2. This provides a
contrast to the SGR ortholog in pea, Mendel’s I gene. Three
types of mutations (three alleles) were found in nine greencotyledon strains that were examined (Y. Sato et al. 2007).
The observation that all of the strains that we surveyed had
the same lesions in GmSGR1 and GmSGR2 must reflect the fact
that the natural occurrence of a double mutant is very rare. The
soybean genome contained three more SGR genes distinct from
the SGRL genes (Rong et al. 2013). These five genes were
thought to be generated by the recent and ancient wholegenome duplication events, which are the basis for the present
octoplodic structure of the soybean genome. Our analysis illustrates an evolutionary history of a gene that experienced two
whole-genome duplication events in soybean by both sequence-based and function-based analyses.
Materials and Methods
Plant materials
Glycine max cv. Tenshin-daiseitou was obtained from Genebank, National
Institute of Agrobiological Sciences (Tsukuba, Japan); the Tachiyutaka,
Aomarukun and Hiratokomame cultivars were obtained from the National
Agriculture and Food Research Organization (Morioka, Japan); the Kariyutaka
cultivar was obtained from the Hokkaido Prefectural Tokachi Agricultural
Experiment Station (Memuro, Japan); the T41 strain was obtained from the
US Department of Agriculture Plant Germplasm Inspection Station (MD, USA);
and the BaiHuaLuDaDou, LuHuangDou, Kangwon, Kyonggi and Chungbuk cultivars were obtained from the National Bioresource Project (Lotus japonicus and
G. max); the Kiyomidori cultivars was purchased from Nakahara-saishujou
(Fukuoka, Japan).
Physiological analysis
Soybean plants were grown in a growth chamber (200 mmol photon m2 s1,
16 h light and 8 h dark at 27 C) for about 2 weeks. The second leaves from the
top were used as pre-senescent leaves. To induce leaf senescence, the leaves
were detached and incubated at 25 C in Petri dishes under dark conditions (Fig
1B, C, D). In other experiments, shoots with two expanded leaves were incubated for 6 d in the dark at 27 C and the second leaves from the top were
analyzed as senescent leaves. For pigment extraction, plant tissues of the same
fresh weight were ground in mortars with liquid nitrogen and extracted with
the same volume of 80% acetone. Chl contents were determined spectrophotometrically according to Porra et al. (1989) or non-destructively with an SPAD
502 Chl meter (Konica Minolta Co. Ltd.). HPLC analysis of photosynthetic pigments was performed according to Yamatani et al. (2013). Fv/Fm values were
measured with a Junior PAM Chl fluorometer (Walz) according to the manufacturer’s protocol. To measure the membrane ion leakage, three leaf discs of
6 mm diameter were floated on 500 ml of distilled water and incubated in the
dark at 25 C. The conductivity was measured with a Twin Cond B-173 conductivity meter (Horiba).
Protein analysis
Green gel analysis was performed according to Morita et al. (2005). Fresh weight
leaf samples (100 mg) were extracted with 250 ml of extraction buffer (0.3 M
Tris, pH 6.8, 1% SDS, 2% Triton X-100, 10% glycerol). For Rubisco small subunit
analysis, the same samples were analyzed by conventional SDS–PAGE and
stained with Coomassie Brilliant Blue G250. Western blot analysis was done
according to Yamatani et al. (2013).
1769
M. Nakano et al. | SGR genes in soybean
Genomic and RT–PCR analysis
For RT–PCR, the first-strand cDNA was synthesized from 100 ng of total RNA
using ReverTra Ace qPCR RT Master Mix with gDNA Remover (Toyobo).
Genomic and RT–PCR were performed using KAPA Taq Extra (KAPA
Biosystems) with primer pairs GmSGR1-2F (50 -CCTTCGCTTTCTCCGCAC
CAC-30 ) and GmSGR1-1R (50 -AACCCCTCCTCCTGGACC-30 ) for GmSGR1,
GmSGR2-2F (50 -GATTGGAACAACAACAGGAA-30 ) and GmSGR2-1R (50 -AGT
TTCTTGCTCACTCTCTTC-30 ) for GmSGR2, GmSGR3-F (50 - GGGCCAGCCATC
TTTGAA-30 ) and GmSGR3-R (50 - TCCACTGACATGGAGGTGAAC-30 ) for
GmSGR3, and Gmb-tub-F (50 -CAAACTCGCCGTGAACCTCATC-30 ) and
Gmb-tub-R (50 -GACTTGAGGCCACCACACCTT-30 ) for b-tubulin (Glyma
08g01740). The PCR profile consisted of initial denaturation for 3 min at
94 C; 35 cycles of 15 s at 94 C, 30 s at 60 C and 3 min at 68 C; and a final
extension of 7 min at 68 C.
obtained from NCBI by a homology search with TBLASTN and
assembled into a single unigene with CLC Main workbench (CLC bio).
The following sequences were used: BB916350, BB925421 and BB919102
for TpSGR; and FF383809, FF394723, FF398594, FF538603, FF543105,
FG809139, FG825778, FG825779, FG829534, FG920394 and FG920395 for
VuSGR. The abbreviations of species are as follows: Gm, Glycine max; Ps,
Pisum sativum; Mt, Medicago truncatula; Lj, Lotus japonicas; Pv, Phaseolus vulgaris; Tp, Trifolium pratense; Vu, Vigna unguiculata; and At, Arabidopsis
thaliana.
Transmission electron microscopy
The ultrastructure of chloroplasts was observed by transmission electron microscopy according to Morita et al. (2009).
Production of RNAi transgenic plants
An RNAi vector, pSGR-RNAi, was constructed according to the following procedures. A 457 bp fragment of GmSGR2 was amplified with the primers GmSGRRNAiF (50 -CGAAGTGGTGGCACAGTGGAA-30 ) and GmSGR-RNAiR (50 -CCCAT
CACAAGGTTCGTAATGG-30 ) using the soybean cDNA as the template DNA.
The amplified product was cloned in an entry vector using the pCR8/GW/
TOPO TA cloning kit (Invitrogen). The region of the hairpin loop RNAi
cassette and GFP (green fluorescent protein) unit, which contains the
Cauliflower mosaic virus 35S (CaMV 35S) promoter and nopaline synthase
terminator from the pBI-sense, antisense-GW (GFP) vector (Inplanta) was inserted
into a binary vector, pMDC123 (Curtis and Grossniklaus 2003). The CaMV 35S
promoter with a duplicated enhancer region in the RNAi cassette was
replaced with a promoter region (AB353075) of the soybean gy1 gene. The resulting binary vector was used as a destination vector. The entry vector and destination
vector were reacted using Gateway LR Clonase (Invitrogen) to construct
pSGR-RNAi.
Agrobacterium-mediated transformation was performed using G. max cv.
Kariyutaka according to the procedure of H. Sato et al. (2007) and Yamada et al.
(2010) with modifications. The cotyledonary node of each explant was
wounded by a stainless steel surgical knife when the explants were prepared
for the infection with Agrobacteirum strain EHA105 harboring pSGR-RNAi. The
transgene was genetically fixed by repeating self-fertilization. T3 seeds were used
in this study. A transgenic soybean plant harboring only a selectable marker
gene, Bar, was used as a control plant. All transgenic plants used in this study
have resistance to the herbicide BASTA.
Supplementary data
Supplementary data are available at PCP online.
Funding
This work was supported by Core Research for Evolutional
Science and Technology [to M.K.] and in part JSPS KAKENHI
Grant Number 26292006 [to M.K.].
Acknowledgment
We thank Yuhi Kono for providing soybean strains
(Tachiyutaka, Aomarukun and Hiratokomame), Kanae Koike
for the ultrastructural analysis of chloroplasts, and Kaori
Kohzuma and Yumi Nagashima for their help.
Disclosures
The authors have no conflicts of interest to declare.
Phylogenetic analysis
A synteny search was performed with a 100 kb genomic region containing
AtSGR1 (Chr 4: 11,966–12,065 kb) against the whole-genome sequence of G.
max v1.1 using the SyMap synteny browser (http://www.symapdb.org/projects/fabaceae/). The synteny map between GmSGR genes and AtSGR1 was
constructed according to the Plant Genome Duplication Database (http://
chibba.agtec.uga.edu/duplication/) with a slight modification. Phylogenetic
analyses were conducted in MEGA5 (Tamura et al. 2011) using the regions
corresponding to 254–403, 1,086–1,153 and 1,394–1,469 bp from the
GmSGR1 genomic sequence, which retains high homology to GmSGR4
[Glyma05g03700.1, G. max genome v1.01 at SoyBase (http://soybase.org/)].
The sequence alignment was conducted using ClustalW with the default
settings. The evolutionary history was inferred using the maximum
likelihood method based on the Tamura-Nei model. All positions containing
gaps and missing data were eliminated. The amino acid sequence alignment
was performed using ClustalW with the default settings, followed by
adjustments that were made by eye. The following sequences were used:
PvSGR, Phvul.002G153100.1 (Phaseolus vulgaris genome v1.0 at Phytozome);
LjSGR genes (LjSGR1, chr2.CM0021.2870.r2.m; LjSGR2, chr4.CM0126.600.r2.m)
(Lotus japonicus genome assembly build 2.5 at http://www.kazusa.or.jp/lotus/);
GmSGR genes (GmSGR1, Glyma11g02980.1; GmSGR2, Glyma01g42390.1;
GmSGR3a, Glyma17g14201.2, GmSGR3b, Glyma17g14210.2) (Glycine max
genome v1.1 at Phytozome); PsSGR, AB303331 (Y. Sato et al. 2007);
and MtSGR, HQ849484 (Zhou et al. 2011). For TpSGR and VuSGR,
the expressed sequence tags that were homologous to AtSGR1 were
1770
References
Arase, S., Hase, Y., Abe, J., Kasai, M., Yamada, T., Kitamura, K. et al. (2011)
Optimization of ion-beam irradiation for mutagenesis in soybean: effects on plant growth and production of visibly altered mutants. Plant
Biotechnol. 28: 323–329.
Armstead, I., Donnison, I., Aubry, S., Harper, J., Hörtensteiner, S., James, C.
et al. (2007) Cross-species identification of Mendel’s I locus. Science 315:
73.
Aubry, S., Mani, J. and Hörtensteiner, S. (2008) Stay-green protein, defective in Mendel’s green cotyledon mutant, acts independent and upstream of pheophorbide a oxygenase in the chlorophyll catabolic
pathway. Plant Mol. Biol. 67: 243–256.
Cannon, S.B. and Shoemaker, R.C. (2012) Evolutionary and comparative
analyses of the soybean genome. Breed. Sci. 61: 437–444.
Curtis, M.D. and Grossniklaus, U. (2003) A gateway cloning vector set for
high-throughput functional analysis of genes in planta. Plant Physiol.
133: 462–469.
Fang, C., Li, C., Li, W., Wang, Z., Zhou, Z., Shen, Y. et al. (2014) Concerted
evolution of D1 and D2 to regulate chlorophyll degradation in soybean.
Plant J. 77: 700–712.
Guiamet, J.J., Schwartz, E., Pichersky, E. and Nooden, L.D. (1991)
Characterization of cytoplasmic and nuclear mutations affecting
Plant Cell Physiol. 55(10): 1763–1771 (2014) doi:10.1093/pcp/pcu107
chlorophyll and chlorophyll-binding proteins during senescence in soybean. Plant Physiol. 96: 227–231.
Horie, Y., Ito, H., Kusaba, M., Tanaka, R. and Tanaka, A. (2009) Participation
of chlorophyll b reductase in the initial step of the degradation of lightharvesting chlorophyll a/b-protein complexes in Arabidopsis. J. Biol.
Chem. 284: 17449–17456.
Huang, W., Chen, Q., Zhu, Y., Hu, F., Zhang, L., Ma, Z. et al. (2013)
Arabidopsis thylakoid formation 1 is a critical regulator for dynamics
of PSII–LHCII complexes in leaf senescence and excess light. Mol. Plant
6: 1673–1691.
Kusaba, M., Ito, H., Morita, R., Iida, S., Sato, Y., Fujimoto, M. et al. (2007)
Rice NON-YELLOW COLORING1 is involved in light-harvesting complex II and grana degradation during leaf senescence. Plant Cell 19:
1362–1375.
Kusaba, M., Tanaka, A. and Tanaka, R. (2013) Stay-green plants: what do
they tell us about the molecular mechanism of leaf senescence.
Photosynth. Res. 117: 221–234.
Morita, R., Kusaba, M., Yamaguchi, H., Amano, E., Miyao, A., Hirochika, H.
et al. (2005) Characterization of chlorophyllide a oxygenase (CAO) in
rice. Breed. Sci. 55: 361–364.
Morita, R., Sato, Y., Masuda, Y., Nishimura, M. and Kusaba, M. (2009)
Defect in non-yellow coloring 3, an a/b hydrolase-fold family protein,
causes a stay-green phenotype during leaf senescence in rice. Plant J. 59:
940–952.
Nagane, T., Tanaka, A. and Tanaka, R. (2010) Involvement of AtNAP1 in
the regulation of chlorophyll degradation in Arabidopsis thaliana.
Planta 231: 939–949.
Park, S.Y., Yu, J.W., Park, J.S., Li, J., Yoo, S.C., Lee, N.Y. et al. (2007) The
senescence-induced staygreen protein regulates chlorophyll degradation. Plant Cell 19: 1649–1664.
Porra, R.J., Thompson, W.A. and Kriedemann, P.E. (1989) Determination
of accurate extinction coefficients and simultaneous equations
for assaying chlorophylls a and b extracted with four different
solvents: verification of the concentration of chlorophyll standards
by atomic absorption spectroscopy. Biochim. Biophys. Acta 975:
384–394.
Pruzinská, A., Tanner, G., Anders, I., Roca, M. and Hörtensteiner, S. (2003)
Chlorophyll breakdown: pheophorbide a oxygenase is a Rieske-type
iron–sulfur protein, encoded by the accelerated cell death 1 gene.
Proc. Natl Acad. Sci. USA 100: 15259–15264.
Ren, G., An, K., Liao, Y., Zhou, X., Cao, Y., Zhao, H. et al. (2007)
Identification of a novel chloroplast protein AtNYE1 regulating chlorophyll degradation during leaf senescence in Arabidopsis. Plant Physiol.
144: 1429–1441.
Rong, H., Tang, Y., Zhang, H., Wu, P., Chen, Y., Li, M. et al. (2013) The StayGreen Rice like (SGRL) gene regulates chlorophyll degradation in rice. J.
Plant Physiol. 170: 1367–1373.
Sakuraba, Y., Schelbert, S., Park, S.-Y., Han, S.H., Lee, B.D., Andrès, C.B. et al.
(2012) STAY-GREEN and chlorophyll catabolic enzymes interact at
light-harvesting complex II for chlorophyll detoxification during leaf
senescence in Arabidopsis. Plant Cell 24: 507–518.
Sato, H., Yamada, T., Kita, Y., Ishimoto, M. and Kitamura, K. (2007)
Production of transgenic plants and their early seed set in Japanese
soybean variety, Kariyutaka. Plant Biotechnol. 24: 533–536.
Sato, Y., Morita, R., Nishimura, M., Yamaguchi, H. and Kusaba, M. (2007)
Mendel’s green cotyledon gene encodes a positive regulator of the
chlorophyll-degrading pathway. Proc. Natl Acad. Sci. USA 104:
14169–14174.
Sato, Y., Morita, R., Katsuma, S., Nishimura, M., Tanaka, A. and Kusaba, M.
(2009) Two short-chain dehydrogenase/reductases, NON-YELLOW
COLORING 1 and NYC1-LIKE, are required for chlorophyll b and
light-harvesting complex II degradation during senescence in rice.
Plant J. 57: 120–131.
Schelbert, S., Aubry, S., Burla, B., Agne, B., Kessler, F., Krupinska, K. et al.
(2009) Pheophytin pheophorbide hydrolase (pheophytinase) is
involved in chlorophyll breakdown during leaf senescence in
Arabidopsis. Plant Cell 21: 767–785.
Schmutz, J., Cannon, S.B., Schlueter, J., Ma, J.X., Mitros, T., Nelson, W. et al.
(2010) Genome sequence of the palaeopolyploid soybean. Nature 463:
178–183.
Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M. and Kumar, S.
(2011) MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony
methods. Mol. Biol. Evol. 28: 2731–2739.
Thomas, H. and Howarth, C.J. (2000) Five ways to stay green. J. Exp. Bot.
51(Suppl. 1): 329–337.
Yamada, T., Watanabe, S., Arai, M., Harada, K. and Kitamura, K. (2010)
Cotyledonary node pre-wounding with a micro-brush increased frequency of Agrobacterium-mediated transformation in soybean. Plant
Biotechnol. 27: 217–220.
Yamatani, H., Sato, Y., Masuda, Y., Kato, Y., Morita, R., Fukunaga, K. et al.
(2013) NYC4, the rice ortholog of Arabidopsis THF1, is involved in the
degradation of chlorophyll–protein complexes during leaf senescence.
Plant J. 74: 652–662.
Zhou, C., Han, L., Pislariu, C., Nakashima, J., Fu, C., Jiang, Q. et al. (2011)
From model to crop: functional analysis of a STAY-GREEN gene in the
model legume Medicago truncatula and effective use of the gene for
alfalfa improvement. Plant Physiol. 157: 1483–1496.
1771