Content uploaded by Wenxu Li
Author content
All content in this area was uploaded by Wenxu Li on Dec 20, 2016
Content may be subject to copyright.
ORIGINAL PAPER
Seed-specific silencing of OsMRP5 reduces seed phytic acid
and weight in rice
Wen-Xu Li •Hai-Jun Zhao •Wei-Qin Pang •
Hai-Rui Cui •Yves Poirier •Qing-Yao Shu
Received: 25 August 2013 / Accepted: 10 March 2014 / Published online: 20 March 2014
ÓSpringer International Publishing Switzerland 2014
Abstract Phytic acid (PA) is poorly digested by
humans and monogastric animals and negatively affects
human/animal nutrition and the environment. Rice
mutants with reduced PA content have been developed
but are often associated with reduced seed weight and
viability, lacking breeding value. In the present study, a
new approach was explored to reduce seed PA while
attaining competitive yield. The OsMRP5 gene, of
which mutations are known toreduce seed PA as well as
seed yield and viability, was down-regulated specifi-
cally in rice seeds by using an artificial microRNA
driven by the rice seed specific promoter Ole18. Seed
PA contents were reduced by 35.8–71.9 % in brown rice
grains of transgenic plants compared to their respective
null plants (non-transgenic plants derived from the same
event). No consistent significant differences of plant
height or number of tillers per plant were observed, but
significantly lower seed weights (up to 17.8 % reduc-
tion) were detected in all transgenic lines compared to
null plants, accompanied by reductions of seed germi-
nation and seedling emergence. It was observed that the
silencing of the OsMRP5 gene increased the inorganic P
(Pi) levels (up to 7.5 times) in amounts more than the
reduction of PA-P in brown rice. This indicates a
reduction in P content in other cellular compounds, such
as lipids and nucleic acids, which may affect overall
seed development. Put together, the present study
demonstrated that seed specific silencing of OsMRP5
could significantly reduce the PA content and increase
Pi levels in seeds; however, it also significantly lowers
seed weight in rice. Discussions were made regarding
future directions towards producing agronomically
competitive and nutritionally valuable low PA rice.
Keywords Artificial microRNA Seed specific
promoter Ole18 Low phytic acid (LPA) Seed
viability OsMRP5 Oryza sativa L.
Introduction
Myo-inositol 1,2,3,4,5,6-hexakisphosphate, known as
phytic acid (PA), is the major storage form of
Electronic supplementary material The online version of
this article (doi:10.1007/s11248-014-9792-1) contains supple-
mentary material, which is available to authorized users.
W.-X. Li H.-J. Zhao (&)W.-Q. Pang
H.-R. Cui Q.-Y. Shu (&)
State Key Laboratory of Rice Biology and Key
Laboratory of Nuclear-Agricultural Sciences of the
Ministry of Agriculture and Zhejiang Province, Institute
of Nuclear Agricultural Sciences, Zhejiang University,
Hangzhou 310029, China
e-mail: hjzhao@zju.edu.cn
Q.-Y. Shu
e-mail: qyshu@zju.edu.cn
W.-X. Li H.-J. Zhao
Wuxi Qiushi Agri-Biological Research Center, Wuxi
214105, Jiangsu, China
Y. Poirier
De
´partement de Biologie Mole
´culaire Ve
´ge
´tale, Biophore,
Universite
´de Lausanne, 1015 Lausanne, Switzerland
123
Transgenic Res (2014) 23:585–599
DOI 10.1007/s11248-014-9792-1
phosphorous (P) in cereal and legume seeds, existing
as mixed salts (phytates) of mineral cations, including
minor amounts of Zn
2?
and Fe
3?
(Lott et al. 2000;
Raboy et al. 2001). Most of the P and minerals in these
salts are not utilized for monogastric animals.
In addition, PA may also interact with minerals in
the intestinal tract and make them indigestible.
Therefore, PA is widely regarded as an anti-nutrient
in food/feed. Moreover, undigested phytic acid P (PA-
P) excreted in animal wastes has increasingly become
an important source of P pollution (Raboy 2009). To
solve the PA-related nutritional and environmental
issues, mutagenesis and transgenic approaches have
been deployed to generate low phytic acid (LPA)
phenotypes, which are largely unavailable in germ-
plasm stocks.
Rice is the staple diet for nearly two billion people
worldwide and the major food for over 50 % of
Asians. Among those 50 %, mineral micronutrient
malnutrition is a common occurrence; hence LPA rice
could be used to alleviate malnutrition in this demo-
graphic. In addition, because the by-product of rice
grains (i.e. hull and bran including pericarp, seed coat,
embryo and aleuronic cells) is an important constit-
uent of daily animal feed, reducing PA content should
also be beneficial to animal production and environ-
mental protection. Therefore, a number of LPA lines
have been developed through chemical and physical
mutagenesis (Larson et al. 2000; Liu et al. 2007; Kim
et al. 2008a,b; Li et al. 2008). However, similar to
most LPA mutants in other crops (Raboy 2009), rice
LPA mutants also appeared to have various degrees of
negative effects on certain agronomic traits, particu-
larly on grain weight and seed viability (Larson et al.
2000; Zhao et al. 2008). For example, KBNTlpa1-1
has an approximate 10 % yield penalty as compared to
its wild type (WT) progenitor Kaybonnet (KNBT) and
other rice varieties (Rutger et al. 2004). Although such
inferior performance is often observed, it might be
ameliorated through further breeding, as was proven
for field emergence of LPA soybean (Spear and Fehr
2007; Trimble and Fehr 2010).
A number of genes have been identified to be
involved in PA metabolism, and mutations of these
genes are known to cause LPA phenotypes in various
crops (Raboy 2009). Among them are the Ins(3)P
1
synthase (MIPS) gene (Hitz et al. 2002; Yuan et al.
2007), which catalyze the cyclization of D-glucose-6-
phosphate to 1D-myo-inositol-3-phosphate [Ins(3)P
1
]
(Loewus and Murthy 2000), the myo-inositol kinase
(MIK) (gene Shi et al. 2005; Kim et al. 2008b), the
inositol polyphosphate kinase (IPK) gene (Shi et al.
2003; Stevenson-Paulik et al. 2005; Yuan et al. 2012),
and the multi-drug resistance-associated protein
(MRP) ATP-binding cassette (ABC) transporter gene
(Maroof et al. 2009; Nagy et al. 2009; Panzeri et al.
2011; Shi et al. 2007; Xu et al. 2009). Furthermore,
Suzuki et al. (2007) and Josefsen et al. (2007)
demonstrated that a diverse range of enzymes, some
of which might be multifunctional, catalyse the
intermediate steps in seed phytate metabolism in rice.
However, because genes involved in PA metabolism
are also expressed in tissues other than seeds, their
knockout may affect their functions in various tissues
and organs, consequently exerting negative effects on
plant growth and yield.
To develop LPA crops without the negative effects
associated with known mutations, genetic engineering
has been explored to specifically reduce PA levels
only in phytate-accumulating seeds. For example,
mutations of MRP genes could cause significant seed
PA reductions in various crops including rice, how-
ever, the LPA phenotype is accompanied with nega-
tive effects on seed weight as well as seed germination
and field emergence (Shi et al. 2007; Maroof et al.
2009; Nagy et al. 2009; Xu et al. 2009; Panzeri et al.
2011). Because the majority of PA is accumulated in
the embryo of maize seeds and ZmMRP4 mutations
may affect endosperm development and consequently
reduce seed weight, Shi et al. (2007) downregulated
ZmMRP4 specifically in maize embryo using embryo-
specific promoters (Ole and Glb) and produced
transgenic lines with a strong LPA phenotype. Eval-
uations of T
1
transformants indicated that some LPA
lines were not significantly different from WT plants
in seed dry weight and germination rate, revealing the
potential of embryo specific silencing of ZmMRP4 in
maize improvement. In rice, PA is mainly accumu-
lated in the embryo and aleuronic layers (O’Dell et al.
1972). To produce transgenic LPA rice without the
yield penalty observed in known mutants, Kuwano
et al. (2006) first attempted to explore the promoter of
the rice seed storage protein glutelin, GluB-1 for seed
specific silencing of OsMIPS1. However, the reduc-
tion in seed PA and the increase of inorganic P (Pi) in
the transgenic line were not as great as in the LPA rice
mutant KBNTlpa1-1. They ascribed the limited effect
of the transgene to the different spatial and temporal
586 Transgenic Res (2014) 23:585–599
123
patterns between OsMIPS1 and GluB-1(Suzuki et al.
2007). Kuwano et al. (2009) further expressed an
antisense fragment of OsMIPS1cDNA under the
control of the promoter of the rice gene oleosin 18
(Ole18). In that study, only *43 % of the transgenic
events had reduced PA contents and most of trans-
genic lines had defects in seeds, but one stable line
(line O-10) was selected with a strong LPA phenotype
(*68 % PA reduction) without significant negative
effects on plant growth and seed weight. Recently, Ali
et al. (2013a,b) specifically silenced the two genes
catalyzing the first and last step of PA biosynthesis, i.e.
OsMIPS and OsIPK1, and substantially reduced
phytate contents in seeds without hampering the
growth and development of transgenic rice plants.
These results indicated that alternative and more
efficient ways should be explored for genetic engi-
neering of LPA rice.
The artificial microRNA (amiRNA) technology is
considered to be a new generation technology for gene
silencing compared with antisense technology (Yadav
and Mukherjee 2012). It has already been proven to be
effective for silencing genes in rice (Warthmann et al.
2008); therefore it is an attractive technology to
generate LPA in rice. Mutations of the rice homologue
of ZmMRP4,OsMRP5, result in seed PA reduction and
negative pleiotropic effects on seed weight, seed
germination and seedling growth in rice (Xu et al.
2009). In the present study, the effectiveness of the
amiRNA technology and the Ole18 promoter for seed
specific down-regulation of OsMRP5 was examined;
the potential of seed specific OsMRP5 silencing as
alternative means to breed competitive LPA rice was
assessed based on changes of seed P content,compo-
sition, and the agronomic performance of transgenic
lines.
Materials and methods
Plant materials
The japonica rice cultivar Nipponbare was used to
develop transgenic plants. All transgenic plants were
grown in the Experimental Farm of Zhejiang Univer-
sity. The field experiments were performed either on
Zijingang Campus, Hangzhou, Zhejiang Province
during the summer seasons of 2011 and 2012, or in
the Hainan Station (Sanya, Hainan Province) for the
winter-spring season of 2011–2012. Standard agro-
nomic practices were applied for transgenic plants,
which were similar for plants in non-transgenic fields.
For comparative studies of agronomic traits and P
contents, at least 24 transgenic plants (4 rows 96
plants/row) in T
2
or T
3
generations and their null
siblings of each line were grown side by side. Six inner
plants of each line were examined for their agronomic
traits in the field or for seed traits and P content in the
laboratory after harvest. Statistical analysis was per-
formed using Student’s ttest by using each plant as a
replicate.
DNA extraction, amplification and sequencing
All rice genomic DNA samples were extracted from leaf
tissues of seedlings or flowering plants. When used for
sequencing, genomic DNA was extracted using Biospin
plant genomic extraction kit (Bioflux, Hangzhou, China)
with treatment of RNase A (Fermentas, Canada) to
remove RNA. When used for target fragments analysis,
genomic DNA was extracted according to a modified
CTAB method as previously described (Tan et al. 2013).
All DNA samples were adjusted to a final concentration
of *25 ng/lL after quantification using the Nanodrop
2000 Spectrophotometer (Thermo Scientific, USA).
PCR primers were designed using the Primer Premier
5 software according to the genome and transcript
sequences of the japonica rice cultivar Nipponbare
(http://www.gramene.org/) and synthesized by Shanghai
Sangon Biological Engineer Technology and Services
Co., Ltd. (Shanghai, China) (Table 1). For amplification
of the Ol8 promoter, PCRs were performed in 25 lL
volumes with 50 ng DNA templates, 2.5 lL109PCR
buffer for KOD-Plus-Neo, 1.5 mM MgSO
4
,
0.2mM dNTPs, 0.5 U KOD-Plus-Neo (TOYOBO,
Japan) and 0.3 lM of each primer. A step down program
was used to increase the specificity of Ol8 promoter
amplification as follows: pre-denaturation at 98 °Cfor
2 min; 5 cycles of 98 °C for 10 s, 74 °Cfor2min;5
cycles of 98 °C for 10 s, 72 °C for 2 min; 5 cycles of
98 °C for 10 s, 70 °C for 2 min; 20 cycles of 98 °Cfor
10 s, 68 °C for 2 min, with a final extension at 68 °Cfor
5 min. For amplification of plasmid DNA, pre-denatur-
ation at 98 °Cfor2min,10 sat98°C, followed by 30
cycles of 30 s at 98 °C, 30 s at 55 °C and 30 s at 68 °C
with a final extension at 68 °C for 7 min. PCR amplicons
were separated on a 1 % agarose gel and target fragments
were cut and purified using the Axy-Prep DNA Gel
Transgenic Res (2014) 23:585–599 587
123
Extraction Kit (Vitagen, Hangzhou, China). The purified
fragments were cloned into pMD-19 T vector (Takara,
Dalian,China) for sub-cloning and sequenced at Nanjing
Genscript Biotech Co., Ltd. (Nanjing, China).
Construction of amiRNA expression vector
The Ole18 promoter region (2063 bp upstream the
?1
ATG site, LOC_Os03g04920) was amplified using
primers pOle18-F and pOle18-R (Table 1), which
introduced a HindIII and a SalI restriction site,
respectively. The target fragment was cloned into
pMD18-simple vector (Sangon, Shanghai, China) by
TA cloning. The recombinant plasmid was named as
pMD-Ole.
To prepare an amiRNA expression vector for
OsMRP5, the protocol of Warthmann et al. (2008)
and guidelines given on the WMD3 website (http://
wmd3.weigelworld.org/) were used. A 21-nt sequence
in the 2nd exon (?2,819 to ?2,839 nt; Fig. 1a) was
first identified as the target for silencing OsMRP5.
Search for amiRNA candidates with up to 5 mis-
matches turned out only one sequence that had a single
mismatch with the target sequence and with a
hybridization energy of -34.69 kcal/mol (Fig. S1).
The mismatch is not in the positions of 2 and 12 of the
amiRNA hence it was used for construction of ami-
RNA expression vector.
Four primers (MRP5-I miR-s, MRP5-IImiR-a, MRP5-
IIImiR*s, MRP5-IVmiR*s, Table 1) were obtained and
used for amiRNA production by fusion PCR with the
vector pNW55 as template (Warthmann et al. 2008). The
amiRNA fragment was inserted into the pMD18-T
(Takara, Dalian, China) vector and the recombinant
plasmid was named as pMD18-amiMRP5. To produce a
seed specific silencing vector, one of the CaMV35S
promoters in the pCAMBIA1301-35SN was first replaced
by Ole18 promoter digested from vector pMD-Ole using
HindIII and SalI, the derived vector was named as
pCAMBIA1301-OleN. Then the amiMRP5 fragment was
cut from pMD18-amiMRP5 (PstIandBamHI), and
inserted to pCAMBIA1301-OleN, and finally the ami-
RNA expression vector, p1301-amiMRP5-OleN, was
developed.
Rice transformation
The p1301-amiMRP5-OleN plasmids were transferred
into Agrobacterium tumefaciens strain EHA105 by
freeze–thaw method (Chen et al. 1994). Transgenic rice
Table 1 The primers used for promoter cloning, expression vector construction, and for the characterization of transgene
Name Forward (F) and reverse (R) sequences (50?30)
a
Product size (bp) Applications
pOle18 F: aagcttATGTCTGCCAGCATTGTGAAG
R: gtcgacTGCTAAGCTAGCTAGCAAGAT
GA
2,063 Cloning of the Ole 18 promoter
Hpt F: GCTTCTGCGGGCGATTTGTGTA
R: CGGTCGCGGAGGCTATGGATG
599 Southern blot
qRT F: CTATACTCGGCGAGATACCCAAATT
R: TTCAGGGAGCAAGCCTCAATAAC
172 qRT-PCR of OsMRP5
Actin F: TGCTATGTACGTCGCCATCCAG
R: AATGAGTAACCACGCTCCGTCA
210
Tami F: TCATCTTGCTAGCTAGCTTAGCA
R: CGGCAACAGGATTCAATCTTAA
405 Transgene identification
I miR-s agTAATTAATGCCCCTATGACCGcaggagattcagtttga For construction of amiMRP5
expression vector
II miR-a tgCGGTCATAGGGGCATTAATTActgctgctgctacagcc
III miR*s ctCGGTCTTAGCGGCATTAATTAttcctgctgctaggctg
IV miR*a aaTAATTAATGCCGCTAAGACCGagagaggcaaaagtgaa
G-4368 CTGCAAGGCGATTAAGTTGGGTAAC
G-4369 GCGGATAACAATTTCACACAGGAAACAG
a
The underlined nucleotides in lowercase are introduced restriction site for subsequent cloning; the nucleotides in lowercase are the
sequences complementary to those in the pNW55vector
588 Transgenic Res (2014) 23:585–599
123
plants were produced by Agrobacterium-mediated
transformation of the cultivar Nipponbare with hygro-
mycin as a selection agent according to Hiei and
Komari (2008). The regenerated plants were acclima-
tized inside a moist growth chamber for 1 week and
then transplanted to the field. Plants regenerated from a
common hygromycin resistant callus were recorded as
a single transgenic event.
Molecular characterization of transgenic plants
GUS staining was performed for leaf discs from young
leaves according to Jefferson et al. (1987). The presence
of transgenes were examined by PCR amplification of
target fragment using PCR primers Tami-F and Tami-R
(Table 1) designed for the particular gene fragment.
PCRs were performed in 20 lL volumes with 50 ng
genomic DNA, 10 lL29master mix (containing 29
PCR buffer, 4 mM MgCl
2
, 0.4 mM dNTPs, 50 units/ml
Taq DNA polymerase, TOYOBO Co., Ltd.), 0.4 lLof
each 10 lM primer with the following program: pre-
denaturation at 94 °C for 5 min, 30 s at 94 °C,
followed by 30 cycles of 30 s at 94 °C, 30 s at 55 °C
and 30 s at 72 °C with a final extension at 72 °Cfor
7min.
The expression level of OsMRP5 in transgenic and
Nipponbare plants was determined using quantitative
real-time polymerase chain reaction (qRT-PCR) with
the Actin gene as an internal control. Total RNA was
extracted from developing seeds of 14-day-old, leaf and
stem tissues of 45-day old plants with RNeasy Plant
Mini kit (Qiagen, Hilden, Germany) and cDNA
synthesis was carried out using PrimeScript II 1st
Strand cDNA Synthesis Kit (Takara, Dalian,China).
The relative transcript abundance was estimated by
qRT-PCR performed on a Stratagene Mx3005p Real-
Time PCR machine (Stratagene, La Jolla, CA, USA)
using SYBR
Ò
Premix Ex Taq
TM
(Takara, Dalian,
China) with the primers qPCR-F and qPCR-R (Table 1;
Fig. 1a). Quantitative RT-PCRs were performed in
triplicates with the total RNAs extracted from 3 plants.
Southern blot analysis was performed using genomic
DNA isolated from T
3
transgenic plants and controls
using a modified CTAB method (Tan et al. 2013).
Southern blot analysis was performed following Sam-
brook and Russell (2001) using the DIG High Prime
DNA Labeling and Detection Starter Kit II (Roche,
Switzerland) according to the manufacturer’s instruc-
tions. Genomic DNAs (*15 lg) were digested with
restriction enzyme (either EcoRI or HindIII), separated
on 0.8 % agarose gel, then transferred onto a nylon
membrane (Hybond-N?, Amersham Biosciences, Pis-
cataway, NJ, USA). After baking at 120 °C for 30 min,
the membrane with transferred DNA was hybridized
with a DIG labeled gene specific probe, which were
amplified by the primers Hpt-F and Hpt-F (Table 1),
and detected by CSPD (chloro-5-substituted adaman-
tyl-1,2-dioxetane phosphate) substrate by means of
autoradiography with X-ray films (Kodak, Japan).
Seed phosphorus analysis
Seed inorganic P (Pi) levels were assayed both
qualitatively and quantitatively according to the
35S’ hptII p35S2 Tnos p35S GUS TnosAmi-MRP5 pOle18
RB
bLB
Tami-R Tami-F
a
Hpt-F Hpt-R
Fig. 1 Schematic presentations of OsMRP5 structure (a) and
its amiRNA expression vector (b). aExons, introns and coding
sequences are depicted as boxes,solid lines, and filled boxes,
respectively; the amiRNA binding site is marked as double solid
lines above the 2nd exon; qPCR-F and qPCR-R are the primers
used for quantitative RT-PCR analysis of OsMRP5 transcripts.
bLB and RB are the left and right border of T-DNA; 35S’, the
CaMV 35S poly A; HptII, hygromycin phosphotransferase II
gene; p35S2, CaMV 35S promoter with a double enhancer
sequence; p35S, the 35S promoter from CaMV; pOle18, the
promoter of rice 18KDa oleosin gene; Intron, intron from the
nitrite reductase gene of Phaseolus vulgaris; GUS, beta-
glucuronidase gene; Tnos, the nopaline synthase terminator;
Ami-MRP5, the amiRNA sequence of OsMRP5 based on osa-
miR528 backbone; Tami-F and Tami-R are the forward and
reverse primers used for identification of Ami-MRP5 transgene;
Hpt-F and Hpt-R are the forward and reverse primers used for
produce the probe for Southern blot
Transgenic Res (2014) 23:585–599 589
123
microdetermination method developed by Chen et al.
(1956) with modifications. The qualitative assay was
used for identification of the high inorganic P (HIP)
phenotype. Seeds were manually cut into two parts,
one with embryo and one without embryo; the former
halves were retained for growing. The latter halves
(after removal of hulls) were transferred to 96-well
plates, extracted in 0.4 M HCl solution (10 lL per mg
sample) overnight at 4 °C; Aliquots of 10 lL super-
natant per sample were used for Pi level determination
according to Larson et al. (2000) with slight modifi-
cations (Liu et al. 2007). Development of a blue color
implies increased level of Pi (HIP), while colorless
samples typified WT levels of parent varieties (Suppl.
Fig. S1).
Seed Pi levels were also quantitatively determined
according to Wilcox et al. (2000) in triplicates as
follows. Brown rice grains were ground into rice flour
and *400 mg rice flour per sample were extracted in
12.5 % (w/v) TCA (trichloroacetic acid containing
25 mM MgCl
2
) by gentle shaking overnight at 4 °C.
After centrifugation at 15,000gfor 10 min, superna-
tants were used for Pi assay according to Raboy et al.
(2001). For determination of Pi in endosperm, brown
rice grains were first polished by milling to *35 % of
brown rice, and then crushed with a hammer into small
particles, after manually removing the remaining of
embryos if needed, for Pi assay similar to that for
brown rice flour.
PA content was determined for brown rice flour
according to Tan et al. (2013). Briefly, after drying at
60 °C for 72 h, dehulled rice grains were ground into
flour by passing through a 60-mesh sieve using a
Cyclone Sample Mill (UDY Corporation, Fort Collins,
Co., USA). Samples were stored in a refrigerator
before analysis. Ten milliliters of 0.6 M HCl was
added into a 50 ml tube with *300.0 mg flour, mixed
and incubated for 30 min in boiling water bath. After
cooling down to room temperature, the tubes were
centrifuged at 10,000gat 4 °C for 15 min. Four
hundred and fifty microliters aliquots of supernatant
were transferred to a second tube and were diluted to
4.5 ml with ddH
2
O. Aliquots of diluted supernatant
solution were passed through an IC-RP column, an IC-
H column and a 0.22 lm syringe filter (Bonna-Agela
Technologies, China). The IC-RP column was used to
get rid of hydrophobic compounds while the IC-H
column to remove residual alkaline earth metal ions,
transition metal ions and carbonate ions.
Analysis of PA was performed on Anion-exchange
Ion Chromatography ICS-2000 (Dionex, Sunnyvale,
CA, USA). Aliquots were fractionated on a Dionex-
IonPac AS11-HC analytical column, equipped with an
IonPac AS11-HC guard column and an EluGenca-
tridge KOH generator tank. The effluent was equili-
brated with 50 mM KOH at a flow rate of 1 ml/min.
PA was determined with 25 lL solution using a
conductivity detector at the suppressor current of
124 mA. An external standard of Na InsP6 (P-3168,
Sigma, St. Louis, MO, USA) was analyzed before and
after every two samples, and each genotype was
analyzed in triplicates.
Total seed phosphorus was determined according to
Hansen et al. (2009) in triplicate. Briefly, *100 mg
flour was digested with 6 ml 65 % HNO
3
and 0.2 ml
H
2
O
2
using a microwave digestion system (Micro-
wave 3000, Anton PAAR, Graz, Austria). After
digestion, the solution was carefully collected and
used for total phosphorus (TP) assay on an inductively
coupled plasma optical emission spectrometer (ICP-
OES) (Optima 8000DV, PerkinElmer, USA).
Results
Construction of amiRNA expression vector
Analysis of the rice Ole18 promoter sequence (Gen-
bank Accession No. AY427563, 1249 bp)isolated by
Qu and Takaiwa (2004) and used by Kuwano et al.
(2009)fortheseed-specificsilencingofMIPS1
revealed a discrepancy at the 50end with the genome
sequence of Nipponbare (http://www.gramene.org/).
Therefore, we designed new primers (pOle18-F/R,
Table 1) and amplified a 50flanking sequence of Ole18
(2063 bp) which was subsequently used as a seed
specific promoter for amiRNA expression of OsMRP5.
For amiRNA expression vector construction, the
candidate amiRNA sequences of OsMRP5 were first
identified using the Designer function available at the
WMD3-Web MicroRNA Designer (http://wmd3.
weigelworld.org/). From the high ranking candidates,
one amiRNA sequence located in the 2nd exon
(Fig. 1a) was chosen according to the previously
described by Warthmann et al. (2008). The four prim-
ers, i.e. MRP5-I miR-s, MRP5-IImiR-a MRP5-III-
miR*-a, and MRP5-IVmiR*-s (Table 1), designed for
the selected amiRNA sequence, the primers G-4368
590 Transgenic Res (2014) 23:585–599
123
and G-4369 were used for engineering the amiRNA
sequence into the rice MIR528 (osa-miR528) precursor
in the vector pNW55 through site-directed mutagene-
sis. Then,the OsMRP5 amiRNA sequence was trans-
ferred into pCAMBIA1301-OleN under the control of
promoter Ole18 to generated the final expression vector
(Fig. 1b).
Production of transgenic plants and high inorganic
P lines
One to several plantlets was regenerated from each of
36 independent initial calluses, which were survived
the hygromycin screening after Agrobacterium-med-
iated transformation. After acclimatization, the plant-
lets were transplanted in the paddy field and grown to
maturity, and plants from a common initial callus were
classified as one independent transgenic event (T
0
).
Leaf GUS assay indicated that 31/36 events were
positive (stained in blue color). The transgenic plants
were further examined for the presence of the
amiRNA transgene by PCR using the Tami primers
(Table 1; Fig. 1b) and plants of 27/36 independent T
0
events (Fig. 2a and data not shown) were positive.
These results indicated that the amiRNA fragment was
highly likely to be integrated into the rice genomes in
these plants. Not long after flowering, a brown plant
hopper outbreak occurred in the experimental field,
which unfortunately also affected our transgenic
plants, particularly those regenerated and transplanted
at a later stage. In addition, two lines appeared to be
completely sterile. Consequently, only 17 independent
events which contained the amiRNA transgene pro-
duced T
1
seeds.
Low seedling emergence rates were observed for T
1
seeds and only 1–14 mature plants were obtained from
20 T
1
seeds sown for each line (Table S1). The T
2
seeds were harvested from each T
1
plant and 8 seeds of
bp
500
400
M+ 12 3 456 78 910111213141516
23.1
9.4
6.6
4.4
(Kb)
EcoRI
HindIII
+ - + -
0.0
1.0
2.0
3.0
4.0
5.0
Seeds Endosperm Leaf Stem
Relative expression (RQ=2
Ct
)
Nipponbare
#18N
#18T
#05N
#05T
**
**
S
a
bc
Fig. 2 Identification and molecular characterization of trans-
genic plants with OsMRP5 amiRNA. aPCR analysis of T
0
transgenic plants with primer Tami-F and Tami-R; M,
100 bp DNA ladder; ?, control plasmid of amiRNA expression
vector; -, Nipponbare; 1–16, individual T
0
transgenic plants.
bSouthern blot analysis of T
3
transgenic lines. Genomic DNA
(15 lg per lane) of non-transgenic plants (NT) and transgenic
lines, digested with EcoRI or HindIII, was separated on an
agarose gel, blotted and hybridized with the Hpt probe; ?
control plasmid of amiRNA expression vector; -, Nipponbare.
The positions and sizes (kb) of markers are indicated on the left.
cQuantitative RT-PCR analysis of OsMRP5 in transgenic rice
seeds (14 days after flowering), endosperm of mature seeds, and
leaf and stem tissues of 45-day old plants. Transgenic lines #18T
and #05T were obtained from Nipponbare transformed with an
Ole18-driven amiRNA sequence of OsMRP5. Relative expres-
sion (RQ =2
-DDCt
) were calculated using housekeeping gene
Actin as internal control and that of Nipponbare leaf tissues set
as ‘1’ (marked with ‘S’). Significant differences between
transgenic lines and their respective null siblings were marked
with an asterisk (P\0.05). Each value represents the
mean ±standard error of three replicates
Transgenic Res (2014) 23:585–599 591
123
each plant were tested for their Pi levels using
colorimetric assay. Among the 17 T
1
lines, two lines
produced T
2
seeds of WT Pi level, indicating that the
amiRNA transgene was lost or was not properly
expressed; the other lines had at least one plant
producing T
2
seeds with a high Pi level (Fig. S2;
Table S1).
For evaluating the effect of OsMRP5 silencing on
agronomic performance, one transgenic and one null
line were developed from each event when possible.
The sibling transgenic and null lines are distinguished
by a suffix of ‘T’ (for transgenic) or a ‘N’ (for null),
e.g. line #01T and #01N stand for the homozygous
transgenic line and homozygous null line from event
#01, respectively. Through Pi analysis of T
2
and T
3
seeds, six homozygous transgenic lines were identified
and used for subsequent assessment together with their
null siblings. The other lines remained segregating for
Pi levels and were not analyzed further.
Seed phosphorus
To analyze the effect of gene silencing on seed P level
and composition, the six independent homozygous
transgenic lines were assessed for their TP, Pi, and PA-
P content, together with their null siblings and
Nipponbare (Table 2). Significant differences were
observed for the contents of Pi, PA-P, and TP in seeds
of Nipponbare produced in Hangzhou and Hainan,
indicating the existence of environmental effects on
seed P content (Table 2). All transgenic lines had
substantially and significantly higher levels of Pi and
lower levels of PA-P compared to their respective null
siblings, which had levels similar to Nipponbare
(Table 2). The results indicated that seed PA biosyn-
thesis was effectively reduced in transgenic lines,
irrespective of growing environments. The TP levels
of three out of five lines tested were significantly less
than that of their respective null siblings, while two
other lines showed no significant differences
(Table 2).
Nipponbare and null lines had Pi levels less than
0.49 mg/g (#06N), while all transgenic lines had Pi
contents ranging from 1.79 to 2.54 mg/g, which
amount to 4.9–7.5 times of their respective null
siblings (Table 2). Nipponbare seeds produced in
Hangzhou had PA-P content of 1.77 mg/g, which was
much less than those produced in Hainan (Table 2).
For the seeds grown in the same site, null lines had PA-
P contents similar to that of Nipponbare (Table 2). In
Table 2 Contents of TP, inorganic P (Pi) and phytic acid P (PA-P) of homozygous transgenic lines (T) and their null siblings (N)
Materials Pi PA-P TP (mg/g) (Pi ?PA-P)/TP
mg/g T/N mg/g (N-T)/N
Brown rice samples (T
3
), harvested from T
2
plants grown in Sanya, Hainan (November 2011 to April 2012)
#06T 2.40 ±0.26 4.9 0.87 ±0.14 62.5 % 4.24 ±0.04* 77.1 %
#06N 0.49 ±0.03 2.32 ±0.08 4.48 ±0.03 62.7 %
#10T 1.95 ±0.02 5.4 1.27 ±0.07 47.7 % NT –
#10N 0.36 ±0.01 2.44 ±0.03 NT –
#15T 2.54 ±0.00 5.6 0.80 ±0.03 65.4 % 3.93 ±0.07* 85.0 %
#15N 0.45 ±0.01 2.21 ±0.10 4.18 ±0.02 63.9 %
#18T 1.79 ±0.23 4.9 1.42 ±0.08 35.8 % 4.31 ±0.03 72.4 %
#18N 0.35 ±0.01 2.21 ±0.12 4.24 ±0.42 60.4 %
#24T 2.30 ±0.01 5.2 0.69 ±0.10 71.9 % 4.39 ±0.01* 68.1 %
#24N 0.44 ±0.01 2.45 ±0.16 4.64 ±0.05 62.3 %
Nipponbare 0.38 ±0.02 2.35 ±0.14 4.41 ±0.09 61.9 %
Brown rice samples (T
4
), harvested from T
3
plants grown in Hangzhou, Zhejiang (May to October 2012)
#05T 1.79 ±0.02 7.5 0.76 ±0.14 57.7 % 3.56 ±0.04 71.6 %
#05N 0.24 ±0.03 1.81 ±0.13 3.65 ±0.03 56.2 %
Nipponbare 0.32 ±0.01 1.77 ±0.02 3.37 ±0.18 62.0 %
Data with an asterisk are significantly different from those of their respective null siblings (P\0.05)
NT not tested
592 Transgenic Res (2014) 23:585–599
123
sharp contrast, all transgenic lines had PA-P levels
substantially lower than their respective null siblings,
with reduction rates ranging from 35.8 to 71.9 %
(Table 2). Furthermore, the increase of Pi was more
than the reduction of PA-P in transgenic lines, and
consequently the proportion of (Pi ?PA-P)/TP was
increased (Table 2). These results suggest a potential
reduced incorporation of P into other compounds, such
as DNA and lipids, to the increase in free Pi content.
Integration of transgenes and seed specific gene
silencing
To confirm transgenes had already been integrated
into rice genome, two homozygous lines, i.e. #05T and
#018T, with PA reductions of 35.8 and 57.7 %
respectively, were analyzed by Southern blot. Results
showed that one single copy was integrated into the
rice genome in both lines (Fig. 2b).
To examine the seed specificity of gene silencing,
the abundances of OsMRP5 transcripts were also
analyzed for the two transgenic lines through qRT-
PCR. While the expressionof OsMRP5 was observed in
all tested tissues, its transcripts were more abundant in
developing seeds than in others (Fig. 2c). No significant
differences were observed among transgenic plants,
their null siblings and Nipponbare in stems and leaves.
However, significant reductions were observed in
developing seeds of transgenic lines, e.g. *55 % in
#18T and *70 % in #05T, respectively, when com-
pared with their respective null siblings (Fig. 2c). The
results indicated that the Ole18 promoter effectively
drove the expression of amiRNA of OsMRP5 specif-
ically in seeds.
To assess whether the Ole18 promoter driven
OsMRP5 silencing also affected the expression of
OsMRP5 and P metabolism in endosperm, rice grains
(T
4
) were polished to remove seed pericarps, aleurone
layers and embryos by milling to the degree of *35 %
and the expression of OsMRP5 and Pi content was
measured for 2 transgenic lines and their respective null
lines. The results showed that the expression of OsMRP5
in endosperm of mature seeds was significantly reduced
(by 60–70 %) in the transgenic lines compared with their
respective null lines (Fig. 2c). The Pi level in the
endosperm of Nipponbare was only *1/3 of that in
brown rice; but the Pi level in transgenic line #15T and
#18T is *6and*3 times that of their respective null
lines. This indicates that the P metabolism in endosperm
was also affected in the transgenic lines (Fig. 3).
Agronomic performance of transgenic lines
The homozygous transgenic and null sibling lines were
assessed for plant height, number of panicles per plant,
and seed weight. Considerable differences were
observed between different lines. However, only two
transgenic lines had significant different plant heights
compared with their respective null siblings (one higher
—#18T and one shorter—#24T), and none of transgenic
lines was significantly different from their respective
null siblings in the number of tillers per plant (Table 3).
These results, although they are very preliminary in
nature due to the limited number of plants analyzed,
indicated that the seed specific silencing of OsMRP5 did
not affect plant growth. However, significantly lower
seed weight was observed in all transgenic lines
compared with their respective null siblings (Table 3),
demonstrating that the OsMRP5 amiRNA also affects
seed development. Both the transgenic and null plants of
line #06 had shorter plant height compared with other
lines and Nipponbare, which may explain their limited
differences of seed weight (Table 3).
Further analysis indicated that seed weight was
significantly negatively correlated with Pi level
(R
2
=-0.923, P\0.01) and the proportion of
(Pi ?PA-P)/TP (-0.844, P\0.01), and significantly
positively correlated with PA level (0.849, P\0.01),
but not the TP levels (Suppl. Table S2).
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
#05 #015 #18 Nipponbare
Iorganic P (mg g-1)
Materials
Null
Transgenic
*
*
*
Fig. 3 Contents of inorganic P in endosperm of homozygous
transgenic lines and their null siblings. Transgenic lines were
obtained from Nipponbare transformed with an Ole18-driven
amiRNA sequence of OsMRP5. Data with an asterisk are
significantly different from those of their respective null siblings
(P\0.05). Each value represents the mean ±standard error of
6 milled rice grains
Transgenic Res (2014) 23:585–599 593
123
To evaluate the effect of gene silencing on seed
germination and growth, three lines with sufficient T
4
seeds were tested together with Nipponbare. Overall
Nipponbare performed better even compared with the
null lines; the transgenic seeds of lines #15 and #18
had significantly lower germination rates than their
null siblings, but no significant differences were
observed for transgenic and null line #05 (Fig. 4a).
The differences of field seedling emergence rates
became more substantial, with all transgenic seeds
having significantly lower emergence than their
respective null siblings (Fig. 4b).
Discussion
With the aim to increase the nutritional quality and
reduce the environmental impacts of PA in rice grains
while attaining competitive yields, a new approach to
reduce PA specifically in rice seeds was explored in
the present study. By using the amiRNA technology
and the rice Ole18 promoter, the expression OsMRP5
was downregulated and consequently reduced the PA
content by 35.8–71.9 % in brown rice. Comparative
analyses of transgenic lines with their respective null
siblings indicated that the silencing of OsMRP5 does
not affect plant height and number of tillers per plant,
but it does cause significant seed weight reduction, and
consequently affects seed germination and seedling
emergence. The results indicated that OsMRP5 may
play important roles in seed biology other than in PA
metabolism, and its seed specific silencing could still
exert negative effects on rice grain yield and seed
viability.
Seed specificity and efficiency of gene silencing
In rice seeds, the hydrophobic protein oleosin (molec-
ular mass, 18 kDa) is abundantly localized on the
surface of oil bodies, the spherical lipid storage
organelles (Wu et al. 1998). Oil bodies are stored in
the embryo and aleurone layer of rice grains, just as the
globoids in which PA is stored. GUS expression
analysis demonstrated that the promoter of the gene
for 18-kDa oleosin, Ole18, specifically drives gene
expression from the early stages of seed development
in the whole embryo and the aleurone layer, not in the
endosperm or vegetative tissues (Qu and Takaiwa,
2004). Our qRT-PCR analysis proved that reduction of
OsMRP5 transcripts was only observed in developing
seeds, not in leaf or stem tissues (Fig. 2c), which
indicated that Ole18 promoter is suitable for driving
amiRNA expression specifically in rice seeds.
New gene silencing technologies, such as virus-
induced gene silencing (VIGS), hairpin RNA interfer-
ence (hpRNAi), and amiRNAs have been developed
and applied in both gene function studies and practical
genetic engineering during the past decade (Yadav and
Mukherjee 2012). Compared with the antisense tech-
nology so far applied in engineering of LPA crops
(Feng and Yoshida 2004;Kuwanoetal.2006; Nunes
et al. 2006; Shi et al. 2007;Kuwanoetal.2009), the
amiRNA technology employed in the present study is
considered to be a 2
nd
generation technology (Yadav
and Mukherjee 2012). Although only one amiRNA
sequence was deployed in the present study, all 17
transgenic events, except two, produced transgenic
plants with the expected phenotype—high-Pi seeds
Table 3 Agronomic traits of homozygous transgenic lines and
their null siblings
Materials PH (cm) NPPP HSW (g)
T
2
plants grown in Sanya, Hainan (November 2011 to April
2012)
#06T 52.6 ±2.3 14.0 ±3.4 2.03 ±0.10*
#06N 54.8 ±3.3 10.3 ±3.3 2.47 ±0.12
#10T 64.8 ±3.5 13.0 ±5.8 2.07 ±0.12*
#10N 65.4 ±2.5 10.5 ±2.7 2.46 ±0.06
#15T 58.1 ±4.2 13.3 ±4.5 2.05 ±0.10*
#15N 63.1 ±6.3 11.2 ±3.8 2.42 ±0.09
#18T 62.9 ±2.5 12.8 ±4.4 2.04 ±0.07*
#18N 58.6 ±3.9 14.0 ±5.8 2.42 ±0.09
#24T 57.8 ±2.0* 9.5 ±3.6 2.05 ±0.06*
#24N 62.8 ±3.4 11.2 ±5.0 2.47 ±0.06
Nipponbare 68.3 ±2.7 11.0 ±3.8 2.54 ±0.09
T
3
plants grown in Hangzhou, Zhejiang (May to October 2012)
#05T 74.2 ±2.1 10.0 ±2.2 2.25 ±0.06*
#05N 76.3 ±3.4 8.8 ±2.3 2.57 ±0.10
#15T 77.6 ±1.7 12.6 ±0.9 2.21 ±0.05*
#15N 78. 3 ±3.9 12.7 ±1.8 2.64 ±0.09
#18T 85.7 ±2.9* 12.7 ±2.3 2.30 ±0.04*
#18N 78.0 ±3.3 12.5 ±2.3 2.48 ±0.5
Nipponbare 86.6 ±2.7 9.8 ±1.6 2.30 ±0.19
Data with an asterisk are significantly different from those of
their respective null siblings (P\0.01)
PH plant height, NPPP number of panicles per plant, HSW
hundred seed weight
594 Transgenic Res (2014) 23:585–599
123
(Table S1). This is in sharp contrast with the results of
OsMIPS1 gene silencing reported by Kuwano et al.
(2009), where they only identified 16 transgenic lines
showing the high-Pi phenotype among the 37 T
0
lines
developed by using antisense technology. These results
suggested that the amiRNA technology could be a more
efficient tool than antisense technology for LPA rice
engineering.
In maize, by using the Ole16 and Glb promoters,
Shi et al. (2007) effectively silenced ZmMRP4 in
embryos and generated transgenic lines with increased
Pi and reduced PA (68–87 % for transgenic lines of
Ole::MRP4 and 32–75 % of Glb::MRP4 lines). In
rice, both the Ole18 promoter and the GluB-1promoter
could drive tissue specific siblings of OsMIPS1 and
reduce PA levels specifically in seeds.The PA reduc-
tion was more significant when the Ole18 promoter
was used (Kuwano et al. 2006,2009), indicating the
similarities between maize and rice. Most transgenic
lines of Ole18::OsMIPS1 generated by Kuwano et al.
(2009) were reported to be defective in seed develop-
ment and only one stable line was obtained with
significant PA reduction, indicating OsMIPS1 may
play a crucial role in seed development. In the present
study, stable transgenic lines were obtained from
about half of the transgenic events (not including those
damaged by brown plant hoppers), indicating that
OsMRP5 might be a better candidate gene in general
for manipulation of LPA rice breeding. PA reductions
in brown rice grains of transgenic lines produced in the
present study ranged from 35.8 to 71.9 % (Table 2),
which is similar to those observed in transgenic
Glb::MRP4 lines in maize (Shi et al. 2007). However
it is difficult to compare our results with those of
Kuwano et al. (2009) because only one line was
analyzed in detail in their study.
Alteration of seed P composition, seed weight
and viability: biological basis?
Because mutations of the indigenous multidrug resis-
tance-associated protein ABC transporter gene not
only cause seed PA reduction but also significantly
affect seed development and yield (Liu et al. 2007;Shi
et al. 2007; Maroof et al. 2009; Nagy et al. 2009;Xu
et al. 2009; Panzeri et al. 2011), a tissue specific
silencing approach was proposed by Shi et al. (2007).
In their study, ZmMRP4 was suppressed using the
embryo-specific promoters Ole16 and Glb, a strong
LPA phenotype ([70 % PA reductions) was obtained
without significant decrease of seed dry weight and
germination rate in two transgenic lines with the Ole16
promoter and more lines with the Glb promoter (Shi
et al. 2007).
In transgenic plants three types of genetic effects
could exist, i.e. the effect of the transgene (amiRNA of
50
55
60
65
70
75
80
85
90
95
100
Seed germination(%)
Materials
Null
Transgenic
*
*
50
55
60
65
70
75
80
85
90
95
100
#05 #15 #18 Nipponbare #05 #15 #18 Nipponbare
Seedling emgernce(%)
Materials
Null
Transgenic
*
*
*
ab
Fig. 4 Seed germination (a) and seedling emergence (b) rate of
homozygous transgenic lines and their null siblings. Transgenic
were obtained from Nipponbare transformed with an Ole18-
driven amiRNA sequence of OsMRP5. Data with an asterisk are
significantly different from those of their respective null siblings
(P\0.05). Each value represents the mean ±standard error of
three replicates
Transgenic Res (2014) 23:585–599 595
123
OsMRP5 in this study), the disruption of an endoge-
nous rice gene by the integration of T-DNA, and
somaclonal variations generated during the process of
transgenic plants production (Shu et al. 2002). How-
ever, only the effects of the former would be consistent
among different transgenic plants, while those of the
latter two are random and would be inconsistent
among different transgenic events (Shu et al. 2002).
The considerable differences of plant height, number
of tillers per plant and seed weight among different
events indicated the existence of somaclonal varia-
tions (Table 3). Therefore homozygous transgenic
plants were compared with their respective null
siblings for each event in the present study to assure
that the differences observed are due to the effect of
amiRNA. Significant negative impacts on seed weight
were observed in all transgenic lines compared with
their respective null siblings, even for line #18T with
PA reduction of only 35.8 % (Tables 2,3), which
indicated the OsMRP5 amiRNA expressor does have
negative effects. Therefore the strategy reportedly
working in maize unfortunately did not work in rice.
The compositional changes of seed P and the
inferior performance of agronomic traits observed in
the transgenic lines could be due to both targeted and
untargeted effects of the OsMRP5 amiRNA. A WMD3
search did reveal that the transcript Os12g05880.1
could be the potential off-target of the OsMRP5
amiRNA, because they had only 5 mismatch (Fig. S1).
However, in-depth analysis showed that there is a
mismatch at position 9 of the amiRNA when aligned
with Os12g05880.1 (Fig. S1), and it is known that no
mismatch between the positions 2 and 9 of amiRNA is
tolerated for effective gene silencing (Yadav and
Mukherjee 2012), hence the off-target effect of the
OsMRP5 amiRNA could be excluded in principle.
Therefore the differences consistently observed
between transgenic lines and their respective null
siblings are highly likely due to the targeted effect of
OsMRP5 amiRNA, and the following reasons may
explain the negative effect on agronomic performance
observed in the present study.
Firstly, while the maize Ole16 gene is only
expressed in embryo (Shi et al. 2007), the rice Ole18
gene was reported to be expressed in both aleurone
layers and embryos (Qu and Takaiwa 2004). Our
present study further indicated the Ole 18 driven
silencing of OsMRP5 was indeed extended to endo-
sperm (Fig. 2c). Therefore the gene silencing modes
driven by the two promoters would be different in rice
and maize. It is yet unknown whether additional
silencing of OsMRP5 in aleurone layers and endo-
sperm would affect seed development, but if it is the
case, the Ole16 promoter driven silencing would
become advantageous due to its restriction to embryos
in maize and explain to some extent the differences
observed in rice and maize regarding the impact of
silencing on seed weight.
Secondly, although PA is mainly stored in embryos
and aleurone layers in rice seeds (O’Dell et al. 1972),
milling analysis showed that low levels of PA and Pi are
present in milled rice (Ren et al. 2007). In the present
study, elevated Pi levels were observed in milled rice of
transgenic lines compared with their respective null
siblings (Fig. 3), which indicated the effect of OsMRP5
silencing was not limited to aleurone layers and
embryos, but was extended to endosperm. Because Pi
is a known inhibitor of starch biosynthesis enzymes
(Preiss 1997), the elevated Pi levels in transgenic lines
could affect starch synthesis and consequently lead to
the reduction of endosperm (seed) weight (Smidansky
et al. 2002,2003). The increased Pi content in
endosperm could be the result of Pi diffusion from the
aleurone layers, where Pi was dramatically increased
due to silencing of OsMRP5. It could also be the direct
result of silencing of OsMRP5 as evidenced by its
reduced transcripts abundance in endosperm (Fig. 2c).
Further studies are needed to clarify whether the
reduced OsMRP5 abundance in endosperm was due
to the effects of amiRNAs generated in the endosperm,
or moved in from aleurone layers/embryos because it is
known that microRNAs are mobile among tissues
(Yadav and Mukherjee 2012).
Thirdly, the silencing of OsMRP5 not only
impaired PA biosynthesis, it also likely affected other
aspects of P metabolism. Our results indicated that the
amount of Pi increase was more than that of PA-P
reduction in transgenic lines, which increased the
proportion of Pi ?PA-P to TP (Table 2). This change
indicates a reduction in the amount of P present in
other cellular components, such as P in membrane
lipids and DNAs, which could certainly affect the cell
metabolism and ultimately affect seed development.
This phenomenon was not observed in OsMIPS and
OsIPK1 silenced transgenic lines reported by Ali et al.
(2013a,b), which suggested that the proportion change
of Pi ?PA-P to TP is unique to the function of
OsMRP5 in rice.
596 Transgenic Res (2014) 23:585–599
123
The seed specific silencing of OsMRP5 further
reduced seed viability. It is noted that the reduction of
seed germination and field emergence was not
proportional to the reduction of seed weight; rather it
seems to be related to the increase of (Pi ?PA)/TP
rate (Table 2), which should be further investigated
with more materials when sufficient seeds become
available.
Seed specific silencing for production of low
phytic acid rice: the way forward
Seed specific silencing of genes involved PA metab-
olism has been proposed a strategy to avoid or
minimize the negative effects that have been observed
in LPA mutants induced by chemical and physical
mutagenesis (Raboy 2009). This technology involves
the selection of proper target gene, the use of a proper
promoter, and the deployment of an efficient gene
silencing approach. For seed specific silencing, the
Ole18 promoter (a counterpart of Ole16 in maize) has
been so far the best choice (Ali et al. 2013a,b; Kuwano
et al. 2009) in rice, which is also consolidated by the
results of our present study. Gene silencing could be
achieved by antisense (Kuwano et al. 2009), hairpin
RNA (Ali et al. 2013a,b) or amiRNA (this study) and
the latter one is generally considered to be advanta-
geous over the other two (Yadav and Mukherjee
2012). Compared with previous reports where only
one or two independent transgenic lines were selected
(Ali et al. 2013a,b; Kuwano et al. 2009), five stable
transgenic lines were selected in the present study,
which indicates that the amiRNA is at least equally
effective, if not better, compared with the other
approaches.
By silencing the expression of MIPS1 specifically in
seeds Kuwano et al. (2009) identified one stable line that
had significantly reduced PA contents in seeds without
defects of any observed agronomic traits. Similar results
were reported by Ali et al. (2013b), but they revealed
that the transgenic seeds are very sensitive to ABA
treatment as compared to respective non-transgenic
control ones. Because sensitivity to ABA would lead to
inhibition of germination in presence of ABA hence
they concluded that MIPS1 is not suitable for production
of transgenic LPA rice (Ali et al. 2013b). For the
development of perfect LPA rice, another gene involved
in PA biosynthesis, the inositol 1,3,4,5,6-pentakisphos-
phate 2-kinase (IPK1) gene, was silenced using the same
approach for MIPS1 (Ali et al. 2013a).Basedonthe
results of detailed characterization of two independent
transgenic lines, they concluded that OsIPK1is the
choice for producing LPA rice by seed specific gene
silencing because the transgenic lines performed as
good as WT ones, including seed viability and tolerance
to artificial aging (Ali et al. 2013a). It is known that
soybean IPK1 mutants had no negative impacts on
agronomic traits and seed viability (Yuan et al. 2007,
2012), but its mutations would simultaneously increase
the content of lower inositol phosphates, e.g. inositol
1,3,4,5,6-pentakisphosphate, which also is an anti-
nutrient because it can still chelate minerals such as
Zn
2?
and Fe
3?
, therefore it is too early to conclude that
seed specific silencing OsIPK1 represents an ideal
approach for the development of LPA rice of both
nutritional value and agronomic competitiveness.
In summary, the present study demonstrated that
seed specific silencing of OsMRP5 using amiRNA
technology and the Ole18 promoter could significantly
reduce seed PA content, but it also significantly lowers
seed weight in rice, which suggested that this approach,
as well as approach reported earlier, may not be good
enough for practical application in LPA rice breeding,
hence new strategies should be further explored.
Acknowledgments The research was financially supported by
the Natural Science Foundation of China through research grant
No. 31071481, and in part by the Sino-Swiss Joint Research
Project (2009 DFA32040 to QS and IZLCZ3 123946I to YP)
and by Wuxi Science and Technology.Department (Grant
#CYES1002), Zhejiang Provincial Innovation Team of
Nuclear Agricultural Science and Technology (2010R50033).
We are grateful to Dr. Yuwei Shen of DNA LandMarks for his
critical comments on and improvement of the manuscript.
Technical assistance of Ms. Lijuan Mao for measurement of
phytic acid content is highly appreciated.
References
Ali N, Paul S, Gayen D, Sarkar SN, Datta K, Datta SK (2013a)
Development of low phytate rice by RNAi mediated seed-
specific silencing of inositol 1,3,4,5,6-pentakisphosphate
2-kinase gene (IPK1). PLoS ONE 8(7):e68161. doi:10.
1371/journal.pone.0068161
Ali N, Paul S, Gayen D, Sarkar SN, Datta SK, Datta K (2013b)
RNAi mediated down regulation of myo-inositol-3-phos-
phate synthase to generate low phytate rice. Rice 6:12.
doi:10.1186/1939-8433-6-12
Chen PS, Toribara TY, Warner H (1956) Microdetermination of
phosphorous. Anal Chem 28:1756–1758
Transgenic Res (2014) 23:585–599 597
123
Chen H, Nelson RS, Sherwood JL (1994) Enhanced recovery of
transformants of Agrob-acterium tumefaciens after freeze-
thaw transformation and drug selection. Biotechniques
16:664–670
Feng XG, Yoshida KT (2004) Molecular approaches for pro-
ducing low-phytic-acid grains in rice. Plant Biotechnol
21:183–189
Hansen TH, Laursen KH, Persson DP, Pedas P, Husted S,
Schjoerring JK (2009) Micro-scaled high-throughput
digestion of plant tissue samples for multi-elemental ana-
lysis. Plant Methods 5:12
Hiei Y, Komari T (2008) Agrobacterium-mediated transfor-
mation of rice using immature embryos or calli induced
from mature seed. Nat Protoc 3:824–834
Hitz WD, Carlson TJ, Kerr PS, Sebastian SA (2002) Bio-
chemical and molecular characterization of a mutation
that confers a decreased raffinosaccharide and phytic
acid phenotype on soybean seeds. Plant Physiol 128:
650–660
Jefferson RA, Kavanagh TA, Bevan MW (1987) GUS fusions:
beta-glucuronidase as a sensitive and versatile gene fusion
marker in higher plants. EMBO J 6:3901–3907
Josefsen L, Bohn L, Sorensen MB, Rasmussen SK (2007)
Characterization of a multifuncti-onal inositol phosphate
kinase from rice and barley belonging to the ATP-grasp
superfamily. Gene 397:114–125
Kim SI, Andaya CB, Goyal SS, Tai TH (2008a) The rice OsLpa1
gene encodes a novel protein involved in phytic acid
metabolism. Theor Appl Genet 117:769–779
Kim SI, Andaya CB, Newman JW, Goyal SS, Tai TH (2008b)
Isolation and characterization of a low phytic acid rice
mutant reveals a mutation in the rice orthologue of maize
MIK. Theor Appl Genet 117:1291–1301
Kuwano M, Ohyama A, Tanaka Y, Mimura T, Takaiwa F,
Yoshida KT (2006) Molecular breeding for transgenic rice
with low-phytic-acid phenotype through manipulating
myo-inositol 3-phosphate synthase gene. Mol Breed
18:263–272
Kuwano M, Mimura T, Takaiwa F, Yoshida KT (2009) Gen-
eration of stable ‘low phytic acid’ transgenic rice through
antisense repression of the 1D-myo-inositol 3-phosphate
synthase gene (RINO1) using the 18-kDa oleosin promoter.
Plant Biotechnol J 7:96–105
Larson SR, Rutger JN, Young KA, Raboy V (2000) Isolation
and genetic mapping of a non-lethal rice (Oryza sativa L.)
low phytic acid 1 mutation. Crop Sci 40:1397–1405
Li CY, Park DS, Won SR, Hong SK, Ham JK, Choi JK, Rhee HI
(2008) Food chemical properties of low-phytate rice cul-
tivar, Sang-gol. J Cereal Sci 47:262–265
Liu QL, Xu XH, Ren XL, Fu HW, Wu DX, Shu QY (2007)
Generation and characterization of low phytic acid germ-
plasm in rice (Oryza sativa L.). Theor Appl Genet
114:803–814
Loewus F, Murthy PPN (2000) Myo-Inositol metabolism in
plants. Plant Sci 150:1–19
Lott JNA, Ockenden I, Raboy V, Batten GD (2000) Phytic acid
and phosphorus in crops seeds and fruits: a global estimate.
Seed Sci Res 10:11–33
Maroof MAS, Glover NM, Biyashev RM, Buss GR, Grabau EA
(2009) Genetic basis of the low-phytate trait in the soybean
line CX1834. Crop Sci 49:69–76
Nagy R, Grob H, Weder B, Green P, Klein M, Frelet-Barrand A,
Schjoerring JK, Brearley C, Martinoia E (2009) The
Arabidopsis ATP-binding cassette protein AtMRP5/AtA-
BCC5 is a high affinity inositol hexakisphosphate trans-
porter involved in guard cell signaling and phytate storage.
J Biol Chem 284:33614–33622
Nunes AC, Vianna GR, Cuneo F, Amaya-Farfan J, de Capde-
ville G, Rech EL, Aragao FJ (2006) RNAi-mediated
silencing of the myo-inositol-1-phosphate synthase gene
(GmMIPS1) in transgenic soybean inhibited seed devel-
opment and reduced phytate content. Planta 224:125–132
O’Dell BL, de Boland AR, Koirtyohann SR (1972) Distribution
of phytate and nutritionally important elements among the
morphological components of cereal grains. J Agric Food
Chem 20:718–721
Panzeri D, Cassani E, Doria E, Tagliabue G, Forti L, Campion
B, Bollini R, Brearley CA, Pilu R, Nielsen E, Sparvoli F
(2011) A defective ABC transporter of the MRP family,
responsible for the bean lpa1 mutation, affects the regula-
tion of the phytic acid pathway, reduces seed myo-inositol
and alters ABA sensitivity. New Phytol 191:70–83
Preiss J (1997) Modulations of starch synthesis. In: Foyer CH,
Quick QP (eds) A molecular approach to primary metab-
olism in higher plants. Taylor and Francis, London,
pp 81–104
Qu LQ, Takaiwa K (2004) Evaluation of tissue specificity and
expression strength of rice seed component gene promoters
in transgenic rice. Plant Biotechnol J 2:113–125
Raboy V (2009) Approaches and challenges to engineering seed
phytate and total phosphorus. Plant Sci 177:281–296
Raboy V, Young KA, Dorsch JA, Cook A (2001) Genetics and
breeding of seed phosphorus and phytic acid. J Plant
Physiol 158:489–497
Ren XL, Liu QL, Fu HW, Wu DX, Shu QY (2007) Density
alteration of nutrient elements in rice grains of a low
phytate mutant. Food Chem 102:1400–1406
Rutger JN, Raboy V, Moldenhauer KAK, Bryant RJ, Lee FN,
Gibbons JW (2004) Registration of KBNT lpa1-1 low
phytic acid germplasm of rice. Crop Sci 44:363
Sambrook J, Russell DW (2001) Molecular cloning: a labora-
tory manual, 3rd edn. Cold Spring Harbor Laboratory
Press, Cold Spring Harbor
Shi JR, Wang HY, Wu YS, Hazebroek J, Meeley RB, Ertl DS
(2003) The maize low-phytic acid mutant lpa2 is caused by
mutation in an inositol phosphate kinase gene. Plant
Physiol 131:507–515
Shi JR, Wang HY, Hazebroek J, Ertl DS, Harp T (2005) The
maize low-phytic acid 3 encodes a myo-inositol kinase that
plays a role in phytic acid biosynthesis in developing seeds.
Plant J 42:708–719
Shi JR, Wang HY, Schellin K, Li BL, Faller M, Stoop JM,
Meeley RB, Ertl DS, Ranch JP, Glassman K (2007)
Embryo-specific silencing of a transporter reduces phytic
acid content of maize and soybean seeds. Nature Bio-
technol 25:930–937
Shu QY, Cui HR, Ye GY, Wu DX, Xia YW, Gao MW, Altosaar
I (2002) Agronomic and morphological characterization of
agrobacterium-transformed but rice plants. Euphytica
127:345–352
Smidansky ED, Clancy M, Meyer FD, Lanning SP, Blake NK,
Talbert LE, Giroux MJ (2002) Enhanced ADP-glucose
598 Transgenic Res (2014) 23:585–599
123
pyrophosphorylase activity in wheat endosperm increases
seed yield. Proc Natl Acad Sci USA 99:1724–1729
Smidansky ED, Martin JM, Hannah LC, Fischer AM, Giroux
MJ (2003) Seed yield and plant biomass increases in rice
are conferred by deregulation of endosperm ADP-glucose
pyrophosphorylase. Planta 216:656–664
Spear JD, Fehr WR (2007) Genetic improvement of seedling
emergence of soybean lines with low phytate. Crop Sci
47:1354–1360
Stevenson-Paulik J, Bastidas RJ, Chiou ST, Frye RA, York JD
(2005) Generation of phytate-free seeds in arabidopsis
through disruption of inositol polyphosphate kinases. Proc
Natl Acad Sci USA 102:12612–12617
Suzuki M, Tanaka K, Kuwano M, Yoshida KT (2007) Expres-
sion pattern of inositol phosphate-related enzymes in rice
(Oryza sativa L.): implications for the phytic acid bio-
synthetic pathway. Gene 405:55–64
Tan YY, Fu HW, Zhao HJ, Lu S, Fu JJ, Li YF, Cui HR, Shu QY
(2013) Functional molecular markers and high-resolution
melting curve analysis of low phytic acid mutations for
marker-assisted selection in rice. Mol Breed 31:517–528
Trimble LA, Fehr WR (2010) Genetic improvement of seedling
emergence of low-phytate soybean lines. Crop Sci
50(1):67–72
Warthmann N, Chen H, Ossowski S, Weigel D, Herve
´P (2008)
Highly specific gene silencing by artificial miRNAs in rice.
PLoS ONE 3:e1829
Wilcox JR, Premachandra GS, Young KA, Raboy V (2000)
Isolation of high seed inorganic P, low-phytate soybean
mutants. Crop Sci 40:1601–1605
Wu LS, Wang LD, Chen PW, Chen LJ, Tzen JT (1998) Genomic
cloning of 18 kDa oleosin and detection of triacylglyc-
erols and oleosin isoforms in maturing rice and postger-
minative seedlings. J Biochem 123:386–391
Xu XH, Zhao HJ, Liu QL, Fank T, Engel KH, An G, Shu QY
(2009) Mutations of the multi-drug resistance-associated
protein ABC transporter gene 5 result in reduction of phytic
acid in rice seeds. Theor Appl Genet 119:75–83
Yadav P, Mukherjee SK (2012) Artificial microRNA and its
applications. In: Mallick B, Ghosh Z (eds) Regulatory
RNAs. Springer, New York, pp 505–521
Yuan FJ, Zhao HJ, Ren XL, Zhu SL, Shu QY (2007) Generation
and characterization of two novel low phytate mutations in
soybean (Glycine max L. Merr.). Theor Appl Genet
115:945–957
Yuan FJ, Zhu DH, Tan YY, Dong DK, Fu XJ, Zhu SL, Li BQ,
Shu QY (2012) Identification and characterization of the
soybean IPK1 ortholog of a low phytic acid mutant reveals
an exon-excluding splice-site mutation. Theor Appl Genet
125:1413–1423
Zhao HJ, Liu QL FUHW, Xu HX, Wu DX, Shu QY (2008)
Effect of non-lethal low phytic acid mutations on grain
yield and seed viability in rice. Field Crops Res
108:206–211
Transgenic Res (2014) 23:585–599 599
123
