See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/340817283

Knockout of MULTI-DRUG RESISTANT PROTEIN 5 genes lead to low phytic

acid contents in oilseed rape

Article  in  Frontiers in Plant Science · April 2020

DOI: 10.3389/fpls.2020.00603

CITATIONS READS

10 111

2 authors:

Niharika Sashidhar Christian Jung

Max Planck Institute for Plant Breeding Research Christian-Albrechts-University of Kiel
6 PUBLICATIONS   151 CITATIONS    251 PUBLICATIONS   7,670 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Genetic analysis of storage root formation in sugar beet View project

Unravelling genetic mechanisms of flowering time control in Quinoa View project

All content following this page was uploaded by Niharika Sashidhar on 28 May 2020.

The user has requested enhancement of the downloaded file.

 ORIGINAL RESEARCH
published: 26 May 2020
doi: 10.3389/fpls.2020.00603

Knockout of MULTI-DRUG
RESISTANT PROTEIN 5 Genes Lead
to Low Phytic Acid Contents in
Oilseed Rape
Niharika Sashidhar, Hans J. Harloff and Christian Jung*
Plant Breeding Institute, Agrar und Ernährungswissenschaftliche Fakultät, Christian-Albrechts-University of Kiel, Kiel,
Germany

Understanding phosphate uptake and storage is interesting to optimize the plant

performance to phosphorus fluctuations. Phytic acid (PA) is the major source of
inorganic phosphorus (Pi) in plants. Genetic analyses of PA pathway transporter
genes (BnMRP5) and their functional characterization might provide clues in better
utilizing the available phosphate resources. Furthermore, the failure to assimilate PA
by monogastric animals results in its excess accumulation in manure, which ultimately
causes groundwater eutrophication. As a first step toward breeding low PA mutants in
Edited by: oilseed rape (Brassica napus L.), we identified knockout mutants in PA biosynthesis and
Nicolas Rispail,
Spanish National Research Council,
transporter genes. The obtained M3 single mutants of Bn.MRP5.A10 and Bn.MRP5.C09
Spain were combined by crossing to produce double mutants. Simultaneously, crosses were
Reviewed by: performed with the non-mutagenized EMS donor genotype to reduce the background
Qingyao Shu,
mutation load. Double mutants identified from the F2 progeny of direct M3 crosses and
Zhejiang University, China
Francesca Sparvoli, BC1 plants showed 15% reduction in PA contents with no significant differences in
Italian National Research Council, Italy Pi. We are discussing the function of BnMRP5 paralogs and the benefits for breeding
*Correspondence: Bnmrp5 mutants in respect to low PA, yield, and stress tolerances.
Christian Jung
c.jung@plantbreeding.uni-kiel.de Keywords: ATP binding cassette, Brassica napus, BnMRP5, lpa, phosphorous, TILLING, EMS, mutant

Specialty section:
This article was submitted to INTRODUCTION
Plant Breeding,
a section of the journal Phosphorus is an essential macronutrient required for plant development. In plants, numerous
Frontiers in Plant Science transporters such as PHT, SPX-MFS, and phosphate antiporters facilitate the uptake of phosphorus
Received: 24 February 2020 from the soil and its mobilization across various tissues (Kopriva and Chu, 2018). The up taken
Accepted: 21 April 2020 phosphorus is utilized in various metabolic processes and is finally stored in seed vacuoles in the
Published: 26 May 2020
form of phytic acid (PA; also referred as inositol hexakis phosphate), which is readily available for
Citation: seedling development. The PA stored in vacuoles is referred to as phytin, which is a complex salt
Sashidhar N, Harloff HJ and of divalent ions. In dry mature seeds of various plant species, up to 80% of total phosphorus is
Jung C (2020) Knockout
stored as PA (Raboy et al., 2000). Therefore, low PA mutants are desirable for a reduced application
of MULTI-DRUG RESISTANT
PROTEIN 5 Genes Lead to Low
of external phosphorus in the form of fertilizers. Furthermore, they could help to decrease the
Phytic Acid Contents in Oilseed Rape. dependency on non-renewable rock phosphates (Lott et al., 2007). In this regard, PA biosynthetic
Front. Plant Sci. 11:603. genes have been knocked out in different plant species to obtain low PA mutants with simultaneous
doi: 10.3389/fpls.2020.00603 increase of inorganic phosphorus (Pi) (Sparvoli and Cominelli, 2015). In some cases, the knockout

Frontiers in Plant Science | www.frontiersin.org 1 May 2020 | Volume 11 | Article 603

Sashidhar et al. Bnmrp5 Shows Low Phytic Acid

of genes involved in PA biosynthesis has proven to have day conditions (16 h of light and 8 h of dark). No external
adverse effects on plant performances such as poor seed set, phosphorus was added to the soil. Three weeks old plants were
delayed germination, and low yield (Shi et al., 2007). Therefore, vernalized for 8 weeks at 4◦ C (16 h light and 8 h dark) and
targeting the transporters that are accumulating PA in seeds after 8 weeks plants were transferred back to 22◦ C (16 h light
gained more attention. and 8 h dark) until harvest. At each time point, five plants were
A multi-drug resistant protein (MRP5), acting as a tonoplast harvested. During flowering, young buds were emasculated and
transporter was shown to have high affinity to PA (Nagy et al., hand-pollinated with Express 617 pollen to mark the day of
2009). MRP5 is a member of the ATP binding cassette (ATP) pollination. For PA and Pi extractions, developing seeds were
super family and belongs to the ABCC subfamily (Kang et al., collected 15, 25, 35, 45, and 55 days after pollination (DAP)
2011). The protein has two transmembrane domains and two and frozen in liquid nitrogen. Samples were kept at −80◦ C
nucleotide-binding domains encasing the walker A, B, and until extraction.
C motifs (Verrier et al., 2008). It has been shown that in M3 seeds were obtained from the breeding company
Arabidopsis the transport of PA is strictly dependent on ATP Norddeutsche Pflanzenzucht Hans-Georg Lembke KG
(Nagy et al., 2009) whereas in rice knockout of SULTR-like (Hohenlieth, Germany). M3 plants homozygous for the
phosphorus distribution transporter (SPDT) genes lead to 25–32% Bn.MRP5.A10 allele (M3 seed code 170482) were crossed with
reduced PA contents in seeds (Yamaji et al., 2016). This indicates homozygous Bn.MRP5.C05 mutants (M3 seed code 170481) to
that the vacuolar loading of phosphorus in seeds is not limited to produce F1 plants. Simultaneously, homozygous M3 plants from
the MRP5 protein but can also be achieved by other transporters. each family were crossed with non-mutagenized Express 617
Furthermore, the accumulation of PA in the vacuoles depends on plants to obtain F1 offspring with reduced background mutation
the availability of Pi, as it is highly important to maintain the ratio load. All F1 plants were selfed to produce F2 seeds (Figure 1).
of vacuolar to cytoplasmic Pi for overall phosphorus homeostasis
(Pratt et al., 2009). Identification of MRP5 Genes and
Oilseed rape (Brassica napus L.) is the third most important oil Mutant Screening
crop in the world and its seed meal is rich in proteins and amino In Arabidopsis, the MRP5 protein is encoded by one gene
acids (Gacek et al., 2018). However, anti-nutritive compounds (At1g0420), which was used for retrieving all paralogous genes
such as glucosinolates and tannins together with a high fiber from B. napus using the Genoscope browser1 . Genes having high
content impede its use as a valuable protein source for human and sequence identity to the Arabidopsis genes were considered as
animal feed. Also, PA has an anti-nutritive effect because it is a true paralogous genes. Protein domains were identified using the
strong chelator of positive ions like magnesium, potassium, zinc, pFam database search2 . All hits were confirmed using TAIR3 ,
and iron, thereby reducing their bioavailability in monogastric NCBI annotations4 and BRAD databases5 .
animals. Moreover, it contributes to phosphate pollution because We have used TbyS for screening the mutations as described
PA cannot be digested due to lack of phytases thereby leading in Sashidhar et al. (2019). Two BnMRP5 paralogs were chosen
to high P concentrations in the manure. Several knockout and for identifying the putative loss of function mutations. Four
knockdown mutants of PA transporter genes were identified in primer pairs were used to screen amplicons 1414–1509 bp in size.
cereal crops such as maize, rice, wheat, and barley. Their seed PA Sanger sequencing was used to confirm the mutations in the M2
contents were reduced by 32–90% with a simultaneous increase generation with the help of paralog specific primers.
of Pi (Dorsch et al., 2003; Shi et al., 2007; Xu et al., 2009; Panzeri Mutation frequencies (F) were calculated based on the
et al., 2011; Bhati et al., 2016). In addition, MRP mutants were mutations per M1 plant (Harloff et al., 2012):
identified from common bean and soybean, which also showed a
substantial reduction of PA contents by up to 90% (Wilcox et al.,
(amplicon size [bp]) ∗ (number of M1 plants)
 
2000; Shi et al., 2007; Campion et al., 2009).
F[1/kb] = 1/
We expected that MRP genes display a similar function in (number of mutations) ∗ 1,000
rapeseed. Therefore, we aimed to select MRP mutants with
reduced PA seed storage. We identified MRP orthologs in
the rapeseed genome and used the sequence information to Nucleic Acid Isolation, PCR, and
select EMS mutants by a novel TILLING by Sequencing (TbyS)
approach (Sashidhar et al., 2019). Only double mutants showed
RT-qPCR
For genotyping experiments, DNA was isolated from the leaves
a significant reduction in seed PA content. These mutants will be
using the CTAB protocol (Rogers and Bendich, 1985) with minor
important for breeding low PA rapeseed varieties.
modifications. PCR was performed using DNA aliquots of 2 µl
as a template for amplifying the genes (92◦ C: 3 min, [92◦ C: 30 s,
MATERIALS AND METHODS 58–63◦ C: 30 s, 72◦ C: 65 s] 36x, 72◦ C: 5 min). Amplicon lengths
1
http://www.genoscope.cns.fr/brassicanapus/
Plant Material and Growth Conditions 2
http://pfam.xfam.org/
The rapeseed winter type variety Express 617 was used to study 3
https://www.arabidopsis.org/servlets/Search?type=gene&action=new_search
PA and Pi accumulation in developing seeds. Plants were grown 4
https://www.ncbi.nlm.nih.gov/gene
in a greenhouse in 11 cm × 11 cm pots at 22◦ C under long 5
http://brassicadb.org/brad/

Frontiers in Plant Science | www.frontiersin.org 2 May 2020 | Volume 11 | Article 603

Sashidhar et al. Bnmrp5 Shows Low Phytic Acid

FIGURE 1 | Crossing scheme for the production of BnMRP5 single and double mutants. The seed codes are written above each genotype (see Table 3 for
respective allele codes). The genotypes highlighted in green were selfed to produce F3 seeds, which were used for phenotyping; “S”: selfing; Bn.MRP5.A10 and
Bn.MRP5.C05 mutations are indicated as mu1 and mu2, respectively.

were verified on 1% agarose gels and the presence of mutations 1.5 ml with double distilled water and incubated at 40◦ C for 1 h.
was confirmed by Sanger sequencing (Table 1). The samples were measured against a reagent blank buffer in a
For expression analysis, the winter type line Express 617 spectrophotometer at 655 nm. A calibration curve ranging from
was grown in the greenhouse as described above. Five plants 0 to 100 nmol was used for both PA and Pi measurements with
were used for expression analysis. Immature flower buds were standards for Pi and PA, respectively.
emasculated followed by hand pollination to ensure the day
of pollination. Seed samples of 50 mg were harvested at 15, Statistical Analysis
25, 35, and 45 DAP and were shock frozen in liquid nitrogen. Data obtained from PA and Pi measurements were used for
RNA was isolated using the peqGOLD RNA isolation kit statistical analysis. Significance was calculated using ANOVA in
(PEQLAB Biotechnologie GmbH, Erlangen, Germany) according R 3.6.0 and the MulticomView package and post hoc test was
to the manufacturer’s protocol. DNase digestion was performed performed by using Tukey’s multiple comparison test; p = 0.05.
after RNA isolation using the peqGOLD DNase I Digest Standard deviation was calculated between five biological samples
Kit (PEQLAB Biotechnologie GmbH, Erlangen, Germany) to along with their three technical repeats.
remove genomic DNA. cDNA first strand synthesis was carried
out using a first strand cDNA synthesis kit (ThermoFisher
Scientific Inc., Waltham, MA, United States) according to RESULTS
the manufacturer’s protocol. RT-qPCR was performed using
Platinum SYBR Green qPCR SuperMix-UDG with ROX
R R

Dynamics of Phytic Acid and Inorganic

(Invitrogen, Karlsruhe, Germany) on a CFX96 Real-Time PCR
detection system (Bio-Rad, Hercules, CA, United States) in a Phosphorous Accumulation in Winter
volume of 20 µl. BnActin2 was used as a reference gene for Type Rapeseed
normalization. Relative expression was calculated using the 1Ct Seed samples from 15, 25, 35, 45, and 55 DAP were used to
method (Table 1). determine the dynamic changes of PA and Pi in developing
seeds. We observed that PA was not measurable at 15 DAP, but
Inorganic Phosphorus and Phytic Acid it accumulated significantly from 15 to 35 DAP. In contrast,
Measurements Pi contents decreased significantly from 15 to 35 DAP. The
coordinated decrease of Pi and increase of PA might indicate
Dry seeds of 200 mg were ground into fine powder and PA was
that most of the Pi is assimilated to PA between 35 and
extracted in three replicates of 50 mg according to Matthäus et al.
45 DAP (Figure 2).
(1995) with minor modifications. PA was measured by HPLC
according to Rounds and Nielsen (1993) calibrating with a PA
standard (Sigma P-8810). For determining Pi contents, 50 µl of Searching the Rapeseed Genome for
the purified column extracts was mixed with 500 µl of coloring MRP5 Orthologs
reagent [10% w/v of ascorbic acid (Roth-Art.Nr.3666.1) and 5% We used the Arabidopsis AtMRP5 sequence as query to search
w/v of ammonium molybdate (Roth, Art. No.3525.2)] adjusted to the Brassica Genoscope database. Four rapeseed sequences were

Frontiers in Plant Science | www.frontiersin.org 3 May 2020 | Volume 11 | Article 603

Sashidhar et al. Bnmrp5 Shows Low Phytic Acid

RT-qPCR primer
RT-qPCR primer
RT-qPCR primer
TILLING primer
TILLING primer
TILLING primer
TILLING primer
Purpose
Amplicon length (bp)

1414
1458
1434
1509
285
156
279

Green letters indicate SNPs between paralogs and red letters indicate modified SNP to increase the specificity. Primer design is done according to Liu et al. (2012).

FIGURE 2 | Dynamics of phytic acid (PA) and inorganic phosphorus (Pi)

Temperature (◦ C)

accumulation during seed maturation in the winter type Express 617. Plants
were grown in a greenhouse under 16 h of light and 8 h of dark at 22◦ C. Five
61.0
63.0
62.0
61.0
59.5
60.0
62.0

plants were used as biological replicates for PA and Pi measurements. Error

bars denote the SD among five biological samples along with three technical
repeats. Statistical significance was calculated using ANOVA and grouping
was done using a Tuckey test. Different letters indicate statistical significance.
GGCTACAAATTCTCGGTTCTTTGC + CACAACAAGCTCTCTGGAGTCTATG
ACGTGCGATTCTCAAGTCCCTT + GGATCGTCTAGTAAATAAATGTCACCG
GAATATGCTTAGAAGTGTGTTCGGA + CTAACAATGCGAACCAGTTCTCTT

identified as putative paralogous genes of AtMRP5 (Table 2). All

CGTCGTTATAGCTTGTAGAATCCACC + CTCTTCCTGAACCTTAGCCGAC
TGGCTTGGGCGAATCCTTAG + GAATGTTCTTCAAGAGGGTCAAGACTC
CTTGGGCGAATCCTCAAGCT + GAATGTTCTTCAAGAGGGTCAAGACTA

AAGGTGTACTTGTCATACATGCGA + CTACTTTAGACTGGTCGCCTTCA

genes showed Walker A, B, and C motifs, which were encased

by the transmembrane domain (Supplementary Figure S1).
The BLAT results from Genoscope browser showed that
Bn.MRP5.A09 had a lower homology and also a shorter protein
sequence when compared to the three other proteins. Upon
protein alignment with AtMRP5, we found that Bn.MRP5.A09
lacked most of the transmembrane domain regions, indicating
that this protein might lack a membrane transport function for
PA. Therefore, we excluded Bn.MRP5.A09 from our analyses
(Supplementary Figure S2).
Messenger RNA abundance was measured in developing
seeds of Express 617 and Ct values were normalized to the
housekeeping gene BnActin2. The transcriptional activities of all
three genes strongly increased between 15 and 25 DAP. They
Primer sequence

peaked between 25 and 35 DAP and decreased 45 DAP (Figure 3).

Finally, the two highest expressed paralogs (Bn.MRP5.A10,
TABLE 1 | Primers used for genotyping and expression analysis.

Bn.MRP5.C05) were used for mutation screening in the EMS

mutant population.

Identification of EMS Mutants and

Production of Double Mutants
NS_P210 + 211
NS_P250 + 251
NS_P208 + 209
NS_P252 + 253

We applied a new TbyS protocol to search for EMS mutants

NS_P99 + 100
Primer name

NS_P33 + 34
NS_P35 + 36

in the Express 617 mutant population (Sashidhar et al., 2019).

We screened for mutations within the translated regions of
Bn.MRP5.A10 (5284 bp) and Bn.MRP5.C05 (5278 bp). The
amplicons used for TbyS covered up to 60% of the coding regions
of Bn.MRP5.A10 and Bn.MRP5.C05, including the Walker A and
C motifs of the genes (Figure 4). As a result, eight stop codon
Paralog name

Bn.MRP5.C05
Bn.MRP5.C05

Bn.MRP5.C05
Bn.MRP5.C09
Bn.MRP5.A10
Bn.MRP5.A10
Bn.MRP5.A10

and three splice site mutations were found in Bn.MRP5.A10

while 10 stop codon and one splice site mutation were found
in Bn.MRP5.C05 (Table 3). All mutations were confirmed
by Sanger sequencing confirming the reliability of the TbyS

Frontiers in Plant Science | www.frontiersin.org 4 May 2020 | Volume 11 | Article 603

Sashidhar et al. Bnmrp5 Shows Low Phytic Acid

TABLE 2 | Gene structure and amino acid similarity of four MRP5 paralogs in Brassica napus L.

Arabidopsis B. napus sequence B. napus Transcript Exon/intron structure Protein Amino acid sequence
gene annotation paralogs length (bp) size (aa) identity with AtMRP5 (%)*
Exons Introns

At1g0420 BnaA10g02400D Bn.MRP5.A10 5284 9 8 1509 93.0

BnaC05g02300D Bn.MRP5.C05 5278 9 8 1510 93.6
BnaC09g10390D Bn.MRP5.C09 5399 12 11 1436 85.3
BnaA09g10230D Bn.MRP5.A09 4168 13 12 1071 84.7

Transcript length is from start to stop codon. *% identity is obtained by using BLAT function from the Genoscope browser.

Bnmrp5 Double Mutants With Reduced

Phytic Acid Seed Content
We measured the effect of single and double mutations with
full and reduced mutation load. Plants from M3 , F2 , and
BC1 generations were grown in the greenhouse. Express 617
and mutant offspring of Express 617 (suffix “e”) were used
as controls. Single mutants from the two F2 families, 180823
(J1 J1 ) and 180850 (K1 K1 ), showed no significant reduction in
PA contents when compared to both Express 617 controls
(Figure 5). However, double mutants obtained from F2 family
180844 (J1 J1 K1 K1 ) showed significant reduction in comparison
to the mutant control plants carrying homozygous Express
617 alleles (Je Je Ke Ke ) but not to non-mutagenized Express 617
(Figure 6). This could be explained due to a high mutation load.
Therefore, we expected clear phenotypic effects in the double
mutants backcrossed with Express 617 (see Figure 1). As a
result, we observed a significant PA reduction in backcrossed
FIGURE 3 | Expression profiles of three BnMRP5 genes in Express 617 at
seed filling stages. Plants were grown at 16 h of light and 8 h of dark at 22◦ C. double mutants from the F2 population 190690 in comparison
Seed samples from four developmental stages of Express 617 were harvested to the internal control plants (Je Je Ke Ke ) as well as to the Express
at 15, 25, 35, and 45 DAP. The expression values were normalized to the 617 plants. We concluded that a reduction of 15% was due to
housekeeping gene BnActin2. Error bars indicate SD of five biological the underlying mutant alleles and not due to the background
samples along with three technical repeats.
mutation load. Indeed, a similar reduction was observed in
the double mutants of the 180844 family when compared to
internal control plants (Je Je Ke Ke ), indicating that the observed
strategy (Sashidhar et al., 2019). For ease of understanding, reduction of 15% PA is due to the two mutated alleles. In addition,
mutations in Bn.MRP5.A10 and Bn.MRP5.C05 were assigned we measured the Pi contents in matured seeds of the 190690
with the allele code J and K, respectively. For phenotyping, progeny to see if the reduced PA content had any effect on Pi
we chose the alleles J1 and K1 , as these mutations were at accumulation. We did not find a significant increase in double
the beginning of the gene and the translation should result mutants as compared to control plants (Figure 6). Besides, plant
in a loss of all the important protein domains (Table 3 and height, silique number per plant, and thousand-kernel weight
Figure 4). were not significantly different in the mutants, indicating that the
It was known from previous studies that single mutants low PA was not affecting the overall plant performance (Table 5).
in rapeseed often do not display a new phenotype due to
gene redundancy. Moreover, EMS mutants suffer from a huge
number of background mutations, which severely hampers the DISCUSSION
phenotypic evaluation (Braatz et al., 2018; Shah et al., 2018).
Therefore, we initiated a crossing program to produce double In this study, we describe low PA mutants in oilseed rape by
mutants and backcross generations with reduced mutation targeting BnMRP5 PA transporter genes. Three paralogs were
load. Homozygous M3 mutants were crossed with non-mutated expressed during seed development, two showed a maximum
Express 617 to obtain F1 seeds. Then, the F1 plants carrying expression between 25 and 35 DAP similar to the expression of
mutations in both the selected copies were crossed with each MRP5 genes in siliques of Arabidopsis (AtMRP5, Kim and Tai,
other (Figure 1) and F2 populations were generated to be 2011) and common bean (PvMRP1, measured 6 and 24 days after
used for PA measurements. In the F2 generation, segregation flowering (Panzeri et al., 2011). The expression pattern of the
ratios were as expected for mono (3:1) and digenic (15:1) paralogs was in line with the PA and Pi accumulation pattern
segregation (Table 4). from Express 617. Shortly after 35 DAP, PA storage reached its

Frontiers in Plant Science | www.frontiersin.org 5 May 2020 | Volume 11 | Article 603

Sashidhar et al. Bnmrp5 Shows Low Phytic Acid

FIGURE 4 | Exon–intron structure showing the positions of the mutations. Exons are shown as white boxes. Primer pairs used for amplicon sequencing are shown
as black and yellow arrows. C motif, Walker A, and Walker B motif are shown as yellow, orange, and blue boxes, respectively. The C motif is also known as ABC
signature motif. Stop codon and splice site mutations are shown as black and red triangles, respectively. The mutant alleles are indicated by respective numbers (see
Table 3).

TABLE 3 | All Brassica napus BnMRP5 mutant alleles detected by TILLING by sequencing (TbyS).

Gene Mutation type Mutation site (bp)§ Allelea Mutation Effect on protein sequence

Bn.MRP5.A10 Non-sense 1077$ J1 G to A W379*

Bn.MRP5.A10 Non-sense 1072 J2 C to T Q378*
Bn.MRP5.A10 Non-sense 1242 J3 G to A W414*
Bn.MRP5.A10 Non-sense 1260 J4 G to A W420*
Bn.MRP5.A10 Non-sense 1561 J5 C to T Q521*
Bn.MRP5.A10 Splice site 3557 J6 G to A –
Bn.MRP5.A10 Splice site 4093 J7 G to A –
Bn.MRP5.A10 Non-sense 4152 J8 G to A W1216*
Bn.MRP5.A10 Non-sense 4271 J9 G to A W1255*
Bn.MRP5.A10 Splice site 4390 J10 G to A –
Bn.MRP5.C05 Non-sense 432 K2 G to A W144*
Bn.MRP5.C05 Non-sense 628$ K1 C to T Q210*
Bn.MRP5.C05 Non-sense 1174 K3 C to T Q392*
Bn.MRP5.C05 Non-sense 1244 K4 G to A W415*
Bn.MRP5.C05 Non-sense 3487$ K5 C to T R1044*
Bn.MRP5.C05 Non-sense 3762 K6 C to T Q1116*
Bn.MRP5.C05 Non-sense 4206 K7 C to T Q1239*
Bn.MRP5.C05 Non-sense 4245 K8 C to T Q1252*
Bn.MRP5.C05 Splice site 4299 K9 G to A –
Bn.MRP5.A10 – – Je – –
Bn.MRP5.C05 – – Ke – –
§ Mutation site from the start codon. a A single letter code was used to display M2 genotypes where “e” denotes the allele of non-mutagenized Express 617 donor variety.
$ Two plants from the EMS population had similar mutation. *Indicates a stop codon.

maximum and Pi depletion its minimum suggesting a major mutants and double mutants with their non-mutagenized
incorporation of Pi into PA until 35 DAP. parent Express 617. Therefore, we could not observe
The double mutants obtained displayed up to 15% significant differences between primary double mutants
reduced PA contents. One of the major concerns regarding (J1 J1 K1 K1 , F2 population 180844) and Express 617.
EMS mutants is their high mutation load. This problem However, double mutants once backcrossed with Express
became obvious when comparing primary (non-backcrossed) 617 (F2 population 190690) had significantly reduced PA

Frontiers in Plant Science | www.frontiersin.org 6 May 2020 | Volume 11 | Article 603

Sashidhar et al. Bnmrp5 Shows Low Phytic Acid

TABLE 4 | Genotypic segregation for Bn.MRP5.A10 and Bn.MRP5.C05 in 4 F2 populations.

F2 seed code Expected segregation pattern Gene Number of plants Chi square
analyzed value (χ2 )

180850 J1 J1 : Je Je : J1 Je Bn.MRP5.A10 50 4.2a

180823 K1 K1 :Ke Ke : K1 Ke Bn.MRP5.C05 50 3.45a
180844 J1 J1 K1 K1: J1 J1 Ke Ke: Je Je K1 K1: Je Je Ke Ke: J1 J1 K1 Ke : Bn.MRP5.A10; 230 3.66b
J1 Je K1 K1 : J1 Je K1 Ke : Je Je K1 Ke : J1 Je Ke Ke Bn.MRP5.C05
190690 J1 J1 K1 K1: J1 J1 Ke Ke: Je Je K1 K1: Je Je Ke Ke: J1 J1 K1 Ke : Bn.MRP5.A10; 200 4.4b
J1 Je K1 K1 : J1 Je K1 Ke : Je Je K1 Ke : J1 Je Ke Ke Bn.MRP5.C05
a 3:1 segregation, χ2 < 5.99, at p = 0.05, df = 2. b 15:1 segregation, χ2 < 7.82, at p = 0.05, df = 3. Genotypes shown in green were used for phenotyping experiments.

FIGURE 5 | Phenotypic analysis of Bnmrp5 mutants. Box plot of phytic acid contents in seeds of Bnmrp5 mutants. Statistical analysis was performed using ANOVA
test at p < 0.05 and grouping was done using the Tukey test. Same letters indicate no significant differences. The non-mutated Express 617 alleles are indicated by
“e.” Bn.MRP5.A10 and Bn.MRP5.C05 mutations are indicated as mu1 and mu2, respectively. Seed codes of F2 families are given above the brackets. Means are
indicated by “+.” Allele codes are given in Table 3.

contents compared to their mutant siblings (Je Je Ke Ke ) and compensate the mutation in OsPLDα1 (Khan et al., 2019). This
compared to Express 617. genetic compensation response has first been described after
We expected that the two knockout mutants resulted in a mutant mRNA degradation for genes with sequence similarities
truncated protein, where the Walker A, B, and C motifs were to the mutated genes in zebrafish (El-Brolosy et al., 2019).
completely lost, thereby leading to a failure to transport PA. Whether this applies also for the mutation of the two MRP5
However, the reduction in PA content was somewhat lower as genes we have studied will be an interesting direction for
in other studies with maize, rice, wheat, barley, Arabidopsis, further studies.
soybean, and common bean where reductions between 32 and In polyploids, the simultaneous knockout of multiple genes
90% have been reported (Sparvoli and Cominelli, 2015). One is required to obtain a phenotypic effect (Wang et al., 2014;
reason for a relatively lower reduction in our study could be that Del Pozo and Ramirez-Parra, 2015). In wheat, a reduction
the third gene might partially contribute to the compensation in PA content by “only” 22% was observed when all three
of the phenotypic effect in the double mutant. In rice it has paralogous genes of TaABCC5 were downregulated by RNAi
been shown that a mutation in OsPLDα1 caused a reduction (Bhati et al., 2016). Therefore, this study and our study might
of PA in seeds, however, in the mutant other PLD genes were indicate the existence of alternative transporters in polyploids,
upregulated, indicating that the other PLD proteins can in part which might compensate for the loss of knockout mutations

Frontiers in Plant Science | www.frontiersin.org 7 May 2020 | Volume 11 | Article 603

Sashidhar et al. Bnmrp5 Shows Low Phytic Acid

et al., 2016), suggest its contribution to PA loading in seeds

(Sacchi and Nocito, 2019). Since, loss of function mutants in
PA transporters were never identified before in oilseed rape,
the hypothesis that other MRP transporters compensate the
knockout effects need to be further validated for instance
by additional mutations in the third paralog Bn.MRP5.C09.
Moreover, they can be crossed with bn2-pgk2 mutants which
we have been recently identified (Sashidhar et al., 2019). Like
the bnmrp5 mutants, bn2-pgk2 mutants did not show any
trade off effects regarding agronomically important characters.
Therefore, creating multiple mutants in both gene families could
be a promising option to further reduce PA seed contents
in oilseed rape.
Knock out mutants of MRP5 in Arabidopsis showed a
major reduction in PA contents. Therefore, we assumed that
this might also be the case in BnMRP5 mutants of the close
relative B. napus. PA transporter mutants in various species do
have higher inositol kinases (Vip/VIH) referred to as inositol
hexakis phosphate kinases (Desai et al., 2014), which synthesize
InsP7 (also referred as PP-InsP5 ) and InsP8 (also referred as
1,5 PP2 -InsP4 ; Supplementary Figure S3). For instance, the
FIGURE 6 | Inorganic P (Pi) measurements in Bnmrp5 double mutants of the
knockout mutants of atmrp5 and zmmrp4 showed an increased
F2 family 190690. The boxplot represents Pi measurements in five plants for accumulation of InsP7. Loss of function in MRP5 might therefore
each genotype from matured seeds. Of each plant, three technical replicates lead to an increased concentration of InsP6 (or PA) in the cytosol,
were taken. Means are indicated by “+.” Statistical analysis was performed which can be used as a substrate by the inositol hexakis phosphate
using ANOVA test at p < 0.05 and grouping was done using the Tukey test.
kinases to produce InsP7 or InsP8, resulting in a lower PA
Same letters indicate no significant differences between the samples. The
non-mutated Express 617 alleles are indicated by “e.” Bn.MRP5.A10 and storage rate. Accumulating InsP7 and InsP8 might be beneficial
Bn.MRP5.C05 mutations are indicated as mu1 and mu2, respectively Allele to plants as they possess energy-rich pyrophosphate moieties
codes are given in Table 3. having a comparable free energy of hydrolysis to ATP (Hand and
Honek, 2007; Desai et al., 2014). It was shown in Arabidopsis
that InsP7/InsP8 binds to SPX domains, which are crucial in
in MRP5 proteins. A study in common bean has shown phosphate sensing and in regulating Pi homeostasis (Jung et al.,
that loss of PvMRP1 was compensated by PvMRP2. However, 2018; Ried et al., 2019). Thereby with an increased accumulation
compensation was not observed in seeds but in leaf tissues of such molecules, the yield capacity can be improved and also
(Cominelli et al., 2018). In addition, reduced PA seed contents fertilizer demands might be reduced (Williams et al., 2015).
in knockout mutants of OsSPDT in rice, an SPDT (Yamaji Finally, inositol pyrophosphates are regulating the jasmonic

TABLE 5 | Evaluation of pleiotropic effects in F3 seeds of Bnmrp5 mutants.

F2 seed code Generation Genotype TKW (g) Plant height (cm) Silique number/plant

180844 M3 (mu1) × M3 (mu2) J1 J1 K1 K1 3.29 ± 0.33 93.06 ± 9.57 39 ± 11.98

J1 J1 Ke Ke 3.37 ± 0.10 101.80 ± 9.12 42 ± 7.08
Je Je K1 K1 3.32 ± 0.38 105.00 ± 9.33 37 ± 10.30
Je Je Ke Ke 3.28 ± 0.14 103.20 ± 3.56 39 ± 9.80
Express 617 3.30 ± 0.22 118.02 ± 6.12 42 ± 6.08
180823 M3 (mu1) × Express 617 J1J1 2.61 ± 0.16 91.53 ± 8.56 46 ± 5.08
JeJe 2.37 ± 0.20 98.88 ± 4.60 52 ± 7.10
180850 M3 (mu2) × Express 617 K1 K1 3.24 ± 0.32 97.50 ± 8.35 39 ± 9.08
Ke Ke 3.09 ± 0.60 97.45 ± 8.40 46 ± 6.50
190690 [M3 (mu1) × Express 617) × (M3 (mu2) × Express 617] J1 J1 K1 K1 2.50 ± 0.15 84.80 ± 3.65 40 ± 4.60
J1 J1 Ke Ke 2.15 ± 0.14 86.20 ± 2.13 45 ± 12.20
Je Je K1 K1 2.67 ± 0.40 90.20 ± 8.25 40 ± 8.79
JeJe Ke Ke 2.38 ± 0.15 83.66 ± 3.44 47 ± 8.28
Express 617 2.59 ± 0.49 89.45 ± 5.40 49 ± 5.23

Plants were grown under 16 h light and 8 h dark in 11 × 11 cm pots at 22◦ C

for all measurements. Five plants for each genotype were used for measuring TKW, plant
height, and silique number. Bn.MRP5.A10 and Bn.MRP5.C05 mutations are indicated as mu1 and mu2, respectively.

Frontiers in Plant Science | www.frontiersin.org 8 May 2020 | Volume 11 | Article 603

Sashidhar et al. Bnmrp5 Shows Low Phytic Acid

acid-based defense mechanisms, which might be beneficial for an experiments, and revised the manuscript. All authors read and
improved growth under stress conditions (Laha et al., 2015). approved the final manuscript.
AtMRP5 Arabidopsis mutants have an elevated drought
tolerance due to a reduced ABA sensitivity resulting in stomatal
closure (Gaedeke et al., 2001; Klein et al., 2003). In atmrp5 FUNDING
mutants, failure to transport PA into vacuoles resulted in reduced
K+ uptake into the guard cells by inhibiting K+ inward rectifying This study was supported by a grant from the German Research
channels (Nagy et al., 2009). The stomatal closure might reduce Foundation, DFG (Grant No. JU 205/26-1).
water loss from leaves and might foster plants to cope with
drought stress and it was shown that atmpr5 mutants showed
reduced water uptake and survived longer than the wild type ACKNOWLEDGMENTS
plants under low water levels (Klein et al., 2003). However, as
we observed only 15% reduction of PA, a complete knockout of We thank Dr. Siegbert Melzer for his valuable discussions
BnMRP5 genes in future might pave the way for a comprehensive and Nirosha Karunarathna in sampling the seed material
analysis of BnMRP5 genes under drought stress. In conclusion, for the accumulation data. We also thank the Institute
the mutants presented here are important for breeding low PA of Clinical Molecular Biology in Kiel for providing Sanger
rapeseed varieties. Moreover, they can be utilized in exploring sequencing as supported in part by the DFG Clusters of
phosphorous sensing and homeostasis mechanisms. Excellence “Precision Medicine in Chronic Inflammation” and
“ROOTS” and we thank T. Naujoks, Dr. D. Langfeldt, and
Dr. B. Löscher for technical support. We thank Dipl. Biol.
DATA AVAILABILITY STATEMENT Jens Hermann and Prof. Dr. Wolfgang Bilger from the
Department of Ecophysiology of Plants at the CAU Kiel for
All datasets generated for this study are included in the HPLC measurements.
article/Supplementary Material.

SUPPLEMENTARY MATERIAL
AUTHOR CONTRIBUTIONS
The Supplementary Material for this article can be found online
NS planned, performed, and analyzed the experiments and wrote at: https://www.frontiersin.org/articles/10.3389/fpls.2020.00603/
the manuscript. HH and CJ designed the study, supervised the full#supplementary-material

REFERENCES El-Brolosy, M. A., Kontarakis, Z., Rossi, A., Kuenne, C., Günther, S., Fukuda, N.,
et al. (2019). Genetic compensation triggered by mutant mRNA degradation.
Bhati, K. K., Alok, A., Kumar, A., Kaur, J., Tiwari, S., and Pandey, A. K. (2016). Nature 568, 193–197.
Silencing of ABCC13 transporter in wheat reveals its involvement in grain Gacek, K., Bartkowiak-Broda, I., and Batley, J. (2018). Genetic and molecular
development, phytic acid accumulation and lateral root formation. J. Exp. Bot. regulation of seed storage proteins (SSPs) to improve protein nutritional value
67, 4379–4389. doi: 10.1093/jxb/erw224 of oilseed rape (Brassica napus L.) Seeds. Front. Plant Sci. 9:890. doi: 10.3389/
Braatz, J., Harloff, H. J., Emrani, N., Elisha, C., Heepe, L., Stanislav, N., et al. (2018). fpls.2018.00890
The effect of INDEHISCENT point mutations on silique shatter resistance in Gaedeke, N., Klein, M., Kolukisaoglu, U., Forestier, C., Müller, A., Ansorge, M.,
oilseed rape (Brassica napus). Theoret. Appl. Genet. 131, 959–971. doi: 10.1007/ et al. (2001). The Arabidopsis thaliana ABC transporter AtMRP5 controls root
s00122-018-3051-4 development and stomata movement. EMBO J. 20, 1875–1887. doi: 10.1093/
Campion, B., Sparvoli, F., Doria, E., Tagliabue, G., Galasso, I., Fileppi, M., et al. emboj/20.8.1875
(2009). Isolation and characterisation of an lpa (low phytic acid) mutant in Hand, C. E., and Honek, J. F. (2007). Phosphate transfer from inositol
common bean (Phaseolus vulgaris L.). TAG Theoret. Appl. Genet. 118, 1211– pyrophosphates InsP5PP and InsP4(PP)2: a semi-empirical investigation.
1221. doi: 10.1007/s00122-009-0975-8 Bioorg. Med. Chem. Lett. 17, 183–188. doi: 10.1016/j.bmcl.2006.09.066
Cominelli, E., Confalonieri, M., Carlessi, M., Cortinovis, G., Daminati, M. G., Harloff, H. J., Lemcke, S., Mittasch, J., Frolov, A., Wu, J. G., Dreyer, F., et al.
Porch, T. G., et al. (2018). Phytic acid transport in Phaseolus vulgaris: a new low (2012). A mutation screening platform for rapeseed (Brassica napus L.) and
phytic acid mutant in the PvMRP1 gene and study of the PvMRPs promoters the detection of sinapine biosynthesis mutants. TAG Theoret. Appl. Genet. 124,
in two different plant systems. Plant Sci. 270, 1–12. doi: 10.1016/j.plantsci.2018. 957–969. doi: 10.1007/s00122-011-1760-z
02.003 Jung, J.-Y., Ried, M. K., Hothorn, M., and Poirier, Y. (2018). Control of plant
Del Pozo, J. C., and Ramirez-Parra, E. (2015). Whole genome duplications in phosphate homeostasis by inositol pyrophosphates and the SPX domain. Curr.
plants: an overview from Arabidopsis. J. Exp. Bot. 66, 6991–7003. doi: 10.1093/ Opin. Biotechnol. 49, 156–162. doi: 10.1016/j.copbio.2017.08.012
jxb/erv432 Kang, J., Park, J., Choi, H., Burla, B., Kretzschmar, T., Lee, Y., et al. (2011). Plant
Desai, M., Rangarajan, P., Donahue, J. L., Williams, S. P., Land, E. S., Mandal, ABC Transporters. Arabidopsis Book 9:e0153. doi: 10.1199/tab.0153
M. K., et al. (2014). Two inositol hexakisphosphate kinases drive inositol Khan, M., Basnet, R., Islam, S., and Shu, Q. (2019). Mutational analysis of OsPLDα1
pyrophosphate synthesis in plants. Plant J. 80, 642–653. doi: 10.1111/tpj. reveals its involvement in phytic acid biosynthesis in rice grains. J. Agric. Food
12669 Chem. 67, 11436–11443. doi: 10.1021/acs.jafc.9b05052
Dorsch, J., Cook, A., Young, K., Anderson, J., Bauman, A., Volkmann, C., Kim, S. I., and Tai, T. H. (2011). Identification of genes necessary for wild-type
et al. (2003). Seed phosphorus and inositol phosphate phenotype of barley levels of seed phytic acid in Arabidopsis thaliana using a reverse genetics
low phytic acid genotypes. Phytochemistry 62, 691–706. doi: 10.1016/S0031- approach. Mol. Genet. Genomics 286, 119–133. doi: 10.1007/s00438-011-
9422(02)00610-6 0631-2

Frontiers in Plant Science | www.frontiersin.org 9 May 2020 | Volume 11 | Article 603

Sashidhar et al. Bnmrp5 Shows Low Phytic Acid

Klein, M., Perfus-Barbeoch, L., Frelet, A., Gaedeke, N., Reinhardt, D., Mueller- Sacchi, G. A., and Nocito, F. F. (2019). Plant sulfate transporters in the low phytic
Roeber, B., et al. (2003). The plant multidrug resistance ABC transporter acid network: some educated guesses. Plants 8:616. doi: 10.3390/plants812
AtMRP5 is involved in guard cell hormonal signaling and water use. Plant J. 0616
33, 119–129. doi: 10.1046/j.1365-313X.2003.016012.x Sashidhar, N., Harloff, H.-J., and Jung, C. (2019). Identification of phytic acid
Kopriva, S., and Chu, C. (2018). Are we ready to improve phosphorus homeostasis mutants in oilseed rape (Brassica napus) by large scale screening of mutant
in rice? J. Exp. Bot. 69, 3515–3522. doi: 10.1093/jxb/ery163 populations through amplicon sequencing. New Phytol. 225, 2022–2034. doi:
Laha, D., Johnen, P., Azevedo, C., Dynowski, M., Weiß, M., Capolicchio, S., 10.1111/nph.16281
et al. (2015). VIH2 regulates the synthesis of inositol pyrophosphate InsP8 and Sashidhar, N., Harloff, H. J., Potgieter, L., and Jung, C. (2020). Gene editing of three
Jasmonate-dependent defenses in Arabidopsis. Plant Cell 27:1082. doi: 10.1105/ BnITPK genes in tetraploid oilseed rape leads to significant reduction of phytic
tpc.114.135160 acid in seeds. Plant Biotechnol. J. 1–10. doi: 10.1111/pbi.13380
Liu, J., Huang, S., Sun, M., Liu, S., Liu, Y., Wang, W., et al. (2012). An improved Shah, S., Karunarathna, N. L., Jung, C., and Emrani, N. (2018). An APETALA1
allele-specific PCR primer design method for SNP marker analysis and its ortholog affects plant architecture and seed yield component in oilseed rape
application. Plant Methods 8:34. doi: 10.1186/1746-4811-8-34 (Brassica napus L.). BMC Plant Biol. 18:380. doi: 10.1186/s12870-018-1606-9
Lott, J. N. A., Ockenden, I., Raboy, V., and Batten, G. D. (2007). Phytic acid and Shi, J., Wang, H., Schellin, K., Li, B., Faller, M., Stoop, J. M., et al. (2007). Embryo-
phosphorus in crop seeds and fruits: a global estimate. Seed Sci. Res. 10, 11–33. specific silencing of a transporter reduces phytic acid content of maize and
doi: 10.1017/S0960258500000039 soybean seeds. Nat. Biotechnol. 25, 930–937. doi: 10.1038/nbt1322
Matthäus, B., Lösingand, R., and Fiebig, H.-J. (1995). Determination of inositol Sparvoli, F., and Cominelli, E. (2015). Seed biofortification and phytic acid
phosphates IP3 – IP6 in rapeseed and rapeseed meal by an HPLC method, reduction: a conflict of interest for the plant? Plants 4, 728–755. doi: 10.3390/
Part 2: investigations of rapeseed and rapeseed meal and comparison with other plants4040728
methods. Lipid/Fett 97, 372–374. doi: 10.1002/lipi.19950971004 Verrier, P. J., Bird, D., Burla, B., Dassa, E., Forestier, C., Geisler, M., et al. (2008).
Nagy, R., Grob, H., Weder, B., Green, P., Klein, M., Frelet-Barrand, A., et al. (2009). Plant ABC proteins – a unified nomenclature and updated inventory. Trends
The Arabidopsis ATP-binding cassette protein AtMRP5/AtABCC5 is a high Plant Sci. 13, 151–159. doi: 10.1016/j.tplants.2008.02.001
affinity inositol hexakisphosphate transporter involved in guard cell signaling Wang, Y., Cheng, X., Shan, Q., Zhang, Y., Liu, J., Gao, C., et al. (2014).
and phytate storage. J. Biol. Chem. 284, 33614–33622. doi: 10.1074/jbc.M109. Simultaneous editing of three homoeoalleles in hexaploid bread wheat confers
030247 heritable resistance to powdery mildew. Nat. Biotechnol. 32, 947–951. doi: 10.
Panzeri, D., Cassani, E., Doria, E., Tagliabue, G., Forti, L., Campion, B., et al. (2011). 1038/nbt.2969
A defective ABC transporter of the MRP family, responsible for the bean lpa1 Wilcox, J., Premachandra, G., Young, K., and Raboy, V. (2000). Isolation of high
mutation, affects the regulation of the phytic acid pathway, reduces seed myo- seed inorganic P, Low-Phytate soybean mutants. Crop Sci. 40, 1601–1605. doi:
inositol and alters ABA sensitivity. New Phytol. 191, 70–83. doi: 10.1111/j.1469- 10.2135/cropsci2000.4061601x
8137.2011.03666.x Williams, S. P., Gillaspy, G. E., and Perera, I. Y. (2015). Biosynthesis and possible
Pratt, J., Boisson, A.-M., Gout, E., Bligny, R., Douce, R., and Aubert, S. (2009). functions of inositol pyrophosphates in plants. Front. Plant Sci. 6:67. doi: 10.
Phosphate (Pi) starvation effect on the cytosolic Pi concentration and Pi 3389/fpls.2015.00067
exchanges across the tonoplast in plant cells: an in vivo 31P-nuclear magnetic Xu, X. H., Zhao, H. J., Liu, Q. L., Frank, T., Engel, K. H., An, G., et al. (2009).
resonance study using methylphosphonate as a Pi analog. Plant Physiol. 151, Mutations of the multi-drug resistance-associated protein ABC transporter
1646–1657. doi: 10.1104/pp.109.144626 gene 5 result in reduction of phytic acid in rice seeds. Theoret. Appl. Genet. 119,
Raboy, V., Gerbasi, P. F., Young, K. A., Stoneberg, S. D., Pickett, S. G., Bauman, 75–83. doi: 10.1007/s00122-009-1018-1
A. T., et al. (2000). Origin and seed phenotype of maize low phytic acid Yamaji, N., Takemoto, Y., Miyaji, T., Mitani-Ueno, N., Yoshida, K. T., and Ma,
1 and low phytic acid 2-1. Plant Physiol. 124:355. doi: 10.1104/pp.124. J. F. (2016). Reducing phosphorus accumulation in rice grains with an impaired
1.355 transporter in the node. Nature 541:92.
Ried, M. K., Wild, R., Zhu, J., Broger, L., Harmel, R. K., Hothorn, L. A., et al.
(2019). Inositol pyrophosphates promote the interaction of SPX domains with Conflict of Interest: The authors declare that the research was conducted in the
the coiled-coil motif of PHR transcription factors to regulate plant phosphate absence of any commercial or financial relationships that could be construed as a
homeostasis. BioRxiv [Preprint], doi: 10.1101/2019.12.13.875393 potential conflict of interest.
Rogers, S. O., and Bendich, A. J. (1985). Extraction of DNA from milligram
amounts of fresh, herbarium and mummified plant tissues. Plant Mol. Biol. 5, Copyright © 2020 Sashidhar, Harloff and Jung. This is an open-access article
69–76. doi: 10.1007/BF00020088 distributed under the terms of the Creative Commons Attribution License (CC BY).
Rounds, M. A., and Nielsen, S. S. (1993). Anion-exchange high-performance liquid The use, distribution or reproduction in other forums is permitted, provided the
chromatography with post-column detection for the analysis of phytic acid and original author(s) and the copyright owner(s) are credited and that the original
other inositol phosphates. J. Chromatogr. A 653, 148–152. doi: 10.1016/0021- publication in this journal is cited, in accordance with accepted academic practice. No
9673(93)80404-V use, distribution or reproduction is permitted which does not comply with these terms.

Frontiers in Plant Science | www.frontiersin.org 10 May 2020 | Volume 11 | Article 603

View publication stats