The Plant Journal A T-DNA mutation in the RNA helicase eIF4A
Transcript of The Plant Journal A T-DNA mutation in the RNA helicase eIF4A
A T-DNA mutation in the RNA helicase eIF4A confers a dose-dependent dwarfing phenotype in Brachypodium distachyon
Philippe Vain1, Vera Thole1, Barbara Worland1, Magdalena Opanowicz2, Max S. Bush2 and John H. Doonan2,*
1Crop Genetics Department, John Innes Centre, Norwich Research Park, Norwich NR4 7UH, UK, and2Department of Cell and Developmental Biology, John Innes Centre, Norwich Research Park, Norwich NR4 7UH, UK
Received 16 December 2010; revised 14 February 2011; accepted 16 February 2011; published online 4 April 2011.*For correspondence (fax +44 (0) 1970 622350; e-mail [email protected]).
SUMMARY
In a survey of the BrachyTAG mutant population of Brachypodium distachyon, we identified a line carrying a
T-DNA insertion in one of the two eukaryotic initiation factor 4A (eIF4A) genes present in the nuclear genome.
The eif4a homozygous mutant plants were slow-growing, and exhibited reduced final plant stature due to a
decrease in both cell number and cell size, consistent with roles for eIF4A in both cell division and cell growth.
Hemizygous plants displayed a semi-dwarfing phenotype, in which stem length was reduced but leaf length
was normal. Linkage between the insertion site and phenotype was confirmed, and we show that the level of
eIF4A protein is strongly reduced in the mutant. Transformation of the Brachypodium homozygous mutant
with a genomic copy of the Arabidopsis eIF4A-1 gene partially complemented the growth phenotype,
indicating that gene function is conserved between mono- and dicotyledonous species. This study identifies
eIF4A as a novel dose-dependent regulator of stem elongation, and demonstrates the utility of Brachypodium
as a model for grass and cereals research.
Keywords: Brachypodium, dwarf, T-DNA mutant, Arabidopsis, eIF4A, stem height.
INTRODUCTION
In recent years, Brachypodium distachyon (hereafter Brac-
hypodium) has emerged as a new genomics and experi-
mental model system to facilitate biological investigations
in grasses (Draper et al., 2001; Garvin et al., 2008; Bevan
et al., 2010; International Brachypodium Initiative, 2010).
Brachypodium, a small wild grass closely related to tem-
perate cereals with a rapid life cycle, provides a convenient
and cost-effective system in which to genetically dissect the
process of growth, to identify new factors that affect plant
height, and to understand the molecular and cellular
mechanisms underpinning growth (Opanowicz et al., 2008).
The species is amenable to genetic analysis, and popula-
tions of mutants are being systematically generated. One of
the first mutagens to be used in Brachypodium was T-DNA
(Vain et al., 2008), and a collection of insertional mutants
has been developed in the context of the BrachyTAG
programme (http://www.brachytag.org/; Thole et al., 2009,
2010). A proportion of the T-DNA mutants thus generated
display recognizable growth phenotypes. Here we describe
the characterization of an insertion in the RNA helicase eIF4A
that confers a dose-dependent dwarfing phenotype.
Eukaryotic initiation factor 4A (eIF4A) is the prototype
DEAD box RNA helicase, with canonically conserved motifs
involved in ATP binding/hydrolysis and RNA binding
(Rogers et al., 2002). eIF4A is believed to play a key role in
mRNA translation to protein, and is a component of
eukaryotic initiation surveillance (‘cap-binding’) complexes.
Effective translation depends on the association of tran-
scripts with polyribosomes, the first step of which is loading
of ribosomes onto the 5¢ m7GpppG cap of mRNAs. This step
binds and positions the mRNA on the 40S ribosomal subunit
as a prequel to scanning and subsequent translation, and is
itself subject to multi-level regulation. RNA helicases, such
as eIF4A, facilitate 40S scanning (Dever, 2002), particularly in
transcripts with a complex secondary structure that would
impede 40S progression. Accordingly, different transcripts
display differential requirements for eIF4A (Bottley et al.,
2010) that may depend on their secondary structure.
Arabidopsis eIF4A-1 and -2 proteins are highly similar,
and their probable main function is in mRNA translation
in the cytoplasm (M. S. Bush, unpublished results). The
functionally divergent eIF4A-3 protein plays a role in nuc-
lear mRNA trafficking (Koroleva et al., 2009). eIF4A1/2 is
recruited into cap complexes in proliferating cells, but is
replaced by other RNA helicases after cells have ceased
growing and became quiescent (Bush et al., 2009). It also
ª 2011 The Authors 929The Plant Journal ª 2011 Blackwell Publishing Ltd
The Plant Journal (2011) 66, 929–940 doi: 10.1111/j.1365-313X.2011.04555.x
interacts with cyclin-dependent kinase A, a cell-cycle regu-
latory protein (Hutchins et al., 2004), and this complex is
only detected in the cap complexes of proliferating cells
(Bush et al., 2009). A T-DNA mutant containing an insertion
in the Arabidopsis eIF4A-1 gene is slow-growing due to
defective cell-cycle progression and late flowering (M. S.
Bush, unpublished results).
Arabidopsis (At3g13920, At1g54270), rice (Os06g48750,
Os02g05330), maize (GRMZM2G116034, GRMZM2G027-
995), sorghum (Sb10g028940, Sb04g003390) and Brachypo-
dium all have two eIF4A orthologues. In Brachypodium,
these are encoded by Bradi1g34170 and Bradi3g03710.
Here, we discuss the extent to which the mutant phenotype
resembles that of the equivalent mutant in Arabidopsis.
Mono- and dicotyledons are distinct branches of the flow-
ering plants, and many aspects of their growth and devel-
opment are quite different. Such comparisons of
fundamental cellular functions may provide insight into
plant growth regulation.
RESULTS
eif4a T-DNA mutant isolation and characterization
As part of a systematic analysis of gene function in Brac-
hypodium, we generated a collection of 4500 T-DNA inser-
tion lines using the standard community diploid line Bd21
(Vain et al., 2008; Thole et al., 2010). We have begun to
screen this collection for lines with segregating growth
phenotypes. One such line, BdAA115, contained two T-DNA
insertions in the primary transgenic plant (T0) that could
be anchored into the annotated Brachypodium genome
sequence (version 1.0; International Brachypodium Initia-
tive, 2010). This provided two flanking sequence tags (FSTs):
one FST (JIC00100_115) indicated that a T-DNA insertion
was present on chromosome 1 in the Bradi1g34170 gene,
while the second FST (JIC00099_115) was located on chro-
mosome 3 in the Bradi3g54010 gene (http://www.model
crop.org/). A BLASTn search of GenBank showed that
Bradi1g34170 is homologous to the Oryza sativa eukaryotic
initiation factor 4A gene (score = 392, expect = e)105), and
that Bradi3g54010 is homologous to the Zea mays cyto-
chrome c oxidase copper chaperone protein family
(score = 197, expect = 5e)47).
To genetically separate these insertion events, we allowed
a population of T1 plants segregating for both T-DNA inserts
to self-pollinate, and T2 seeds were collected (Figure 1). T2
plant families were grown and the progenies screened by
PCR using primer pairs specific to each T-DNA:Bd21 junction
sequence (Table 1). This enabled us to identify a T1 plant,
A167, and its T2 progeny that contained the T-DNA inser-
Figure 1. Flow chart of Brachypodium eif4a T-DNA mutant purification, characterization and complementation.
‘nil’ plants do not contain a T-DNA insertion in the eIF4A gene (Bradi1g34170 on chromosome Bd1); ‘He’ plants are hemizygous for the T-DNA insertion (i.e. contain
one T-DNA allele and one wild-type allele); ‘Ho’ plants are homozygous for the T-DNA insertion (i.e. contain two T-DNA alleles). From the T2 generation onwards, the
plants studied were free of the second T-DNA insert in Bradi3g54010 (on chromosome Bd3). hpt, hygromycin phosphotransferase; gfp, green fluorescent protein;
bar, phosphinothricin resistance gene. T-DNA insertions are represented as green triangles. Plants expressing GFP are represented as green circles.
930 Philippe Vain et al.
ª 2011 The AuthorsThe Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 66, 929–940
tion in Bradi1g34170 (eIF4A) but not the insertion in
Bradi3g54010 (cytochrome c oxidase). PCR fragments
amplified from the T-DNA:Bradi1g34170 junction were
validated by sequence analysis. Consequently, progeny
from line A167 were used for most of the subsequent
analyses of eIF4A function as described below. Twelve T2
plants derived from A167 were scored for GFP fluorescence,
indicating that the insertion segregated as a simple Mende-
lian trait (i.e. three plants were GFP-negative and nine were
GFP-positive). These twelve T2 plants were genotyped by
PCR for the T-DNA insertion in Bradi1g34170 (eIF4A) using
primer pairs specific to the T-DNA:Bradi1g34170 junction or
the wild-type allele of Bradi1g34170 (primers spanning the
insertion site) (Figure 2b,c). We designated the plants
lacking the T-DNA insertion in Bradi1g34170 as ‘nil’
segregants (Thole et al., 2010); ‘He’ and ‘Ho’ indicate plants
that are hemi- and homozygous for the T-DNA insertion,
respectively.
The T2 plants were allowed to reach maturity and exam-
ined for overall plant stature. Nil segregants (n = 3) were tall
(height 22.7 � 1.2 cm at maturity; mean � SE) and fertile.
Hemizygous plants (n = 5) were tall (19.9 � 0.65 cm) and
fertile. Two of the five hemizygous plants exhibited retarded
growth at 3 weeks, but caught up with the other hemizy-
gous plants at 5 weeks (Figure 2a). Overall, hemizygotes
remained significantly smaller (88%) compared to the nil
segregants (Figure 2d). The fertility of hemizygous plants
(mean 64 seeds, n = 5) was also 39.2% lower than the
fertility of the nil segregants (mean 106 seeds, n = 3).
Homozygous plants (n = 4) exhibited a dwarf phenotype
(height 9.9 � 0.6 cm at maturity) (Figure 2a). This pheno-
type was very pronounced during the early stages of
development (34% of nil plant height), and stabilized at
around less than half (46%) of the height of nil plants at
maturity. Homozygous plants were completely sterile, but
inflorescences emerged approximately at the same time as
for nil segregants or hemizygous plants. In total, 108 T2
plants produced from seven T1 plants (A166, A167, A168,
A169, A215, A218 and A220) were genotyped and examined
for plant stature. eif4a mutants were further characterized at
the T3 generation using progeny plants from A167 and A169
T1 plants. A flow chart of plant material production is shown
in Figure 1.
The T-DNA insertion is predicted to be in the first intron of
the Bradi1g34170 coding sequence (Figure 2b), and could
lead to either a truncated peptide or an overall decrease
in eIF4A protein levels. Protein levels in plants of defined
genotypes (i.e. nil and homozygous insertion lines derived
from A167) were assessed using Western blots probed with
an anti-eIF4A antibody. This indicated that levels of eIF4A
protein were strongly reduced in the homozygous mutant
compared to nil and hemizygous plants (Figure 3).
Immunoprecipitation using the anti-eIF4A antibody followed
by Western blotting (Figure 3) confirmed this, and indicated
that the T-DNA insertion leads to strong knock-down of
eIF4A in Brachypodium. The small amount of eIF4A protein
that can be detected in the homozygous mutant could
be produced from Bradi3g03710, the other copy of eIF4A,
although we cannot be certain that the T-DNA insertion in
Bradi1g34170 leads to a complete elimination of function.
Hemizygous plants (containing one T-DNA and one wild-
type allele) exhibited a gene dosage effect, with reduced
growth and fertility compared to the controls despite having
similar levels of eIF4A protein.
Growth and architecture of the Brachypodium eif4a
T-DNA mutant
We next undertook a detailed phenotypic analysis of T3
plants (Figure 4) and comparison with T2 plants (Figure 2). In
both generations, the homozygous mutants consistently
exhibited a dwarf phenotype (43–46% of the height of the nil
segregants) (Movie S1). Likewise, hemizygous plants from
both generations were slightly smaller (81–88%) than the nil
segregants at maturity. This semi-dwarf plant phenotype
suggests a dosage effect of eIF4A in Brachypodium. Exam-
ination of the primary tillers from T3 plants (Figure 5)
showed that each component of each internode (stem, leaf
and sheath) was smaller in the homozygous mutant
Table 1 PCR primers used for T-DNA flank-ing sequence retrieval and characterization Primer type Primer name Primer sequence
T-DNA:Bd21 junctionsT-DNA (LB) TDNA4 5¢-CGGCCGCATGCATAAGCTTA-3¢Bradi1g34170 1g34170-R1 5¢-CCAAACCATAGTCAAGCTGTT-3¢T-DNA (RB) TDNA1 5¢-CTGATAGTGACCTTAGGCGA-3¢Bradi3g54010 3g54010-R1 5¢-GGCATCGCTTGTGAGCTTCTA-3¢
Bd21 wild-type allelesBradi1g34170 1g34170-F1 5¢-CTCAGTCCATTGTAGCCATCA-3¢Bradi3g54010 3g54010-F1 5¢-GCTGCATGATTCCGAGAGTTCA-3¢
ComplementationT-DNA (35S:bar) BAF1 5¢-GGTCTGCACCATCGTCAACC-3¢T-DNA (35S:bar) BAR2 5¢-GTCATGCCAGTTCCCGTGCT-3¢T-DNA (AteIF4A-1) 4A1F 5¢-CTTCGATCTCCCAACTCAG-3¢T-DNA (AteIF4A-1) 4A1R 5¢-TGGGAGAGAGAGGAGAGAACAACAACTAAGAAA-3¢
Brachypodium eif4a T-DNA mutant 931
ª 2011 The AuthorsThe Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 66, 929–940
compared to equivalent structures in either the hemizygous
or nil segregant plants. The lengths of the top two leaves and
sheaths in the homozygous eif4a mutant were 60 and 69% of
those of the nil segregants, respectively (regardless of their
position relative to the inflorescence). However, the stem
length between each node was dramatically reduced to 43%
(peduncle) and 30% (internode under peduncle) of that of the
nil segregant. Hemizygous plants had leaves and sheaths
comparable in length with those of the nil plants, but the
stem length was only 81% (peduncle) and 79% (internode
under peduncle) of that observed in nil plants. The overall
reduction in stem length completely explained the de-
creased height of hemizygous T3 plants at maturity (i.e. 81%
of the total height of nil segregants) (Table S1). The stem
length reduction in the semi-dwarf hemizygous plants also
supports a dosage effect of eIF4A on plant height.
To determine the cellular basis of reduced leaf growth in
Brachypodium, we examined the leaf cell size of T3 homo-
zygous and nil plants at 4.5 weeks and at maturity (7 weeks).
Observation of abaxial and adaxial (rib and furrow) epider-
mal leaf cells (Figure 6) – excluding stomata and hairs –
showed that all types of epidermal cells in the homozygous
plants were on average 20% (4.5 weeks) to 13% (7 weeks)
shorter and more rectangular compared to the correspond-
ing cells in nil plants (Table S2). This is consistent with
whole-plant height observations showing that the mutation
phenotype is more pronounced during the early stages
of development (Figure 2). Epidermal cell width was less
affected than cell length (approximately 50% of cell length
variation) by the eif4a mutation, and varied according to cell
type. The eif4a mutation led to a significant reduction in
palisade cell density in leaves (Table S2). This resulted
from a combination of larger cell size and a less com-
pact arrangement of cells (with larger intercellular spaces)
(Figure 6).
Inflorescence and grain development of Brachypodium
eif4a T-DNA mutants
Examination of spikelets of T3 plants (Figure 7) indicated
that the eif4a mutation resulted in a significantly lower
number of smaller spikelets (2.3 per tiller) compared to
hemizygous (3.4) or nil segregant (3.4) plants (Table S1). The
second and third spikelet (when present) in the inflores-
cence of homozygous mutants exhibited the strongest size
reduction. Fertility was reduced in hemizygous plants and
homozygous plants were sterile. Pollen from the eif4a
homozygous mutants was either deformed in shape or
absent. Normal pollen grains were observed in hemizygous
and nil plants (Figure 8a,b). Alexander staining (Alexander,
1969) revealed that, in contrast with the round uniformly
(a)
(b) (d)
(c)
Figure 2. Phenotyping and genotyping the eIf4a
T-DNA mutation in Brachypodium.
(a) Phenotype of segregating T2 plants (3 and
5 weeks old). ‘nil’ plants do not contain a T-DNA
insertion in the eIF4A gene (Bradi1g34170); ‘He’
and ‘Ho’ plants are hemi- and homozygous for
the T-DNA insertion in eIF4A, respectively.
(b, c) Genotyping of segregating T2 plants. The
presence of the wild-type Bradi1g34170 allele
and T-DNA:Bradi1g34170 junction were asses-
sed in each plant using specific PCR primer pairs
(indicated as arrows; sequences given in
Table 1). The introns (line), exon (blue box)
and untranslated regions (white box) of the
Bradi1g34170 transcript and the insertion site of
the T-DNA (green box) are shown.
(d) Total height of 3- and 5-week-old T2 plants.
Circles represent individual plant height mea-
surements. Percentage represents the mean size
compared to ‘nil’ plants.
932 Philippe Vain et al.
ª 2011 The AuthorsThe Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 66, 929–940
stained pollen from nil segregants and hemizygotes, pollen
from the homozygotes was much more diverse in size and
shape. In many cases, the pollen from the homozygotes did
not take up the Alexander stain, which is indicative of a lack
of pollen viability, or appeared collapsed (Figure 8c).
Grain development was also compromised in eif4a
homozygous mutants compared to hemizygous and nil T3
plants from the same segregating population (Figure 9).
Grains were staged using anthesis as a reference point, and
their growth was followed for up to 6 days after anthesis
(DAA). The fertilized caryopsis in Brachypodium is approx-
imately 1 mm long. There were no significant changes in
caryopsis length and width in eif4a homozygous mutants,
but grain length in hemizygous and nil plants increased
4.6–4.7-fold during the first 6 DAA (Figure S1). During this
period, the hemizygous grain size (length and width)
remained approximately 90% smaller than that of the nil
control, suggesting a dosage effect on grain development,
as for stem and total plant height. We histologically exam-
ined homozygous, hemizygous and control (nil) grains,
focusing on the early stages of development. Transverse
sections revealed that the caryopsis in hemizygous and nil
plants developed similarly to wild-type Brachypodium Bd21
grains (Opanowicz et al., 2011), and formed pericarp, syn-
cytial endosperm and a circular central vacuole at 0 DAA.
The central vacuole in the hemizygote and nil segregants
occupied about 250 lm, and post-fertilization development
occurred as previously reported for Bd21 (Opanowicz et al.,
2011). The corresponding stages in the homozygous mutant
showed a highly reduced central vacuole that did not change
substantially during development. At 6 DAA, the homozy-
gous caryopsis was a similar size to that at 0 DAA, but
became flattened dorso-ventrally. The pericarp remained
the predominant tissue in the mutant caryopsis. The nucellar
epidermis and other internal tissue layers within the
(a)
(b)
(d)
(e)
(c)
Figure 3. eIF4A protein levels in Brachypodium T-DNA mutant plants.
Protein levels in leaves of 4-week-old T3 Brachypodium plants. ‘nil’ plants do
not contain a T-DNA insertion in the eIF4A gene (Bradi1g34170); ‘He’ and
‘Ho’ plants are hemi- and homozygous for the T-DNA insertion in eIF4A,
respectively; ‘c25’, ‘c40’ and ‘c54’ plants are homozygous for the T-DNA
insertion and independently transformed with the Arabidopsis eIF4A-1 gene
(At3g13920). ‘At’ corresponds to a 2-day-old Arabidopsis thaliana cell culture
(Bush et al., 2009). Ladders indicate molecular weight (kDa).
(a, d) Coomassie staining of polyacrylamide gels loaded with 20 lg soluble
cell extract (SCE) or 15 ll protein samples immunoprecipitated with an anti-
eIF4A antibody (IP). The large band in the SCE lanes is Rubisco, whilst that in
the IP lanes is the heavy immunoglobulin polypeptide from the anti-eIF4A
antibody. The band likely to be eIF4A is indicated with an arrowhead.
(b, c, e) Enhanced chemiluminescence detection of immunolabelled eIF4A
protein in SCE (10 lg loaded per lane) (b, e) and immunoprecipitated (IP)
samples (c). The arrowheads indicate the eIF4A protein.
(a)
(b)
Figure 4. Growth of the Brachypodium eif4a T-DNA mutant.
(a) Phenotype of T3 Brachypodium plants at maturity (6 weeks old). ‘nil’ plants
do not contain a T-DNA insertion in the eIF4A gene (Bradi1g34170); ‘He’ and
‘Ho’ plants are hemi- and homozygous for the T-DNA insertion in eIF4A,
respectively. ‘Ho + AteIF4A-1’ plants are homozygous for the T-DNA insertion
and transformed with the Arabidopsis eIF4A-1 gene (At3g13920).
(b) Total plant height of ‘nil’ (n = 10), ‘He’ (n = 19), ‘Ho + AteIF4A-1’ (n = 26)
and ‘Ho’ (n = 26) plants.
Brachypodium eif4a T-DNA mutant 933
ª 2011 The AuthorsThe Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 66, 929–940
caryopsis never became clearly differentiated. Nuclear dis-
tribution was irregular around the periphery of the central
vacuole, which formed very late in grain development.
We conclude that there is a pre-fertilization defect in grain
development in the mutant.
Complementation of the Brachypodium T-DNA mutant
with Arabidopsis thaliana eIF4A-1
To confirm that the plant phenotype was due to the insertion
in Bradi1g34170, we investigated whether it could be com-
plemented by expression of the wild-type Arabidopsis
eIF4A-1 gene containing its own regulatory sequences.
Twenty-nine embryogenic callus lines were established
from T3 immature embryos derived from four hemizygous
T2 plants (B13, B14, B15 and B20) (Figure 1) and scored for
GFP fluorescence. Seven callus lines were GFP-negative and
22 were GFP-positive, confirming a 3:1 segregation. The 29
callus lines were screened by PCR using primer pairs specific
to the T-DNA:Bradi1g34170 junction or the Bradi1g34170
wild-type allele (eIF4A) to determine which callus lines were
nil segregants or homozygous/hemizygous for the T-DNA
insertion in Bradi1g34170. The regions flanking both sides of
the T-DNA insert(s) were analysed in the seven callus lines
selected (Figure 1). The FST JIC00100_115 [characteristic of
the T-DNA insertion in Bradi1g34170 (eIF4A)] was recovered
from the six hemizygous/homozygous callus lines but not
from the nil segregant line. FST JIC00099_115 [characteristic
of the T-DNA insertion Bradi3g54010 (cytochrome c oxi-
dase)] was not recovered from any of the callus lines, con-
firming the absence of the T-DNA insertion in Bradi3g54010.
Two hemizygous callus lines (L20, L24), four homozygous
callus lines (L9, L13, L23, L29) and one nil callus line (L3)
were selected for transformation. Half of the callus from
each callus line was kept for plant regeneration, and the
other half was used for transformation with AteIF4A-1.
Plasmid pMDC123-AteIF4A-1 was used for Agrobacterium-
mediated transformation of Brachypodium callus lines, and
phosphinothricin selection was used to recover transgenic
plants. Ten independently transformed plants and ten non-
transformed plants from each homozygous callus line were
regenerated in parallel. In total, 80 independently comple-
mented and non-complemented plants were produced from
the four callus lines L9, L13, L23 and L29 (Figure 1). Seventy-
three per cent of the T3 homozygous mutants comple-
mented with AteIF4A-1 exhibited a restored wild-type or
near wild-type (intermediate) plant phenotype. The remain-
ing 27% retained small plant stature similar to the mutant,
and may correspond to plants expressing no or low levels of
the AteIF4A-1 transgene. In comparison, all non-comple-
mented plants (40/40) regenerated from the homozygous
callus lines displayed a mutant phenotype. When consider-
ing all plants, the AteIF4A-1 gene, on average, significantly
increased (see Table S1) total plant height from 43% of
control plant height (mutant) to 71% of control plant height
(complemented) at maturity (Figure 4). Complemented
plants had a phenotype that resembled that of hemizygous
plants (i.e. 81% of control plant height). During the first
3 weeks, there was little difference between nil, hemizygous
and complemented plants. However, from the time of
emergence of the inflorescences (4 weeks and thereafter)
the difference between nil and complemented plants
increased to 71% of control plant height. Untransformed
Figure 5. Tiller architecture of Brachypodium
eif4a T-DNA mutant plants.
Arrows indicate the mean inflorescence plus
peduncle as well as internode lengths in 5-week-
old T3 Brachypodium plants. The mean leaf
length is indicated in green above each leaf
(measurements are detailed in Table S1). ‘n’
indicates the number of primary tillers mea-
sured. ‘nil’ plants do not contain a T-DNA
insertion in the eIF4A gene (Bradi1g34170);
‘He’ and ‘Ho’ plants are hemi- and homozygous
for the T-DNA insertion in eIF4A, respectively.
‘Ho + AteIF4A-1’ plants are homozygous for the
T-DNA insertion and transformed with the Ara-
bidopsis eIF4A-1 gene (At3g13920).
934 Philippe Vain et al.
ª 2011 The AuthorsThe Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 66, 929–940
mutant plants remained, on average, smaller than all other
plant types during their entire life cycle (Figure 4b). To
characterize the complementation in more detail, we
examined and measured isolated tillers. Complementation
of homozygous mutants with the AteIF4A-1 transgene
restored leaf and sheath length to 90–91% of that of nil
plants, and restored stem length to 66–72% of that of
nil plants (Figure 5). Complementation of the homozygous
eif4a mutants also increased the number of florets of T3
plants from 2.3 to 3.2 per tiller, which is similar to the
number (3.4) observed in nil plants (Figure 7). Interestingly,
the complemented plants exhibited a mixture of normal and
malformed pollen grains in their anthers (Figure 8a). How-
ever, the complemented homozygous plants did not
produce seeds. It remains unclear whether complementa-
tion restored some level of fertility, as the transformation
process itself – involving phosphinothricin selection – also
strongly reduced the fertility of controls (nil) or hemizygous
plants. As determined by Western blot analysis, eIF4A
protein levels in 4-week-old complemented plants were
restored to normal or near normal levels (Figure 3).
DISCUSSION
We report the characterization of a T-DNA-induced mutation
in the model grass Brachypodium. Brachypodium is rapidly
becoming the model organism of choice for grasses, and
is a counterpart for the rapidly cycling small dicotyledonous
model Arabidopsis. The widespread use of Arabidopsis has
Figure 7. Inflorescences of Brachypodium eif4a T-DNA mutant plants.
Inflorescence and individual spikelet of 5-week-old T3 Brachypodium plants.
Black bar = 1 cm. The mean number of spikelets and length of inflorescences
(in parentheses) are shown. Spikelet length measurements (n = 378) are
detailed in Table S1. Close-up of the top spikelet at maturity (7.5-week-old
plants) is shown in the insets (black background). White bar = 0.5 cm. ‘nil’
plants do not contain a T-DNA insertion in the eIF4A gene (Bradi1g34170); ‘He’
and ‘Ho’ plants are hemi- and homozygous for the T-DNA insertion in eIF4A,
respectively. ‘Ho + AteIF4A-1’ plants are homozygous for the T-DNA insertion
and transformed with the Arabidopsis eIF4A-1 gene (At3g13920).
(a)
(b)
Figure 6. Epidermal and palisade leaf cells in Brachypodium eif4a T-DNA mutant plants.
Abaxial and adaxial epidermis was observed by scanning electron microscopy of fresh 4.5-week-old T3 Brachypodium leaves. Ri, rib; Fu, furrow. Palisade cells were
observed using light microscopy of cleared leaves at plant maturity (7.5-week-old T3 plants). Black bar = 50 lm. White bar = 100 lm.
(a) ‘Ho’ plants are homozygous for the T-DNA insertion in the eIF4A gene (Bradi1g34170).
(b) ‘nil’ plants do not contain a T-DNA insertion.
Brachypodium eif4a T-DNA mutant 935
ª 2011 The AuthorsThe Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 66, 929–940
revolutionized our understanding of plant biology, in part
due to the many genetic and genomics resources available.
Similar resources for Brachypodium will enable us to
address many questions in the developmental and com-
parative biology of grasses. The availability of T-DNA
insertion lines, in combination with the recently released
Brachypodium genome sequence (International Brachypo-
dium Initiative, 2010), is critical to facilitate biological
investigations in grasses. Since 2008, the BrachyTAG
mutant collection has contributed to this effort. To date, 16%
of the collection (4500 lines) has been analysed (Thole et al.,
2010), and 17% of the tagged genes have been distributed to
laboratories in nine countries (http://www.brachytag.org/
users.htm). The tagging, traceability (GFP-based), high level
of mutation indexing, and prediction of tagged gene func-
tion have greatly facilitated mutant uptake by the research
community. International efforts are underway to mutage-
nize the entire Brachypodium gene complement (25 532
protein-coding loci) during the next few years. Additional
tagged mutant collections (T-DNA or transposon), as well as
chemically mutagenized populations for targeting induced
local lesions in genomes (TILLING) are being generated.
Although mutant populations are being developed, no such
mutants have been characterized in detail to date. Here,
we report the detailed analysis and complementation of a
T-DNA mutation in the Brachypodium eIF4A gene (Bradi-
1g34170). The entire process from identification of the
insertion, phenotypic characterization and complementation
was undertaken over a 9–12 month period, which is similar
to that currently enjoyed by the Arabidopsis research
community.
The eIF4A proteins and genes are very highly conserved,
but their functions are not well understood. Although
considered to be general translation factors, biochemical
and genetic analysis in other species suggests that this role
is most prominent in proliferating cells (Dorn et al., 1993;
(a) (b) (c) Figure 8. Anthers and pollen of Brachypodium
eif4a T-DNA mutant plants.
‘nil’ plants do not contain a T-DNA insertion in
the eIF4A gene (Bradi1g34170); ‘He’ and ‘Ho’
plants are hemi- and homozygous for the T-DNA
insertion in eIF4A, respectively. ‘Ho + AteIF4A-1’
plants are homozygous for the T-DNA insertion
and transformed with the Arabidopsis eIF4A-1
gene (At3g13920).
(a) Scanning electron micrographs of anthers
from 4.5-week-old T3 Brachypodium plants.
White bar = 50 lm.
(b) Light microscopic images of anthers from
3–4-week-old T3 Brachypodium plants. Black
bar = 100 lm.
(c) Light microscopic images of fresh pollen
following Alexander staining. Black bar =
30 lm.
(a) (b)
(c)
Figure 9. Grain development in eif4a T-DNA mutant plants.
‘nil’ plants do not contain a T-DNA insertion in the eIF4A gene (Bradi1g34170);
‘He’ and ‘Ho’ plants are hemi- and homozygous for the T-DNA insertion in
eIF4A, respectively. ODAA/6DAA = 0 and 6 days after anthesis, respectively.
(a) Ovary and grain of 3–4-week-old T3 Brachypodium plants. White
bar = 1 mm.
(b, c) DAPI-stained transverse sections of homozygous and nil caryopses at
0 days after anthesis (b) and 6 days after anthesis (c). The pericarp (p), central
vacuole (cv) and nucellar epidermis (ne) are indicated. White bars = 300 lm
(b) and 200 lm (c).
936 Philippe Vain et al.
ª 2011 The AuthorsThe Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 66, 929–940
Daga and Jimenez, 1999; Bush et al., 2009). The ability of
the Arabidopsis gene to complement the Brachypodium
mutant indicates that the eIF4A genes are also functionally
conserved. However, the signals controlling gene expres-
sion are likely to be only partially conserved, as comple-
mentation of the vegetative phenotype is observed in the
majority of plants but fertility is not restored. Restoration of
the level of eIF4A protein to normal levels in the comple-
mented lines suggests that the dicotyledonous promoter is
sufficiently active during vegetative growth in Brachypodium
to support effective complementation. Notably, the first
generation of complemented lines resembles most closely
the hemizygous plants, which also contain similar levels of
protein to wild-type plants but are phenotypically different
from wild-type plants in terms of plant height and fertility. As
the effect of eIF4A gene is clearly dose-dependent, but this
is not reflected at the protein level in either hemizygous or
complemented lines, the precise level of eIF4A may be
crucial for certain stages of development, for example stem
elongation and grain development. As a first step towards
identifying a possible molecular explanation, we compared
eIF4A gene and promoter structures in Arabidopsis and
Brachypodium. Structurally, both Arabidopsis (At3g13920)
and Brachypodium (Bradi1g34170) eIF4A promoter regions
harbour a similar type and number of predicted transcription
factor binding sites (Figure S2), but the Arabidopsis pro-
moter has a higher concentration of binding sites and
consensus motifs clustered near the transcription start site
compared to the Brachypodium gene. It is possible that such
differences could lead to altered expression profiles.
The high level of sequence homology (80–82% nucleotide
identity) makes it difficult to establish relationships between
the proteins (95–97% amino acid similarity) across species
with a high degree of confidence. It is likely that the presence
of two copies of the eIF4A gene in both Arabidopsis and
Brachypodium is due to independent genome duplications
that occurred after the divergence of the di- and monocot-
yledonous lineages (Blanc et al., 2000; International Brac-
hypodium Initiative, 2010). One such duplication in the
Brachypodium lineage is thought to have occurred
30–40 million years ago, while a genome duplication took
place in Arabidopsis approximately 200 million years ago
(Vision et al., 2000). Inspection of the Plant Genome Dupli-
cation Database (http://chibba.agtec.uga.edu/duplication/
index/home) indicates that the genomic regions around
Arabidopsis eIF4A-1 and eIF4A-2 contain additional dupli-
cated genes suggestive of an ancestral duplication event.
Using the Bradi1g34170 gene (on chromosome Bd1) as a
query reveals that the region is highly syntenic with a region
on chromosome 6 of rice. Likewise, a query using Bradi-
3g03710 reveals high synteny with a region on rice chromo-
some 2. In both cases, the synteny of the region around each
Brachypodium eIF4A gene is greater with the equivalent
region in rice than it is between the two regions in
Brachypodium. This suggests that the eIF4A duplication
event in grasses precedes the separation of rice and
Brachypodium/Triticae at 32–45 million years ago.
The phenotype of the Brachypodium insertion mutant
reported here resembles that of the Arabidopsis eif4a-1
mutant in that it is slow-growing. However, in Arabidopsis,
plants grow to a normal final size. Flowering time is not
significantly affected in Brachypodium by down-regulation
of the eIF4A function, unlike Arabidopsis, in which knock-
down of several eIF4-related functions led to prolonged
vegetative growth (Geraldo et al., 2009; Lellis et al., 2010).
This prolonged growth period means that the final stature
and appearance of the Arabidopsis mutants are similar
to those of other late-flowering mutants. Conversely, the
Brachypodium mutant progresses through its life cycle at
the normal rate and flowers at a similar time to wild-type
plants. In Arabidopsis, the floral repressor FLC regulates
flowering time. The level of FLC protein is very sensitive to
perturbation of translation, and mutations in many genes
encoding translation factors delay flowering time (Simpson
et al., 2004). FLC function does not appear to have been
conserved in the grasses, including Brachypodium (Higgins
et al., 2010). On the basis of the eif4a mutant phenotype and
the absence of FLC, we speculate that flowering time in
grasses is not as sensitive as Arabidopsis is to perturbation
of mRNA translation.
Leaf length is reduced in the Brachypodium mutant
compared to wild-type plants. At the cellular level, this is
due to a reduction in both cell number and cell size. The
intermediate stature of the hemizygous mutants reveals
a dosage effect (a phenomenon previously noted for eIF4A
in Drosophila development; Galloni and Edgar, 1999) that
particularly affects stem elongation. This intermediate stem
phenotype resembles that of semi-dwarf varieties of wheat
and rice. These varieties have underpinned the success of
the ‘Green Revolution’ because they can be fertilized heavily
to increase grain yield but stem growth remains restricted.
The shortened stem has two main advantages: losses due
to lodging are reduced, and the harvest index (the ratio of
straw to grain) is improved. Therefore, such varieties are
well adapted to modern agricultural practice. However, their
genetic base is quite narrow, and the genes so far utilized in
agriculture encode components of the gibberellin (reviewed
by Sakamoto and Matsuoka, 2004) or brassinosteroid
(Morinaka et al., 2006) biosynthetic or signalling pathways.
As in many dwarf and semi-dwarf cereals (Worland et al.,
2001), reduction in eIF4A is associated with reduction in grain
yield. Grain development in eif4a mutant lines is affected
in two ways. First, there is a dose-dependant decrease in
grain growth rate and a reduction in the number of grains
produced per plant in the hemizygous lines. Second, the
homozygous lines are completely infertile. Similar unin-
tended negative consequences on growth, pathogen resis-
tance or yield (Worland et al., 2001; Knopf et al., 2008) have
Brachypodium eif4a T-DNA mutant 937
ª 2011 The AuthorsThe Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 66, 929–940
been observed previously for other dwarfing alleles, and
these can restrict their use in breeding programmes. In
wheat, multiple alleles of just three dwarfing genes are used
to modify plant stature in commercial lines (Knopf et al.,
2008). A deeper understanding of the regulation of height in
cereals could reveal additional genes affecting this important
trait, and provide novel approaches to breed or engineer
plant architecture. Eliminating unwanted side-effects of such
dwarfing mutations may allow application of strategies that
would restrict plant height but leave fertility intact. For
example, knockdown of eIF4A selectively in the stem during
the cell proliferation phase might be a strategy that would
permit dwarfing without adversely affecting yield. A more
comprehensive collection of mutant lines, including knock-
down of other genes affecting plant height, will allow us to
address these important issues.
EXPERIMENTAL PROCEDURES
Plant culture and phenotyping
The plant line BdAA115 is part of the previously described Brachy-TAG T-DNA mutant collection (Thole et al., 2010) produced usingAgrobacterium-mediated transformation of Brachypodium com-pact embryogenic calli (Alves et al., 2009) and transformation vectorpVec8-GFP (35Spro:hyg-int:NOS-ter + UBIpro-UBI-int:sGFPS65T:NOS-ter; pVec8-GFP GenBank accession number FJ949107). Seedsfrom the mutant line BdAA115 were sterilized, kept for 2 days in thedark at 5�C on wet sterile filter paper and then for 5–7 days at 25�C,first in the dark (2 days) then under a 16 h photoperiod. Germinatedseedlings were potted and grown in 2 · 2 cm cell trays in a con-trolled environment room at 22�C with a 20 h photoperiod asdescribed by Alves et al. (2009). When plants started to form tillers,they were transferred into Rootrainer� pots (Roonash, http://www.ronaash.co.uk/) and grown under the same conditions. T1, T2
and T3 plants from line BdAA115 were phenotyped by measuringthe total plant height as well as the length (and in some cases width)of the spikelets, stems, leaves and sheaths in the top three tillers ofeach plant. Approximately 100 mg leaf material was sampled perindividual plant, and stored at )80�C for molecular analysis.
Characterization of T-DNA insertions
DNA from 2–4-week-old wild-type and T0, T1, T2 and T3 plants, aswell as from 6-week-old T2 callus lines (about 90 mg callus) wasextracted using a DNeasy plant mini kit (Qiagen, http://www.qiagen.com/) according to the manufacturer’s instructions. Forvalidation of the T-DNA insertion(s), confirmation of complemen-tation, and assessment of zygocity level, PCRs were performed with15 ng genomic DNA and Taq DNA polymerase (New EnglandBiolabs, http://www.neb.com/) as detailed below. The identity of thePCR fragments was verified by sequencing.
The T-DNA insertion in Bradi1g34170 (eIF4A) was validated usingprimers 1g34170-R1 and TDNA4, amplifying a 482 nt T-DNA:Bradi-1g34170 junction sequence (Table 1). The 1g34170-R1 and 1g34170-F1 primers amplify a 807 nt promoter:Bradi1g34170 fragment thatdoes not contain the T-DNA insertion. The size of the T-DNAinsertion prevents amplification from primers 1g34170-R1 and1g34170-F1 in the mutant alleles. In homozygous T-DNA lines, onlythe 482 nt fragment is detected, while in hemizygous T-DNA linesboth the 482 and 807 nt fragments are found. Wild-type Bd21 plantsand nil segregants only produce the 807 nt fragment. The cycling
conditions used for both primer combinations were denaturation at94�C for 1.5 min, followed by 35 cycles of denaturation at 94�C for30 sec, annealing at 54�C for 30 sec and extension at 72�C for 1 min,and a final extension at 72�C for 10 min.
The T-DNA insertion in Bradi3g54010 (cytochrome c oxidase)was validated using primers 3g54010-R1 and TDNA1 (Table 1),amplifying a 526 nt T-DNA:Bradi3g54010 junction sequence (Gen-Bank accession number ER987355). The primers 3g54010-R1 and3g54010-F1 (Table 1) amplify a 638 nt Bradi3g54010 fragment thatdoes not contain the T-DNA insertion. The cycling parameters aresimilar to those described above, with an annealing temperature of57�C instead of 54�C.
Retrieval of T-DNA flanking sequences in BdAA115 T3 callus linesused for complementation assays (see below) was performed using100 ng of genomic DNA digested with BfaI, and adapter-ligationPCR amplification of the regions flanking both sides of the T-DNAinsert(s) as described previously (Thole et al., 2009).
Complementation assays
For complementation assays, the full-length genomic ArabidopsiseIF4A-1 gene (At3g13920) was cloned into transformation vectorpMDC123 (Curtis and Grossniklaus, 2003) containing the2x35Spro:bar:35Ster selection cassette. The full-length (3199 bp)At3g13920 genomic DNA fragment was amplified using primers5¢-GGGGACAAGTTTGTACAAAAAAGCAGGCTacaccaattctaccataaccg-3¢ (upper case indicates the attB1 site and lower case represents thegenomic sequence) and 5¢-GGGGACCACTTTGTACAAGAAAGCTG-GGTTaggtttgtgtggatgtga-3¢ (upper case indicates the attB2 site andlower case represents the genomic sequence) and Pfu Ultra IIFusion HS DNA polymerase (Agilent Technologies, http://www.agilent.com). The PCR product was gel-purified, cloned into vectorpDONR207 (Invitrogen, http://www.invitrogen.com/), and subse-quently into plasmid pMDC123. The pMDC123-AteIF4A-1 vector wasmobilized into Agrobacterium tumefaciens strain AGL1 via a freeze–thaw method. The presence of the binary vector in A. tumefacienswas confirmed by PCR using bar-specific primers (BAF1 and BAR2;Table 1) amplifying a 420 nt fragment of the bar gene (denaturationat 95�C for 1 min, followed by 30 cycles of denaturation at 95�C for30 sec, annealing at 60�C for 30 sec and extension at 72�C for 1 min,and a final extension at 72�C for 10 min) and AteIF4A-1 specificprimers (4A1F and 4A1R; Table 1) amplifying a 519 nt fragment ofthe AteIF4A-1 gene and its 3¢ UTR (denaturation at 94�C for 1.5 min,followed by 35 cycles of denaturation at 94�C for 30 sec, annealingat 57�C for 30 sec and extension at 72�C for 1 min, and a finalextension at 72�C for 10 min).
Compact embryogenic callus lines were produced as describedby Alves et al. (2009) from segregating T3 immature embryos fromfour T2 hemizygous plants (B13, B14, B15 and B20) (Figure 1). Thecallus lines were scored for GFP expression (Vain et al., 2008) andgenotyped for T-DNA insertions in the Bradi1g34170 and Bradi-3g54010 genes as described above. Four lines homozygous for boththe T-DNA insertion in the Bradi1g34170 gene (L9, L13, L23 and L29)(Figure 1) and the wild-type Bradi3g54010 gene were used for plantregeneration (without a second round of transformation) and com-plementation. Agrobacterium-mediated transformation of 7-week-old compact embryogenic callus (approximately 100 compactembryogenic calli per T3 callus line) was performed as describedpreviously (Alves et al., 2009) using strain AGL1 harbouringpMDC123-AteIF4A-1 with a modified selection system adapted tothe bar gene (i.e. transformed tissues were kept under 5 mg L)1
phosphinothricin during callus culture and 3 mg L)1 phosphino-thricin during regeneration and germination of transformed plantsinstead of using hygromycin).
938 Philippe Vain et al.
ª 2011 The AuthorsThe Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 66, 929–940
Protein analysis
Soluble cell extracts were prepared as previously described forArabidopsis (Bush et al., 2009) using approximately 100 mg of fro-zen leaf samples of BdAA115 T3 plants previously genotyped for theT-DNA insertion in Bradi1g34170 (i.e. hemi-, homozygous plantsand nil segregants identified as described above, as well as com-plemented homozygous plants c40, c25 and c54). Cell extracts wereadjusted to equal protein concentration, and samples were boiledin Laemmli buffer and separated by SDS–PAGE by loading equalvolumes per lane according to standard procedures (Sambrooket al., 1989). Immunoprecipitation was performed by incubatingsoluble cell extracts with 50 ll of a Protein A–anti-wheat eIF4Apolyclonal antibody complex overnight at 4�C (Bush et al., 2009).Washed antibody complexes were boiled in 50 ll Laemmli buffer,and then 15 ll samples of immunoprecipitated proteins were sep-arated by SDS–PAGE, transferred to nitrocellulose, probed with theanti-wheat eIF4A antibody and enhanced chemiluminescent detec-tion was performed using a SuperSignal West Pico chemilumines-cent substrate kit (Pierce, http://www.piercenet.com/). Duplicategels were stained with Coomassie brilliant blue to confirm uniformprotein loading.
Tissue preparation and staining for cellular analyses and
histology
For cellular analysis of leaves and stems, samples were collectedfrom homozygous and nil segregant T3 plants and immersed inCarnoy’s fixative (three parts 95% ethanol to one part glacial aceticacid, v/v) overnight at 4�C. Following removal of chlorophyll in 100%v/v ethanol, leaves and stems were then hydrated through a seriesof alcohol solutions (90, 70, 50 and 30%; 1 h in each), and finallytransferred to water. The tissue was immersed in a choral hydrate/glycerol/water solution (8 g:2 ml:1 ml) overnight, and mounted onslides in choral hydrate solution.
For 4¢,6¢-diamidino-2-phenylindole (DAPI) staining, grains werestaged from the time of anthesis, and collected 0, 2, 4 and 6 days afteranthesis. Collected grains were fixed in formalin/acetic acid/alcohol(FAA; 37% formaldehyde, 5% acetic acid, 50% ethanol), and vacuuminfiltrated for 15 min. After overnight fixation, samples were trans-ferred to a TissueTek vacuum infiltration processor (Bayer, http://www.bayer.co.uk/) for an automated dehydration/infiltration pro-cess as follows: 70% ethanol for 1 h at 35�C; 80% ethanol for 1.5 h at35�C; 90% ethanol for 2 h at 35�C; 100% ethanol for 1 h at 35�C; 100%ethanol for 1.5 h at 35�C; 100% ethanol for 2 h at 35�C; 100% xylenefor 0.5 h at 35�C; 100% xylene for 1.0 h at 35�C; 100% xylene for1.5 h at 35�C; molten paraffin wax (VWR International, http://www.vwr.com/) for 2 h at 60�C. Samples were then transferred tothe TissueTek embedding console for embedding. Wax-embeddedgrains were sectioned (10 lm thick) on a Leica microtome (http://www.leica.com/) and organized sequentially on slides coated withpoly-lysine. After overnight incubation at 42�C, slides were de-waxedas described by Drea et al. (2005), and used for staining.
De-waxed sections of staged Brachypodium grains were stainedwith DAPI solution (Partec, http://www.partec.com/) for 20 min, andviewed with a fluorescent microscope. Freshly collected antherswere dissected under a stereomicroscope, mounted on slides inAlexander stain (Alexander, 1969), and observed under a lightmicroscope.
Microscopy and image processing
For cellular analyses, samples were observed using a LeicaDM6000B microscope (20· objective), and images were recordedusing a digital camera. Pollen and anthers from 3–4-week-old
plants, as well as leaves and stems cleared in chloral hydrate wereobserved under bright-field conditions. Sections of DAPI-stainedgrains 0 and 6 days after anthesis were observed under UV illumi-nation.
For scanning electron microscopy (SEM), leaf samples, caryopsesand anthers were mounted on an aluminium stub using OCTcompound (BDH, https://www.vwrsp.com/), plunged into liquidnitrogen slush, then transferred onto the cryostage of an ALTO 2500cryo-transfer system (Gatan, http://www.gatan.com/) attached toa scanning electron microscope (Zeiss Supra 55 VP FEG, http://www.zeiss.com/). Samples were sputter-coated with platinum(90 sec at 10 mA, )110�C), then imaged at 3 kV on the cryostagein the main chamber of the microscope at approximately )130�C.
ACKNOWLEDGEMENTS
We thank Karen Browning (Department of Chemistry and Bio-chemistry, University of Texas at Austin, TX) for antibodies andfusion proteins to eIF4 proteins, the John Innes Centre horticulturalstaff for plant husbandry, and Andrew Davis (John Innes Centre Bio-imaging Facility) for photography. This work was supported by theUK Biotechnology and Biological Sciences Research Council; M.O.was funded by an EU Re-Integration Fellowship and J.H.D.acknowledges support from the Gatsby Charitable Foundation, theLeverhulme Trust and the EU Framework VII ‘AGRO-NOMICs’Research Programme. We thank Simon Griffith, Kay Trafford,Sinead Drea and John Snape for useful discussions and commentson the manuscript.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the onlineversion of this article:Movie S1. Video clip of Brachypodium eif4a T-DNA mutant plantgrowth compared with wild-type.Figure S1. Grain development in eif4a T-DNA mutant plants.Figure S2. Comparison of Arabidopsis (At3g13920) and Brachypo-dium (Bradi1g34170) eIF4A promoter regions.Table S1. Plant height, tiller and spikelet length of the Brachypo-dium eif4a T-DNA mutants.Table S2. Cell size in leaves of the Brachypodium eif4a T-DNAmutants.Please note: As a service to our authors and readers, this journalprovides supporting information supplied by the authors. Suchmaterials are peer-reviewed and may be re-organized for onlinedelivery, but are not copy-edited or typeset. Technical supportissues arising from supporting information (other than missingfiles) should be addressed to the authors.
REFERENCES
Alexander, M. (1969) Differential staining of aborted and nonaborted pollen.
Stain Technol. 44, 117–122.
Alves, S.C., Worland, B., Thole, V., Snape, J.W., Bevan, M.W. and Vain, P.
(2009) A protocol for Agrobacterium-mediated transformation of Brachyp-
odium distachyon community standard line Bd21. Nat. Protoc. 4, 638–649.
Bevan, M.W., Garvin, D.F. and Vogel, J.P. (2010) Brachypodium distachyon
genomics for sustainable food and fuel production. Curr. Opin. Biotechnol.
21, 211–217.
Blanc, G., Barakat, A., Guyot, R., Cooke, R. and Delseny, M. (2000) Extensive
duplication and reshuffling in the Arabidopsis genome. Plant Cell, 12,
1093–1101.
Bottley, A., Phillips, N.M., Webb, T.E., Willis, A.E. and Spriggs, K.A. (2010)
eIF4A inhibition allows translational regulation of mRNAs encoding pro-
teins involved in Alzheimer’s disease. PLoS ONE, 5, e13030.
Bush, M.S., Hutchins, A.P., Jones, A.M., Naldrett, M.J., Jarmolowski, A.,
Lloyd, C.W. and Doonan, J.H. (2009) Selective recruitment of proteins to 5¢cap complexes during the growth cycle in Arabidopsis. Plant J. 59, 400–412.
Brachypodium eif4a T-DNA mutant 939
ª 2011 The AuthorsThe Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 66, 929–940
Curtis, M.D. and Grossniklaus, U. (2003) A Gateway cloning vector set for
high-throughput functional analysis of genes in planta. Plant Physiol. 133,
462–469.
Daga, R.R. and Jimenez, J. (1999) Translational control of the cdc25 cell cycle
phosphatase: a molecular mechanism coupling mitosis to cell growth.
J. Cell Sci. 112, 3137–3146.
Dever, T.E. (2002) Gene-specific regulation by general translation factors. Cell,
108, 545–556.
Dorn, R., Morawietz, H., Reuter, G. and Saumweber, H. (1993) Identification of
an essential Drosophila gene that is homologous to the translation initia-
tion factor eIF-4A of yeast and mouse. Mol. Gen. Genet. 237, 233–240.
Draper, J., Mur, L.A., Jenkins, G., Ghosh-Biswas, G.C., Bablak, P., Hasterok, R.
and Routledge, A.P. (2001) Brachypodium distachyon. A new model sys-
tem for functional genomics in grasses. Plant Physiol. 127, 1539–1555.
Drea, S., Leader, D.J., Arnold, B.C., Shaw, P., Dolan, L. and Doonan, J.H. (2005)
Systematic spatial analysis of gene expression during wheat caryopsis
development. Plant Cell, 17, 2172–2185.
Galloni, M. and Edgar, B.A. (1999) Cell-autonomous and non-autonomous
growth-defective mutants of Drosophila melanogaster. Development, 126,
2365–2375.
Garvin, D.F., Gu, Y.Q., Hasterok, R., Hazen, S.P., Jenkins, G., Mockler, T.C.,
Mur, L.A.J. and Vogel, J.P. (2008) Development of genetic and genomic
research resources for Brachypodium distachyon, a new model system for
grass crop research. Crop Sci. 48, S69–S84.
Geraldo, N., Baurle, I., Kidou, S., Hu, X. and Dean, C. (2009) FRIGIDA delays
flowering in Arabidopsis via a cotranscriptional mechanism involving
direct interaction with the nuclear cap-binding complex. Plant Physiol. 150,
1611–1618.
Higgins, J.A., Bailey, P.C. and Laurie, D.A. (2010) Comparative genomics of
flowering time pathways using Brachypodium distachyon as a model for
the temperate grasses. PLoS ONE, 5, e10065.
Hutchins, A., Roberts, G., Lloyd, C.W. and Doonan, J.H. (2004) In vivo inter-
action between CDKA and eIF4A: a possible mechanism linking translation
and cell proliferation. FEBS Lett. 556, 91–94.
International Brachypodium Initiative (2010) Genome sequencing and anal-
ysis of the model grass Brachypodium distachyon. Nature, 463, 763–768.
Knopf, C., Becker, H., Ebmeyer, E. and Korzun, V. (2008) Occurrence of three
dwarfing rht genes in German winter wheat varieties. Cereal Res. Commun.
36, 553–560.
Koroleva, O.A., Calder, G., Pendle, A.F., Kim, S.H., Lewandowska, D., Simp-
son, C.G., Jones, I.M., Brown, J.W. and Shaw, P.J. (2009) Dynamic
behavior of Arabidopsis eIF4A-III, putative core protein of exon junction
complex: fast relocation to nucleolus and splicing speckles under hypoxia.
Plant Cell, 21, 1592–1606.
Lellis, A.D., Allen, M.L., Aertker, A.W., Tran, J.K., Hillis, D.M., Harbin, C.R.,
Caldwell, C., Gallie, D.R. and Browning, K.S. (2010) Deletion of the eIFiso4G
subunit of the Arabidopsis eIFiso4F translation initiation complex impairs
health and viability. Plant Mol. Biol. 4, 249–263.
Morinaka, Y., Sakamoto, T., Inukai, Y., Agetsuma, M., Kitano, H., Ashikari, M.
and Matsuoka, M. (2006) Morphological alteration caused by brassinos-
teroid insensitivity increases the biomass and grain production of rice.
Plant Physiol. 141, 924–931.
Opanowicz, M., Vain, P., Draper, J., Parker, D. and Doonan, J.H. (2008) Brac-
hypodium distachyon: making hay with a wild grass. Trends Plant Sci. 13,
172–177.
Opanowicz, M., Hands, P., Betts, D., Parker, M.L., Toole, G.A., Mills, E.N.,
Doonan, J.H. and Drea, S. (2011) Endosperm development in Brachypodi-
um distachyon. J. Exp. Bot. 62, 735–748.
Rogers, G.W. Jr, Komar, A.A. and Merrick, W.C. (2002) eIF4A: the godfather
of the DEAD box helicases. Prog. Nucleic Acid Res. Mol. Biol. 72, 307–
331.
Sakamoto, T. and Matsuoka, M. (2004) Generating high-yielding varieties by
genetic manipulation of plant architecture. Curr. Opin. Biotechnol. 15, 144–
147.
Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A
Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Labora-
tory Press.
Simpson, G.G., Quesada, V., Henderson, I.R., Dijkwel, P.P., Macknight, R. and
Dean, C. (2004) RNA processing and Arabidopsis flowering time control.
Biochem. Soc. Trans. 32, 565–566.
Thole, V., Alves, S.C., Worland, B., Bevan, M.W. and Vain, P. (2009) A protocol
for efficiently retrieving and characterizing flanking sequence tags (FSTs) in
Brachypodium distachyon T-DNA insertional mutants. Nat. Protoc. 4, 650–
661.
Thole, V., Worland, B., Wright, J., Bevan, M.W. and Vain, P. (2010) Distribution
and characterization of more than 1000 T-DNA tags in the genome of
Brachypodium distachyon community standard line Bd21. Plant Biotech-
nol. J. 8, 734–747.
Vain, P., Worland, B., Thole, V., McKenzie, N., Opanowicz, M., Fish, L.J.,
Bevan, M.W. and Snape, J.W. (2008) Agrobacterium-mediated transfor-
mation of the temperate grass Brachypodium distachyon (genotype
Bd21) for T-DNA insertional mutagenesis. Plant Biotechnol. J. 6, 236–
245.
Vision, T.J., Brown, D.G. and Tanksley, S.D. (2000) The origins of genomic
duplications in Arabidopsis. Science, 290, 2114–2117.
Worland, A., Sayers, E. and Korzun, V. (2001) Allelic variation at the dwarfing
gene Rht8 locus and its significance in international breeding programmes.
Euphytica, 119, 155–159.
940 Philippe Vain et al.
ª 2011 The AuthorsThe Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 66, 929–940