-
AtABCA9 transporter supplies fatty acids for lipidsynthesis to
the endoplasmic reticulumSangwoo Kima, Yasuyo Yamaokab, Hirofumi
Onob, Hanul Kima, Donghwan Shima, Masayoshi Maeshimac,Enrico
Martinoiaa,d, Edgar B. Cahoone, Ikuo Nishidab,1,2, and Youngsook
Leea,f,1,2
aDivision of Molecular Life Sciences, Pohang University of
Science and Technology, Pohang 790-784, Korea; bDivision of Life
Science, Graduate School ofScience and Engineering, Saitama
University, Saitama 338-8570, Japan; cGraduate School of
Bioagricultural Sciences, Nagoya University, Nagoya 464-8601,Japan;
dInstitute of Plant Biology, University of Zurich, 8008 Zurich,
Switzerland; eCenter for Plant Science Innovation, Department of
Biochemistry, Universityof Nebraska-Lincoln, Lincoln, NE 68588; and
fPohang University of Science and Technology–University of Zurich
Global Research Laboratory, Division ofIntegrative Biology and
Biotechnology, Pohang University of Science and Technology, Pohang
790-784, Korea
Edited by Maarten J. Chrispeels, University of California at San
Diego, La Jolla, CA, and approved November 28, 2012 (received for
review August 17, 2012)
Fatty acids, the building blocks of biological lipids, are
synthesizedin plastids and then transported to the endoplasmic
reticulum (ER)for assimilation into specific lipid classes. The
mechanism of fattyacid transport from plastids to the ER has not
been identified. Herewe report that AtABCA9, an ABC transporter in
Arabidopsis thali-ana, mediates this transport. AtABCA9 was
localized to the ER, andatabca9 null mutations reduced seed
triacylglycerol (TAG) contentby 35% compared with WT. Developing
atabca9 seeds incorpo-rated 35% less 14C-oleoyl-CoA into TAG
compared with WT seeds.Furthermore, overexpression of AtABCA9
enhanced TAG deposi-tion by up to 40%. These data strongly support
a role for AtABCA9as a supplier of fatty acid substrates for TAG
biosynthesis at the ERduring the seed-filling stage. AtABCA9 may be
a powerful tool forincreasing lipid production in oilseeds.
ABCA transporter | ABCA9 | acyl-CoA | fatty acid transporter
Fatty acids are essential for all forms of life, serving as
com-ponents of membrane and storage lipids and as precursors
ofsignaling molecules. Photosynthetic organisms, including
plantsand algae, are major suppliers of biological fatty acids for
otherorganisms. Fatty acids are synthesized in plastids, and a
subset isincorporated into plastidial glycerolipids. However, the
majorityof de novo synthesized fatty acids is assembled into
phospholipidsand neutral lipids at the endoplasmic reticulum (ER)
(1), ne-cessitating the transport of fatty acids from the plastid
to the ER.Although each step of fatty acid and lipid synthesis has
been
studied extensively for many decades, the process of fatty
acidtransport into the ER remains obscure. The identification
oftransporters of fatty acids or other lipophilic compounds is
in-trinsically difficult owing to the lack of an appropriate assay
systemfor the transport of such compounds across membranes. Thus,
theidentification of such transporters often depends on
biochemicalanalyses of the lipid levels in selected candidate
mutants. Trans-porters identified for lipophilic compounds belong
mostly to ATP-binding cassette (ABC) proteins in ABCA and ABCG
subfamiliesin animals (2) and in the ABCG subfamily in plants
(3).ABC proteins are ubiquitous in all living organisms, and
their
structures and functions are highly conserved (4). In animals,
manyimportant genetic diseases are associated with defects in
lipid-transporting ABC proteins (2, 5). Changing the activity of
someanimal ABCA proteins causes overaccumulation of lipids, such
astriacylglycerol (TAG) and cholesterol esters, in specific tissues
(6).In plants, many ABCGproteins secrete lipidic molecules that
formcutin and wax layers (3). Defects in the expression of such
proteinsresult in reduced surface lipids and/or developmental
defects, suchas organ fusion (7, 8).To identify the transporters
that deliver fatty acid substrates to
the ER for glycerolipid synthesis, we focused on ABCA andABCG
proteins. Our database searches revealed that no ABCGproteins and
only one animal ABCA protein, ABCA17, localizeat the ER membrane
(9). In the model plant Arabidopsis thaliana,
12 genes encode ABCA transporters, none of which has
beencharacterized previously (3, 10).We hypothesized that ABCA
proteins in plants mediate the
transport of fatty acids into the ER, and that the activity of
suchproteins is a limiting step in TAG biosynthesis in oilseed
plantsduring the seed-filling stage, when TAG biosynthesis is
greatest,and thus the flux of acyl-CoAs from the plastid to the ER
isgreatest as well. This hypothesis predicts that a knockout of
sucha transporter would decrease TAG synthesis, resulting in
re-duced seed storage lipid content.
ResultsScreening of ABCA Subfamily Genes for Altered Seed
Phenotypes. Toidentify the ABC transporters involved in fatty acid
transport tothe ER, we grew seeds of KO mutants of eight ABCA
familymembers on half-strengthMurashige and Skoog (1/2MS)mediumwith
or without sucrose, and compared early seedling growth withWT. The
rationale behind this test is that early seedling growth ofoilseed
plants depends on storage lipids in the absence of sucrose,but not
in its presence (11, 12). Thus, plants defective in expressionof
fatty acid-transporting ABC transporter would be expected toexhibit
reduced growth in the absence of sucrose, but normalgrowth in its
presence.Among the mutants tested, the growth of abca9-1 seedlings
was
most retarded on medium lacking sucrose (Fig. 1A). In
contrast,the growth of abca9-1 seedlings onmedium containing 1%
sucrosewas comparable to that of WT. The mutant phenotypes
wereconfirmed for two other abca9 alleles (abca9-2 and abca9-3;
Fig.S1). Thus, abca9 seeds are defective in the accumulation of
storagelipids or the conversion of lipids to sucrose.
abca9 Seeds Exhibit Reduced Seed Size and/or Abnormal
Morphology.Given that storage lipids account for ∼35–40% of the dry
seedweight of Arabidopsis thaliana (1), seeds defective in storage
lipidaccumulation are expected to be smaller ormisshapen. Among
theeight ABCA KO seeds tested, abca9-1 seeds displayed the
greatestvariability in size and shape (Fig. 1B). Using a dissecting
micro-scope, we classified abca9 seeds into four groups: normal,
flat,small, and brown and shrunken. Only 20% of abca9 seeds
werenormal, compared with 95% of WT seeds (Fig. S2A). Image
Author contributions: S.K., M.M., E.M., E.B.C., I.N., and Y.L.
designed research; S.K., Y.Y.,H.O., H.K., and D.S. performed
research; and S.K., E.M., E.B.C., I.N., and Y.L. wrotethe
paper.
Conflict of interest statement: S.K., Y.Y., H.O., I.N., and Y.L.
have filed patent PCT/KR2011/006826 and the title of patent is
“Composition for increasing seed size or content ofstorage lipid in
seed, comprising the ABC transporter protein-coding gene”.
This article is a PNAS Direct Submission.1I.N. and Y.L.
contributed equally to this work.2To whom correspondence may be
addressed. E-mail: [email protected]
[email protected].
This article contains supporting information online at
www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplemental.
www.pnas.org/cgi/doi/10.1073/pnas.1214159110 PNAS | January 8,
2013 | vol. 110 | no. 2 | 773–778
PLANTBIOLO
GY
Dow
nloa
ded
by g
uest
on
June
16,
202
1
http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplemental/pnas.201214159SI.pdf?targetid=nameddest=SF1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplemental/pnas.201214159SI.pdf?targetid=nameddest=SF1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplemental/pnas.201214159SI.pdf?targetid=nameddest=SF2mailto:[email protected]:[email protected]://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplementalhttp://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplementalwww.pnas.org/cgi/doi/10.1073/pnas.1214159110
-
analysis using ImageJ (http://rsbweb.nih.gov/ij) revealed
thatabca9 seeds were smaller and more varied in size than WT
seeds(Fig. S2B). Such variability in seed phenotype was observed
insubsequent generations as well, irrespective of the size and
shapeof the seeds from which the plants were generated (Fig.
S2C).To demonstrate genetic complementation of abca9, we trans-
formed the mutant plants with ABCA9 under the control of
itsnative promoter.We generated two independent
complementationlines for each of the three alleles of KO mutants.
The resulting sixtransgenic lines produced normal seeds and grew
normally onmedium lacking sucrose (Fig. 1B and Fig. S3), proving
that ABCA9is required for proper seed morphology and early
seedlingestablishment.
ABCA9 Is Expressed Specifically in Maturing Seeds.We evaluated
thespatial and temporal patterns ofABCA9 expression by
quantitativeRT-PCR analysis. ABCA9 transcripts were amplified only
slightlyfrom the whole seedlings, rosette leaves, stems, and
flowers,whereas high transcript levels were found in siliques,
especiallyduring the middle (S2) and late (S3) stages of seed
development(Fig. 1C). Together with the defective seed morphology,
this ex-pression pattern suggests that ABCA9 functions during seed
de-velopment. To test this, we evaluated developing seeds at 4, 8,
12,16, and 20 d after flowering (DAF) (Fig. 1D). Approximately
50%of abca9 seeds were paler and/or smaller than WT seeds at 8,
12(early S2 stage), and 16 (S3 stage) DAF. The temporal
manifes-tation of abnormal seed morphology in mutant siliques
coincidedwith the timing of ABCA9 transcript accumulation in WT
siliques,supporting our conclusion that the abnormal seed
phenotypes arecaused by disruption of ABCA9.
abca9 Seeds Have Reduced TAG Content. We then tested whetherthe
abca9 seeds indeed had reduced lipid content. Dry seedweight was
20% lower and total lipid content per seed was 16%
lower in mutant seeds compared with WT seeds, but proteincontent
did not differ (Fig. 1E). TAG, the major component oftotal lipids
in seeds, was ∼35% less abundant in abca9 seedscompared with WT
(Fig. 1F); however, the fatty acid compositionof TAG was comparable
in dry mutant and WT seeds (Fig. 1G),except for slightly lower
linolenate (18:3) levels in the mutants.These findings indicate
that abca9 does not affect the selectivity offatty acid
incorporation into TAG or further desaturation steps.
ABCA9 Is Localized at the ER. If ABCA9 is involved in fatty
acidtransport to theER, then it should be localized at this
organelle.Weexamined the subcellular localization of ABCA9 using
transgenicplants expressing ABCA9-sGFP under the control of the
CaMV35S promoter (Pro35S::ABCA9gDNA-sGFP). Green fluorescencewas
observed in the petioles (Fig. 2A) and isolated mesophyll
pro-toplasts (Fig. 2B) of the transgenic plants, in a distribution
re-sembling that of the ER.To further test the possibility that
ABCA9 is localized to the
ER, we fractionated crude membrane samples from
Pro35S::ABCA9gDNA-sGFP transgenic plants by sucrose density
gradientcentrifugation and evaluated ABCA9-sGFP using anti-sGFP
an-tibody. The distribution of ABCA9-sGFP matched that of the
ERmembrane marker BiP (Fig. 2C). To confirm the ER localizationof
ABCA9, we performed an ER membrane-shifting assay byadding EDTA,
which causes ribosomes to dissociate from the ER,thereby shifting
the distribution of ER markers to lower-densityfractions. Under
these conditions, both the BiP andABCA9 bandsshifted to lower
sucrose density fractions than in the presence of4 mM MgCl2,
further supporting the ER localization of ABCA9(Fig. 2D).
abca9 Is Defective in TAG Synthesis. Our findings of reduced
TAGcontent in abca9 seeds and localization of ABCA9 to the
ERsupport the idea that ABCA9 is involved in TAG biosynthesis
at
**
**
A
abca9-1WTabca9-1WT
1% sucroseno sucroseB
1-C1 2-C1
abca9-1WT
DC
**
**
*
E GF
Fig. 1. Characterization of ABCA9 KO mutants. (A)abca9-1 is
delayed in seedling growth on 1/2 MSmedium without sucrose (Left),
but not on mediumcontaining 1% sucrose (Right). (B) Seed
phenotypeof the WT, abca9-1, and two complementation lines,1-C1 and
2-C1. (Scale bar: 500 μm.) (C) Transcriptlevels of ABCA9 in various
tissues of WT. WS, wholeseedlings; L, rosette leaves; S, stems; F,
flowers; S1,siliques from 4∼6 DAF; S2, siliques from 10∼12 DAF;S3,
siliques from 16∼18 DAF. (D) WT and abca9 seedsduring the
seed-filling stages. Red arrows indicatedefective seeds. (E) Seed,
lipid, and protein weightper seed. (F) TAG content per seed. (G)
Fatty acidcomposition of TAG. *P < 0.05; **P < 0.01. Error
barsindicate SEs from three replicates.
774 | www.pnas.org/cgi/doi/10.1073/pnas.1214159110 Kim et
al.
Dow
nloa
ded
by g
uest
on
June
16,
202
1
http://rsbweb.nih.gov/ijhttp://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplemental/pnas.201214159SI.pdf?targetid=nameddest=SF2http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplemental/pnas.201214159SI.pdf?targetid=nameddest=SF2http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplemental/pnas.201214159SI.pdf?targetid=nameddest=SF3www.pnas.org/cgi/doi/10.1073/pnas.1214159110
-
the ER membrane, the site of TAG biosynthesis (1). To
furthertest this possibility, we fed 14C-acetate to actively
growing seeds(at 10 DAF) isolated from WT and abca9 siliques, and
thenmeasured the radioactivity incorporated into the TAG
fraction.We used 14C-acetate because it is readily converted into
acetyl-CoA and malonyl-CoA for fatty acid biosynthesis (13).
Theamount of radioactivity incorporated into TAG in developingabca9
seeds was approximately half that in WT seeds (Fig. 3A).Reduced TAG
synthesis in abca9 seeds was confirmed in time-chasing experiments
(Fig. 3B).
To test whether ABCA9 can supply fatty acids for TAG syn-thesis,
we examined whether ABCA9 facilitates assimilation ofexogenously
supplied acyl-CoAs and free fatty acids into TAG. At10 DAF,
developing seeds of WT and abca9 were fed 14C-oleoyl-CoA and
14C-oleic acid. The amount of radioactivity incorporatedinto TAG
from 14C-oleoyl-CoA and oleic acid was significantlylower in abca9
seeds than in WT seeds (Fig. 3A). These results arein agreement
with the reduced TAG content in dry abca9 seeds,indicating that
ABCA9 supplies fatty acid substrates for seed TAGbiosynthesis at
the ER membrane.
A
C
ABCA9-sGFP
BiP
PIP2;1
γ-TIPER
PMVM
17.8
19.0
20.6
21.4
22.4
23.6
24.6
26.0
27.2
28.6
29.6
31.0
32.4
34.4
35.4
37.0
38.0
39.4
41.0
42.0
42.8
43.8
44.8
16.0Sucrose
w/w (%)
(4 mM MgCl2)
ER
PM
VM
ABCA9-sGFP
BiP
PIP2;1
γ-TIP
Sucrosew/w (%)17
.6
19.0
20.0
21.0
22.4
23.4
24.6
25.6
27.0
28.6
29.4
30.8
32.4
34.0
35.2
36.6
38.0
39.6
41.0
42.0
42.6
44.0
45.0
16.0
(4 mM EDTA)D
ABCA-GFP BiP-RFP MergedB
Fig. 2. Localization of ABCA9 at the endoplasmicreticulum. (A)
sGFP signal from the petiole of thePro35S::ABCA9gDNA-sGFP T3 single
homozygousline. (Scale bar: 20 μm.) (B) Overlap of GFP andRFP
fluorescence in protoplasts isolated from thePro35S::ABCA9gDNA-sGFP
T3 single homozygousline and transformed with the ER marker
Pro35S::BiP-RFP. (Scale bar: 5 μm.) (C and D) Intracelluar
lo-calization of ABCA9-sGFP detected by fractionationof microsomes
on a sucrose density gradient in thepresence of 4 mM MgCl2 (C) or 4
mM EDTA (D). The24 fractions were collected and immunoblotted
withantibodies to sGFP and ER (BiP) and with vacuolarmembrane (VM;
γ-TIP) and plasma membrane (PM;PIP2;1) markers.
A
B C
**
*
*
abca9
Acyl-CoAs Fatty acids
WT
Normal acyl-CoA Pool
TAG
Reduced acyl-CoA Pool
TAG
ER
ER
Cytosol
Cytosol
ADPATP
ADPATP
ABCA9
Fig. 3. Reduced TAG biosynthesis in developingabca9 seeds. (A)
Incorporation into TAG from 14C-acetate, 14C-oloeyl-CoA, and
14C-oleic acid was testedusing 50 developing seeds at 10 DAF. Total
proteincontent did not differ between the WT and abca9seeds (WT,
637.8 ± 19.61 μg; abca9, 618.2 ± 15.73 μg).Six replicates were
averaged, and the SE is shown.*P < 0.05; **P < 0.01, Student
t test. (B) Time de-pendence of 14C-acetate incorporation into TAG.
(C)Working hypothesis. The ABCA9 transporter (in red)transports
both acyl-CoAs and fatty acids to the ER,thereby facilitating TAG
synthesis.
Kim et al. PNAS | January 8, 2013 | vol. 110 | no. 2 | 775
PLANTBIOLO
GY
Dow
nloa
ded
by g
uest
on
June
16,
202
1
-
Overexpression of ABCA9 Increases Seed Oil Content. To
testwhether ABCA9 mediates a limiting step in lipid
accumulationduring seed filling, we generated Arabidopsis thaliana
plantsexpressing full-length genomic DNA of ABCA9 under the CaMV35S
promoter (Pro35S::ABCA9gDNA) (Fig. S4A). Interestingly,the
ABCA9-overexpressing plants produced enlarged seeds (Fig.4A), with
dry weight up to 126% of that of WT (Fig. 4D). Cot-yledon cells of
ABCA9-overexpressing embryos were larger thanthose of WT (Fig. 4B),
and ABCA9-overexpressing cells weremore densely packed with oil
bodies compared with WT (Fig.4C). In contrast, protein body
frequency and size did not sig-nificantly differ between
ABCA9-overexpressing and WT cells(Fig. 4C). The TAG content per
seed in ABCA9-overexpressinglines was up to 140% of that of WT,
with no significant differ-ences in fatty acid composition (Fig. 4E
and Fig. S4B).We performed biochemical analyses of protein and
carbohy-
drates to investigate whether increased lipid levels affected
otherseed reserves. Comparison of ABCA9-overexpressing andWT
linesrevealed significantly increased lipids in
ABCA9-overexpressinglines with no reduction in other seed reserves
(Table S1). Given that
silique number per plant (average±SE, 130.7±10.5 forWT,
125.4±9.1 for overexpression line 2, 112.4 ± 11.2 for
overexpression line4, and 116.2± 10.2 for overexpression line 6)
and seed number perhalf-silique (average ± SE, 26.7 ± 0.8 for WT,
26.9 ± 0.4 foroverexpression line 2, and 26.6 ± 0.9 for
overexpression line 6)were not significantly different between
ABCA9-overexpressinglines and WT (Fig. S4 C and D), we conclude
that overexpressionof ABCA9 can increase the total seed oil yield
per plant.
DiscussionA transporter that mediates the transport of acyl-CoAs
and/orfree fatty acids from plastids to the ER has not yet been
iden-tified in any living system. A recent transcriptome analysis
ofoilseed-specific gene expression revealed increased levels
ofmRNAs involved in fatty acid synthesis and acyl-CoA efflux
fromplastids during seed filling (14); however, cytosolic
acyl-CoAsinhibit plastidial fatty acid biosynthesis by inhibiting
plastidialmetabolite transporters (15, 16). In addition, cytosolic
acyl-CoAsand/or free fatty acids are toxic to the cell, and thus
must be
** ** **
** **** **
D
WT OX2OX2 OX5OX2 WTWTA B
Line Area of cotyledon, mm2 Area per cell, μm2
WT 0.09 (100) 150.23 4.31 (100)OX2 0.11 (122) 179.15 4.49**
(119)
COX4WT
PB
PBOB
* *** ***
E*
Fig. 4. ABCA9-overexpressing plants produce enlarged seedswith
elevated averageweight and lipid content. (A) Developing seeds at
12DAF (Left) and dry seeds(Right) fromWTandABCA9-overexpressing
plants. (B)Mature embryos isolated frommature dry seeds, imbibed
for 1 h, of theWTandABCA9–overexpressing line(Upper) and cotyledon
cells from these embryos (Lower). (Scale bars:Upper, 100 μm; Lower,
10 μm.) Surface areas of cotyledons andof individual cells of
cotyledonsmeasured from images of the embryos are listed in the
table. (C) Increased density of oil bodies in
anABCA9-overexpressing line. Note that OX4 cells containmanyoil
bodies that exclude toluidine blue dye (Right), which stains the
cytosol ofWT cells purple (Left). OB, oil body; PB, protein body.
(Scale bars: 5 μm.) (D and E) Dryseed weight (D) and TAG level (E)
in WT and ABCA9-overexpressing plants. *P < 0.05; **P < 0.01,
Student t test. Error bars indicate SE from four replicates.
776 | www.pnas.org/cgi/doi/10.1073/pnas.1214159110 Kim et
al.
Dow
nloa
ded
by g
uest
on
June
16,
202
1
http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplemental/pnas.201214159SI.pdf?targetid=nameddest=SF4http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplemental/pnas.201214159SI.pdf?targetid=nameddest=SF4http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplemental/pnas.201214159SI.pdf?targetid=nameddest=ST1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplemental/pnas.201214159SI.pdf?targetid=nameddest=SF4www.pnas.org/cgi/doi/10.1073/pnas.1214159110
-
rapidly transported into the ER to maintain the high rates
offatty acid and TAG biosynthesis during seed filling.We propose
that ABCA9 acts as such a transporter.We base this
conclusion on the following lines of evidence: (i) TAG levels
arereduced in abca9 and increased in ABCA9-overexpressing
seeds;(ii) ABCA9 is expressed specifically in seeds at the middle
and latestages of maturation, when storage lipids are rapidly
accumulatedand the rate of TAG synthesis is greatest (17); (iii)
the temporalprofile ofABCA9 transcript accumulation is closely
correlated withthe manifestation of abnormal seeds in developing
abca9 siliques;(iv) TAG synthesis is reduced in abca9 seeds, as
demonstrated byassimilation experiments with 14C-acetate,
14C-oleoyl-CoA, and14C-oleic acid; (v) ABCA9 is localized at the ER
and belongs toa subfamily of lipid transporters, ABCA; and (vi) the
relativeproportions of fatty acids in TAGs are similar in WT and
abca9seeds, suggesting that the desaturation of esterified fatty
acids onthe glycerol backbone is not affected in abca9 and in turn,
that onlythefirst step at theER, but not the later steps of
lipidmetabolism, isaltered in the mutant (18–21). If any steps
further downstream inthe TAG biosynthesis pathway were defective,
then a marked al-teration in the fatty acid composition of TAGs
would be expected,because the ER is the site of extensive
desaturation of fatty acids.Thus, we conclude that ABCA9
facilitates acyl-CoA uptake
into the ER, thereby enhancing the assembly of acyl-CoAs
intoTAGs. Cytosolic acyl-CoA binding proteins also may contributeto
the transfer of acyl-CoAs from plastids to the ER (22).
Theseproteins reportedly bind acyl-CoAs and maintain the
acyl-CoApool size in the cytosol. Whether they are critical for TAG
bio-synthesis in developing seeds remains to be determined,
however.In ABCA9 overexpression lines, embryo cells are larger
and
contain more oil bodies than WT cells (Fig. 4C). As a
result,ABCA9 overexpression increases seed size without changing
thenumber of seeds per silique or the number of silique per
plant(Fig. 4A and Fig. S4 C and D), resulting in an overall
increase inseed oil yield per plant. Of note, there were no changes
in proteinor carbohydrate levels in ABCA9-overexpressing seeds
(Table S1).Thus, ABCA9 seems to increase sink capacity specific to
storagelipids by facilitating the transport of activated fatty
acids into theER. The highly expressed ABCA9most likely increases
the size ofthe acyl-CoA pool in the ER, which may facilitate TAG
synthesisby increasing substrate concentrations or extending the
criticalperiod of TAG synthesis during seed filling (Fig. 3C). The
in-creased oil production by ABCA9 overexpression suggests
thatuptake of acyl-CoAs into the ER is a limiting step in
metaboliteflux during seed filling.A previous genetic engineering
strategy was to increase seed oil
content by introducing enzymes and transcription factors
involvedin lipid metabolism. Overexpression of acetyl-CoA
carboxylase orfatty acid synthase had little effect on the level of
lipid accumu-lation (23, 24), whereas overexpression of enzymes
catalyzingTAG biosynthesis, such as glycerol-3-phosphate
acyltransferase,lysophosphatidic acid acyltransferase, and
diacylglycerol acyl-transferase, significantly increased seed oil
yield, by 10–40% (25–29). Overexpression of the transcription
factorWRINKLED1 alsoincreased seed oil yield by 20% (30). ABCA9
overexpressionprovides a valuable option for increasing the TAG
content inseeds, demonstrating up to a 40% increase in seed oil
yield with noreduction in protein and carbohydrate content. The use
of a lipidtransporter is a unique approach to increasing the
vegetable oilcontent of seeds that can be combined with other
methodsenforcing seed metabolic functions. Given that the global
con-sumption of vegetable oils is expected to double by 2030
(31),this strategy may be valuable in exploring ways to meet
theurgent need for increased oil production.
Materials and MethodsPlant Materials and Growth Conditions.
Arabidopsis thaliana seeds were sur-face-sterilized, placed in the
dark at 4 °C for 2 d, and then randomly sownon1/2MS-agar plates
with 1% sucrose. Plates were incubated for 2–3 wk (22/18 °C;
16/8 h day/night). For further analyses, plants were transferred
to soil andgrown in a greenhouse (18/16 °C; 16/8 h day/night).
Isolation of ABCA9 KO Mutants. Seeds of three alleles ofABCA9
KOArabidopsisthaliana mutants—SALK_058070, SALK_023744, and
SALK_084342—wereobtained from the Salk Institute Genomic Analysis
Laboratory (http://signal.salk.edu/cgi-bin/tdnaexpress). Genomic
DNA (gDNA) was extracted from plantsgrown for 4 wk on soil.
Homozygous ABCA9 KO plants were isolated by PCRusing a
T-DNA–specific primer (pROKLBb1: 5′-GCGTGGAACCGCTTGCTGCAACT-3′) and
four ABCA9-specific primers (SALK_058070LP:
5′-CTACATATGGCTCGT-GGGAAC-3′; SALK_058070RP:
5′-AAAGAGGTGGAGGTGCTCTTC; SALK_084342LP:5′-ATGACTCTGCGAGAAGGCTT-3′;
and SALK_084342RP: 5′-GAAAGAGACCAAA-CCACACC-3′).
DNA Constructs. To generate complemented and tagged lines of
abca9, sGFPwas amplified from the 326-sGFP (kindly providedby
InhwanHwang, POSTECH,Korea) vector using primers containing SpeI
and PmlI restriction sites (5′-ACTAGTATGGTGAGCAAGGGCGAGGA-3′ and
5′-CACGTGTTACTTGTACAGCTCGTCCATG-3′) and inserted into the SpeI and
PmlI sites of pCAMBIA1302(Cambia, Canberra, Australia). For the
complementation line, the ABCA9 pro-moter was amplified using
primers containing KpnI and SpeI sites
(5′-GGTACCGCACGGTGTGAACATTAATT-3′ and
5′-ACTAGTGATCACAGAGGAAGAAGAAG-3′)and inserted into the KpnI and
SpeI sites of each construct. Finally, full-lengthgenomic
DNAofABCA9was amplified usingprimers containing the SpeI site
(5′-ACTAGTATGACTCTGCGAGAAGGCTT-3′ and
5′-ACTAGTTTCATTGTTAGATTCA-TAAT-3′) and ligated into the SpeI site
of the construct.
To generate the ABCA9 overexpression construct, genomic DNA of
ABCA9was amplified using primers containing the SpeI site
(5′-ACTAGTATGA-CTCTGCGAGAAGGCTT-3′ and
5′-ACTAGTTTCATTGTTAGATTCATAAT-3′) andinserted into the SpeI site of
pCAMBIA1302. All constructs were verifiedby sequencing.
Seed Lipid, Protein, and Carbohydrate Analysis. A total of 500
Arabidopsisthaliana seeds were immersed in 1 mL of boiling
isopropanol and heated for 5min at 80 °C. After cooling, 2 mL of
chloroform was added to the sample, andthe plant material was
finely ground with a Polytron homogenizer (HitachiKoki). The
extract was centrifuged at 1,600 × g for 10 min. The resultant
su-pernatant was decanted to a new 10-mL screw-capped glass tube,
and the pelletwas reextracted with 2 mL of chloroform and 1 mL of
methanol by vortexing.After centrifugation at 1,600 × g for 10 min,
the supernatant was recoveredby decantation, combined with the
first supernatant, and then washed with 1.2mL of 0.9% KCl by
vigorous shaking. Following centrifugation at 1,600 × g for15 min,
the lower layer was recovered into a weighted 20-mL pear-shaped
flask,and the solvent was evaporated on a rotary evaporator. The
residual solventwas dried using a vacuum desiccator, and lipid dry
weight was determined.Dried lipid residues were dissolved in
chloroform at a concentration of 10 mg/mL and stored in a 1-mL
screw-capped sample tube at −30 °C until use.
Total lipids (1 mg) were separated by silica gel TLC using a
solvent mixture[80:30:1 (by volume) hexane/diethylether/acetic
acid] that facilitated theseparation of neutral lipids. Lipid spots
were visualized by spraying with0.01% (wt/vol) primuline reagent
(Sigma-Aldrich), and each lipid class wastransformed into fatty
acid methyl esters at 80 °C for 3 h in 3 mL of 5% (wt/vol) HCl in
methanol, with an additional 47.8 nmol pentadecanoic acid as
aninternal standard and 50 nmol 2,6-di-t-butyl-4-methylphenol as an
antioxi-dant. The resultant fatty acid methyl esters were extracted
with 3 mL ofhexane and quantified by GC.
For total protein assays, 50 seeds were homogenized in 200 μL of
extractionbuffer [1% (wt/vol) SDS 6 M urea]. Protein content was
measured in 200 μL ofthe crude homogenate using the Bio-Rad DC
Protein Assay Kit with BSA asa standard. Carbohydrates were
analyzed as described by Siloto et al. (32) withsome modifications.
A total of 300 seeds were homogenized in 80% (vol/vol)ethanol and
incubated at 70 °C for 90 min. After centrifugation at 16,000 ×
gfor 5 min, the supernatant was transferred to a new test tube. The
pellet wasextracted three times with 500 μL of 80% (vol/vol)
ethanol, and the solvent ofthe combined supernatants was evaporated
at room temperature undera vacuum. This residue was dissolved in
0.1 mL of water and used for sucrosequantification. The pellet
remaining after ethanol extraction was homoge-nized in 200 μL of
0.2 M KOH and then incubated at 95 °C for 1 h. After theaddition of
35 μL of 1 M acetic acid and centrifugation for 5 min at 16,000 ×
g,the supernatant was used for starch quantification. Sucrose
content and starchcontent were measured using kits from
Sigma-Aldrich.
Tissue-Specific Expression of ABCA9. To investigate the level of
ABCA9transcript in different tissues, real-time PCR was performed
using totalRNA extracted from each organ. To eliminate the
possibility of genomic
Kim et al. PNAS | January 8, 2013 | vol. 110 | no. 2 | 777
PLANTBIOLO
GY
Dow
nloa
ded
by g
uest
on
June
16,
202
1
http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplemental/pnas.201214159SI.pdf?targetid=nameddest=SF4http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214159110/-/DCSupplemental/pnas.201214159SI.pdf?targetid=nameddest=ST1http://signal.salk.edu/cgi-bin/tdnaexpresshttp://signal.salk.edu/cgi-bin/tdnaexpress
-
DNA contamination, real-time PCR was performed using ABCA9
cDNA-specific primers (5′-TGCTGTAAAGGGTTTGTGGA-3′ and
5′-TGCCAGTAG-TCGGTTCATCT-3′).
Membrane Fractionation Using Sucrose Gradient Centrifugation
Arabidopsisthaliana plants (Pro35S::ABCA9gDNA-sGFP) were cultured
for 2 wk in liquid 1/2MS medium. Then 10 g of plant seedlings (50
seedlings) was homogenized in40mL of homogenized solution [250 mM
sorbitol, 50 mM Tris-acetate (pH 7.5),1 mM EGTA-Tris (pH 7.5), 2 mM
DTT, 1× protease inhibitor mixture (PIC, EDTA-free; Roche Applied
Science), 2% (wt/vol) polyvinylpyrrolidone, and 4 mMEDTA or MgCl2]
using a mortar and pestle. The sample was filtered throughnylon
mesh and centrifuged at 2,000 × g for 10 min at 4 °C. The
supernatantwas recentrifuged twice at 10,000 × g for 10 min at 4
°C. The supernatantwas ultracentrifuged at 100,000 × g for 1 h at 4
°C (SW28 rotor; BeckmanCoulter). The pellet was suspended in
microsome resuspension buffer con-taining 5% (wt/vol) sucrose, 20
mM Tris-acetate (pH 7.5), 0.5 mM EGTA-Tris,1× PIC, and 4 mM EDTA or
MgCl2. Then 2.5 mg of the crude membranesuspension was applied to
the sucrose gradient solution [10–50% sucroselinear gradient, 20 mM
Tris·HCl (pH 7.5), 0.5 mM EGTA-Tris, 1× PIC, and 4 mMEDTA or
MgCl2], centrifuged (slow acceleration, no break) at 100,000 × g
for16 h at 4 °C (SW41Ti rotor; Beckman Coulter) and fractionated
into 28fractions of 410 μL each. The fractionated proteins were
separated by SDS/PAGE and transferred to a PVDF membrane. The
membrane was blockedwith 5% (wt/vol) skim milk and incubated with
organelle-specific primaryantibodies (BiP, γ-TIP, and PIP2;1 for
the ER, vacuole, and plasma membrane,respectively). HRP-conjugated
goat anti-mouse or goat anti-rabbit IgG wasapplied as the secondary
antibody.
ABCA9:sGFP was detected using anti-sGFP (Clontech) and goat
anti-mouseantibody. The transfer solution contained 48 mM Tris, 39
mM glycine, 0.02%SDS, and 20% (vol/vol) methanol. Blocking solution
contained TBST buffer [0.5M Tris base (pH 7.6), 9% (wt/vol) NaCl,
and 1% (vol/vol) Tween-20] and 5% (wt/vol) skim milk. The membrane
was washed with TBST buffer. Antibody con-centrations were 1:2,000
for anti-sGFP and 1:3,000 for anti-BiP, anti-PIP2;1,anti–γ-TIP,
anti-mouse, and anti-rabbit. Incubation with primary antibodieswas
carried out overnight at 4 °C, and incubation with secondary
antibodieswas done for 1 h at room temperature (∼25 °C) with gentle
shaking.
Assay Monitoring Incorporation of Precursors into TAG. To
compare the ratesof incorporation of precursors into TAG, 50
Arabidopsis thaliana seeds were
collected from WT and abca9 siliques at 10 DAF and transferred
to 200 μL of20 mM MES buffer (pH 5.8). One of the following
14C-labeled compoundswas added at the specific activities
indicated: 0.5 μCi of 14C-acetate (50 mCi/mmol), 0.5 μCi of
14C-oleoyl-CoA (40–60 mCi/mmol), or 0.5 μCi of 14C-oleicacid (40-60
mCi/mmol). The seeds were incubated for 18 h in the dark whilebeing
rotated at 100 rpm.
In the 14C-acetate time-course assay, the seeds were incubated
for 9, 12,and 18 h under the same conditions. The seeds were then
washed with 1 mLof ice-cold water and homogenized in 50 μL of
chloroform:methanol:formicacid (10:10:1 by volume). The organic and
aqueous phase were separated byadding 12.5 μL of solution
consisting of 1 M KCl and 0.2 M H3PO4 andcentrifuging at 16,000 × g
for 5 min. The lipids in the lower phase wereseparated on a silica
TLC plate (1.05721.0001; Merck) with hexane:diethylether:acetic
acid (80:30:1 by volume). Silica material containingthe TAG was
scraped from the TLC plate and mixed with the scintillationmixture.
Radioactivity was measured by scintillation counting.
Oil Body Observation. To observe oil bodies in embryo cells, dry
seeds of WTand ABCA9-overexpressing line were imbibed for 1 h and
then cut in halfwith a razor blade. The samples were fixed
overnight at 4 °C in 20 mMcacodylate buffer (pH 7.0) containing 3%
(wt/vol) paraformaldehyde and2.5% (vol/vol) glutaraldehyde, and
then rinsed in cacodylate buffer andfurther fixed in 1% (wt/vol)
osmium tetraoxide for 1 h at 4 °C. The sampleswere dehydrated using
a graded ethanol series and embedded in LR Whiteresin (Electron
Microscopy Sciences). Serial sections of the samples werestained
with 1% (wt/vol) toluidine blue and observed by light
microscopy.
ACKNOWLEDGMENTS. We thank the Salk Institute Genomic Analysis
Labora-tory for the Arabidopsis thaliana mutant seeds and J. L.
Harwood andW. Dewitte for their insightful comments on the
manuscript. Research in theY.L. laboratory was supported by grants
from the Global Research Laboratoryprogram of the Ministry of
Science and Technology; the Next-Generation Bio-green 21 Program
(Grant PJ008102), Rural Development Administration; andthe Global
Frontier Program (Grant 2011-0031345) of the Republic of
Korea.Research in the I.N. laboratory was funded by Grants-in-Aid
for Scientific Re-search 21570034 and 24570040 from theMinistry of
Education, Culture, Sports,Science and Technology of Japan.
Research in the E.B.C. laboratory was sup-ported by the Center for
Advanced Biofuel Systems, an Energy Frontier Re-search Center
funded by the US Department of Energy, Office of Science,Office of
Basic Energy Sciences (Award DE-SC0001295).
1. Li-Beisson Y, et al. (2010) Acyl-lipid metabolism.
Arabidopsis Book 8:e0133.2. van Meer G, Voelker DR, Feigenson GW
(2008) Membrane lipids: Where they are and
how they behave. Nat Rev Mol Cell Biol 9(2):112–124.3. Kang J,
et al. (2011) Plant ABC transporters. Arabidopsis Book 9:e0153.4.
Higgins CF (1992) ABC transporters: From microorganisms to man.
Annu Rev Cell Biol
8:67–113.5. Kim WS, Weickert CS, Garner B (2008) Role of
ATP-binding cassette transporters in
brain lipid transport and neurological disease. J Neurochem
104(5):1145–1166.6. Oram JF, Vaughan AM (2006) ATP-binding cassette
cholesterol transporters and car-
diovascular disease. Circ Res 99(10):1031–1043.7. Pighin JA, et
al. (2004) Plant cuticular lipid export requires an ABC
transporter. Science
306(5696):702–704.8. Bird D, et al. (2007) Characterization of
Arabidopsis ABCG11/WBC11, an ATP binding
cassette (ABC) transporter that is required for cuticular lipid
secretion. Plant J 52(3):485–498.
9. Ban N, Sasaki M, Sakai H, Ueda K, Inagaki N (2005) Cloning of
ABCA17, a novel rodentsperm-specific ABC (ATP-binding cassette)
transporter that regulates intracellularlipid metabolism. Biochem J
389(Pt 2):577–585.
10. Verrier PJ, et al. (2008) Plant ABC proteins—a unified
nomenclature and updatedinventory. Trends Plant Sci
13(4):151–159.
11. Cernac A, Andre C, Hoffmann-Benning S, Benning C (2006) WRI1
is required for seedgermination and seedling establishment. Plant
Physiol 141(2):745–757.
12. Andre C, Froehlich JE, Moll MR, Benning C (2007) A
heteromeric plastidic pyruvatekinase complex involved in seed oil
biosynthesis in Arabidopsis. Plant Cell 19(6):2006–2022.
13. Benning C (2008) A role for lipid trafficking in chloroplast
biogenesis. Prog Lipid Res47(5):381–389.
14. Bourgis F, et al. (2011) Comparative transcriptome and
metabolite analysis of oil palmand date palm mesocarp that differ
dramatically in carbon partitioning. Proc NatlAcad Sci USA
108(30):12527–12532.
15. Fox SR, Rawsthorne S, Hills MJ (2001) Fatty acid synthesis
in pea root plastids is in-hibited by the action of long-chain
acyl-coenzyme as on metabolite transporters.Plant Physiol
126(3):1259–1265.
16. Yurchenko OP, Weselake RJ (2011) Involvement of low
molecular mass soluble acyl-CoA-binding protein in seed oil
biosynthesis. New Biotechnol 28(2):97–109.
17. Ruuska SA, Girke T, Benning C, Ohlrogge JB (2002)
Contrapuntal networks of geneexpression during Arabidopsis seed
filling. Plant Cell 14(6):1191–1206.
18. Xu C, Fan J, Froehlich JE, Awai K, Benning C (2005) Mutation
of the TGD1 chloroplastenvelope protein affects phosphatidate
metabolism in Arabidopsis. Plant Cell 17(11):3094–3110.
19. Lu C, Xin Z, Ren Z, Miquel M, Browse J (2009) An enzyme
regulating triacylglycerolcomposition is encoded by the ROD1 gene
of Arabidopsis. Proc Natl Acad Sci USA106(44):18837–18842.
20. Bates PD, Browse J (2011) The pathway of triacylglycerol
synthesis through phos-phatidylcholine in Arabidopsis produces a
bottleneck for the accumulation of unusualfatty acids in transgenic
seeds. Plant J 68(3):387–399.
21. Andrianov V, et al. (2010) Tobacco as a production platform
for biofuel: over-expression of Arabidopsis DGAT and LEC2 genes
increases accumulation and shifts thecomposition of lipids in green
biomass. Plant Biotechnol J 8(3):277–287.
22. Xiao S, Chye ML (2009) An Arabidopsis family of six
acyl-CoA-binding proteins hasthree cytosolic members. Plant Physiol
Biochem 47(6):479–484.
23. Roesler K, Shintani D, Savage L, Boddupalli S, Ohlrogge J
(1997) Targeting of theArabidopsis homomeric acetyl-coenzyme A
carboxylase to plastids of rapeseeds. PlantPhysiol
113(1):75–81.
24. Thelen JJ, Ohlrogge JB (2002) Metabolic engineering of fatty
acid biosynthesis inplants. Metab Eng 4(1):12–21.
25. Jain RK, Coffey M, Lai K, Kumar A, MacKenzie SL (2000)
Enhancement of seed oilcontent by expression of
glycerol-3-phosphate acyltransferase genes. Biochem SocTrans
28(6):958–961.
26. Zou J, et al. (1997) Modification of seed oil content and
acyl composition in thebrassicaceae by expression of a yeast sn-2
acyltransferase gene. Plant Cell 9(6):909–923.
27. Jako C, et al. (2001) Seed-specific over-expression of an
Arabidopsis cDNA encodinga diacylglycerol acyltransferase enhances
seed oil content and seed weight. PlantPhysiol 126(2):861–874.
28. Lardizabal K, et al. (2008) Expression of Umbelopsis
ramanniana DGAT2A in seedincreases oil in soybean. Plant Physiol
148(1):89–96.
29. Oakes J, et al. (2011) Expression of fungal diacylglycerol
acyltransferase2 genes toincrease kernel oil in maize. Plant
Physiol 155(3):1146–1157.
30. Liu J, et al. (2010) Increasing seed mass and oil content in
transgenic Arabidopsis bythe overexpression of wri1-like gene from
Brassica napus. Plant Physiol Biochem48(1):9–15.
31. Carlsson AS, Yilmaz JL, Green AG, Stymne S, Hofvander P
(2011) Replacing fossil oilwith fresh oil: With what and for what?
Eur J Lipid Sci Technol 113(7):812–831.
32. Siloto RMP, et al. (2006) The accumulation of oleosins
determines the size of seedoilbodies in Arabidopsis. Plant Cell
18(8):1961–1974.
778 | www.pnas.org/cgi/doi/10.1073/pnas.1214159110 Kim et
al.
Dow
nloa
ded
by g
uest
on
June
16,
202
1
www.pnas.org/cgi/doi/10.1073/pnas.1214159110