Arabidopsis Vacuolar Sorting Mutants (green fluorescent ... · Green Fluorescent Protein in Their...
Transcript of Arabidopsis Vacuolar Sorting Mutants (green fluorescent ... · Green Fluorescent Protein in Their...
Arabidopsis Vacuolar Sorting Mutants (green fluorescent seed)Can Be Identified Efficiently by Secretion of Vacuole-TargetedGreen Fluorescent Protein in Their Seeds W OA
Kentaro Fuji,a Tomoo Shimada,a Hideyuki Takahashi,a Kentaro Tamura,a Yasuko Koumoto,a Shigeru Utsumi,b
Keito Nishizawa,b Nobuyuki Maruyama,b and Ikuko Hara-Nishimuraa,1
a Department of Botany, Graduate School of Science, Kyoto University, Kyoto 606-8502, Japanb Graduate School of Agriculture, Kyoto University, Gokasho, Uji, Kyoto 611-0011, Japan
Two Arabidopsis thaliana genes have been shown to function in vacuolar sorting of seed storage proteins: a vacuolar
sorting receptor, VSR1/ATELP1, and a retromer component, MAIGO1 (MAG1)/VPS29. Here, we show an efficient and simple
method for isolating vacuolar sorting mutants of Arabidopsis. The method was based on two findings in this study. First,
VSR1 functioned as a sorting receptor for b-conglycinin by recognizing the vacuolar targeting signal. Second, when green
fluorescent protein (GFP) fusion with the signal (GFP-CT24) was expressed in vsr1, mag1/vps29, and wild-type seeds, both
vsr1and mag1/vps29 gave strongly fluorescent seeds but the wild type did not, suggesting that a defect in vacuolar sorting
provided fluorescent seeds by the secretion of GFP-CT24 out of the cells. We mutagenized transformant seeds expressing
GFP-CT24. From ;3,000,000 lines of M2 seeds, we obtained >100 fluorescent seeds and designated them green fluorescent
seed (gfs) mutants. We report 10 gfs mutants, all of which caused missorting of storage proteins. We mapped gfs1 to VSR1,
gfs2 to KAM2/GRV2, gfs10 to the At4g35870 gene encoding a novel membrane protein, and the others to different loci. This
method should provide valuable insights into the complex molecular mechanisms underlying vacuolar sorting of storage
proteins.
INTRODUCTION
Higher plants synthesize large amounts of storage proteins that
account for almost 80% of the total protein in seeds during seed
maturation. In maturing seeds, most storage proteins are syn-
thesized in the endoplasmic reticulum (ER) as precursors, which
are transported into protein storage vacuoles (PSVs), where they
are converted into mature forms (Hara-Nishimura et al., 1993;
Hara-Nishimura and Maeshima, 2000; Shimada et al., 2003b).
Multiple pathways are used for the transport of storage proteins
into PSVs (Vitale and Hinz, 2005). One of the pathways is an
aggregation sorting that is mediated by precursor-accumulating
(PAC) vesicles that are involved in the mass transport of storage
proteins to PSVs (Hara-Nishimura et al., 2004). A type I mem-
brane protein, PV72, was found on the membrane of the PAC
(Shimada et al., 1997, 2002). Arabidopsis thaliana has seven
homologs of PV72, which we designated At VSR1 to At VSR7 (for
vacuolar sorting receptor of Arabidopsis) (Shimada et al., 2003a).
The homolog most closely related to PV72 was designated
At VSR1, which is also referred to as ATELP (Ahmed et al., 1997).
We demonstrated that VSR1 functions as a receptor to sort seed
storage proteins to PSVs (Shimada et al., 2003a). vsr1 missorts
storage proteins by secreting them from cells, resulting in an
enlarged and electron-dense extracellular space in the seeds.
The vsr1 seeds abnormally accumulate the precursors of two
major storage proteins, 12S globulin and 2S albumin. These
findings show a receptor-mediated transport of seed storage
proteins to PSVs in higher plants.
Sorting of vacuolar proteins requires the interaction between a
vacuolar targeting signal and a sorting receptor that recognizes
the signal (Kirsch et al., 1994, 1996). Despite many articles that
discuss the interactions between VSR and a vacuolar targeting
signal (Kirsch et al., 1994; Sanderfoot et al., 1998; Ahmed et al.,
2000; Laval et al., 2003), there have been only a few in vivo
demonstrations of the interaction between VSR and the de-
termined vacuolar targeting signals: between VSR1 and an
N-terminal propeptide of either aleurain or sporamin (Ahmed
et al., 2000; Watanabe et al., 2004) and between pea (Pisum
sativum) BP-80 and the N-terminal propeptide of sporamin
(daSilva et al., 2005). Vacuolar targeting signals for PSVs have
been found in soybean (Glycine max) b-conglycinin (Nishizawa
et al., 2003, 2004), common bean (Phaseolus vulgaris) phaseolin
(Frigerio et al., 1998, 2001b), 2S albumin of Brazil nut (Saalbach
et al., 1996), barley (Hordeum vulgare) lectin (Muntz, 1998), and
castor bean (Ricinus communis) ricin (Frigerio et al., 2001a).
These signals exhibit no consensus amino acid sequence. It has
been shown that VSR binds to 2S albumin signals (Shimada
et al., 2003a; Jolliffe et al., 2004) and the ricin signal (Jolliffe et al.,
2004). The question raised is whether VSR binds to these
1 To whom correspondence should be addressed. E-mail [email protected]; fax 81-75-753-4142.The author responsible for distribution of materials integral to thefindings presented in this article in accordance with the policy describedin the Instructions for Authors (www.plantcell.org) is: Ikuko Hara-Nishimura ([email protected]).W Online version contains Web-only data.OA Open Access articles can be viewed online without a subscription.www.plantcell.org/cgi/doi/10.1105/tpc.106.045997
The Plant Cell, Vol. 19: 597–609, February 2007, www.plantcell.org ª 2007 American Society of Plant Biologists
different signals. It was also reported that At RMR1 (for Arabi-
dopsis receptor homology region transmembrane domain ring
H2 motif protein1) interacts with the C-terminal sorting domain of
phaseolin, which also belongs to the 7S globulin family and is a
homolog of b-conglycinin (Park et al., 2005). A reverse genetic
analysis with an RMR1-deficient mutant is necessary to demon-
strate that RMR1 functions as a sorting receptor for 7S globulins.
Both VSR1 and PV72 bind to peptides derived from storage
proteins, including 2S albumin (Watanabe et al., 2002) and 12S
globulin (Shimada et al., 2003a), in a Ca2þ-dependent manner.
Both VSRs have a binding ability not only to the storage protein
peptides but also to a vacuolar targeting signal for lytic vacuoles
(Watanabe et al., 2002; Shimada et al., 2003a). Overexpression
of the ER-localized soluble form of a seed-specific VSR (PV72)
interfered with the transport of a lytic vacuolar protein (ALEU) in
Arabidopsis leaves (Watanabe et al., 2004). This means that both
VSR for PSVs and VSR for lytic vacuoles share a similar mech-
anism for recognizing a vacuolar targeting signal.
Although VSR is unique to plants, VSR has a short cytosolic tail
containing the sequence YMPL, which is similar to Tyr-based
motifs (YXXF, in which X represents any amino acid and F
represents a hydrophobic residue with a bulky side chain) of
yeast VPS10 (Cooper and Stevens, 1996) and mammalian man-
nose 6-phosphate receptor (Honing et al., 1997). The Tyr-based
motif is recognized by the adapter complex of clathrin-coated
vesicles in yeast (Deloche et al., 2001) and animals (Klumperman
et al., 1998). Recently, the Tyr-based motif of BP-80 was
reported to be required for the transport of vacuolar proteins
(daSilva et al., 2006).
We recently isolated a mutant (maigo1 [mag1]) that missorts
storage proteins by secreting them out of the seed cells and
identified the responsible gene as At3g47810, which encodes a
homolog of the yeast retromer component VPS29 (Shimada
et al., 2006). MAG1/VPS29 might be involved in recycling VSR1
for the efficient sorting of seed storage proteins. VSR1 and
MAG1/VPS29 are the only factors that have been demonstrated
in planta to function in vacuolar sorting of seed storage proteins,
although VACUOLELESS (Rojo et al., 2001) and DRP2A (ADL6)
(Jin et al., 2001) were reported to be involved in trafficking to lytic
vacuoles. Forward genetics should be a useful approach to fully
clarify the sorting machinery for PSVs. We have isolated vacuolar
sorting mutants (mag) that accumulate storage protein precur-
sors in seeds (Shimada et al., 2006; Li et al., 2006). In this study,
we established a more efficient and easier method (green fluo-
rescent seed [gfs] method) of isolating vacuolar sorting mutants
and identified factors for vacuolar sorting of storage proteins.
RESULTS
VSR1 Is Involved in Vacuolar Sorting of the Exogenously
Expressed a9-Subunit of b-Conglycinin by Recognizing the
C-Terminal Vacuolar Targeting Signal in Arabidopsis Seeds
Two vacuolar targeting signals have been identified in 7S glob-
ulins. One is the 10–amino acid C-terminal peptide (CT10; Figure
1A) of the a9-subunit of b-conglycinin of soybean (Nishizawa
et al., 2003), and the other is the 4–amino acid C-terminal peptide
of phaseolin of common bean (Frigerio et al., 1998). In Arabi-
dopsis seeds, CT10 functions as a necessary and sufficient
vacuolar targeting signal (Nishizawa et al., 2003). To determine
whether VSR binds to the CT10 of b-conglycinin, we applied
microsomal proteins from Arabidopsis siliques, which contained
VSR1, to an affinity column conjugated with CT10. VSR1 bound
to CT10 in the presence of 1 mM CaCl2 and was eluted with CT10
or EGTA (Figure 1B), indicating that VSR1 binds to the vacuolar
targeting signal of b-conglycinin in a Ca2þ-dependent manner.
This means that VSR1 binds to the targeting signal of
b-conglycinin, as it binds to the signal of the storage protein 2S
albumin (Shimada et al., 2003a), even though the targeting
signals of these storage proteins have no consensus amino
acid sequences.
To determine whether VSR is required for vacuolar sorting of
b-conglycinin in Arabidopsis seeds, we expressed b-conglycinin
exogenously under the control of a seed-specific promoter in the
wild type (WT; Columbia) to produce the transformed plant
WT/conglycinin and in two vsr1 mutants to produce the trans-
formed plants vsr1-1/conglycinin and vsr1-2/conglycinin. WT/
conglycinin seeds gave several bands on an immunoblot with
anti-b-conglycinin antibody (Figure 2A, lane 2), although non-
transformant seeds gave no signal on the blot (Figure 2A, lane 1).
b-Conglycinin, when expressed in Arabidopsis seeds, was
shown to be transported to PSVs and to be partially degraded
to give several bands on the immunoblot, probably because it is
less stable in Arabidopsis PSVs than in soybean PSVs (Nishizawa
Figure 1. VSR1 Can Bind a Vacuolar Targeting Signal of a Soybean
Seed Storage Protein (b-Conglycinin) in a Calcium-Dependent Manner.
(A) Scheme of the preproprotein precursor of the soybean b-conglycinin
a9-subunit, which contains a signal peptide (SP) at the N terminus and a
vacuolar targeting signal composed of 10 amino acids (CT10) at the C
terminus.
(B) Affinity chromatography with a CT10-conjugated column. Micro-
somes were prepared from Arabidopsis siliques including maturing
seeds and were applied to the affinity column in the presence of 1 mM
CaCl2. The bound VSR1 was eluted competitively with the CT10 peptide
and was also eluted with EGTA. Total microsomes, the flow-through
fraction, the wash fraction, and the eluted fractions (fractions 1, 3, 5, and
7) were subjected to immunoblot analysis with anti-VSR1 antibody.
Molecular masses are given at left.
598 The Plant Cell
et al., 2003). The level of the nonprocessed form of b-conglycinin
was higher in both vsr1-1 and vsr1-2 seeds than in wild-type
seeds, although the protein profiles of b-conglycinin in the
mutants and the wild type were similar (Figure 2A).
The extracellular space of the vsr1-1/conglycinin seeds was
abnormally enlarged and filled with electron-dense material
(Figure 2C). To determine the localization of b-conglycinin, we
performed immunocytochemistry with an anti-b-conglycinin an-
tibody that does not give any signal on nontransformant seeds
(Nishizawa et al., 2003). b-Conglycinin was distributed in the
electron-dense region of the extracellular space in the vsr1-1/
conglycinin seeds (Figure 2E), whereas b-conglycinin was local-
ized in the aggregates within PSVs of WT/conglycinin seeds
(Figure 2D). We previously showed that vsr1 mutants missort
the storage proteins 12S globulin and 2S albumin by secreting
them out of seed cells (Shimada et al., 2003a). Our findings
revealed that b-conglycinin, like authentic storage proteins, is
abnormally secreted in vsr1 seeds, indicating that VSR is in-
volved in the vacuolar sorting of b-conglycinin in maturing seeds
of Arabidopsis.
vsr1 Generates Green Fluorescent Seeds by Secreting the
Green Fluorescent Protein Fusion Protein with the Vacuolar
Targeting Signal CT24 of b-Conglycinin
Previously, we generated an Arabidopsis transformant (WT/
GFP-CT24) that expressed SP-GFP-CT24, which was com-
posed of a signal peptide and green fluorescent protein (GFP)
followed by the b-conglycinin C-terminal 24 amino acids (CT24),
including CT10, under the control of a seed-specific promoter
(Nishizawa et al., 2003). WT/GFP-CT24 gave GFP fluorescence
that was completely merged with the red autofluorescence of
PSVs in seeds, showing that GFP-CT24 was transported into
PSVs (Figure 3A).
We generated vsr1-1/GFP-CT24 and vsr1-2/GFP-CT24 by
crossing each of two vsr1 mutants (vsr1-1 and vsr1-2) with WT/
GFP-CT24, respectively. Figure 3B shows that vsr1-1/GFP-CT24
gave a much higher intensity of GFP fluorescence in the extra-
cellular space than WT/GFP-CT24, suggesting that most of the
GFP-CT24 molecules were secreted out of the vsr1 seed cells.
Immunocytochemistry revealed that GFP-CT24 (Figures 3C and
3D) and 12S globulin of the authentic storage proteins (Figure 3E)
were detected in the electron-dense and enlarged extracellular
space of vsr1-1 seeds. vsr1-1 missorted GFP-CT24 by secreting
it out of the seed cells together with the authentic storage
proteins. vsr1-2 gave the same result as vsr1-1 (data not shown).
These results indicate that VSR1 sorts and delivers GFP-CT24 to
PSVs by recognizing CT24 of b-conglycinin. Considering that
VSR1 binds to CT10 of b-conglycinin in vitro (Figure 1B), CT10
might be sufficient to be recognized by VSR1.
Figure 4A shows an immunoblot of the transformant seeds
with anti-GFP antibody to evaluate the amount of GFP. The
seeds gave two bands on the blots, which might correspond to
28-kD GFP and 30-kD GFP-CT24. We found that both vsr1-1/
GFP-CT24 and vsr1-2/GFP-CT24 accumulated much higher
amounts of GFP in their seeds than WT/CT24 (Figure 4A). It is
possible that GFP-CT24 is more stable in the extracellular space
than in PSVs of Arabidopsis seeds. Interestingly, the high accu-
mulation of GFP in the extracellular space of vsr1 caused the
seeds to be fluorescent. Figures 4D and 4E show that all of the
seeds of both vsr1-1/GFP-CT24 and vsr1-2/GFP-CT24 exhibited
strong GFP fluorescence. On the contrary, the seeds of WT/GFP-
CT24, in which GFP was accumulated in PSVs, exhibited ex-
tremely low levels of GFP fluorescence (Figure 4C).
Figure 2. b-Conglycinin Is Missorted by Secreting Out of the Cells When
It Is Expressed in vsr1 Seeds.
(A) b-Conglycinin was expressed in both the wild type (Columbia) and
vsr1 mutants under the control of a seed-specific promoter. Seeds of
nontransformant (lane 1), WT/conglycinin (lane 2), vsr1-1/conglycinin
(lane 3), and vsr1-2/conglycinin (lane 4) plants were subjected to immu-
noblot analysis with either anti-VSR1 antibody or anti-b-conglycinin
antibody. Molecular masses are given at left.
(B) and (C) Ultrastructures of the seeds of both WT/conglycinin (B) and
vsr1-1/conglycinin (C) plants. The extracellular space of the transformed
vsr1-1 seeds was abnormally enlarged and filled with electron-dense
material. Arrowheads indicate the extracellular spaces. Bars ¼ 5 mm.
(D) and (E) Immunoelectron micrographs of WT/conglycinin (D) and
vsr1-1/conglycinin (E) seeds with anti-b-conglycinin antibody. Exoge-
nously expressed b-conglycinin was localized in the electron-dense
extracellular space (ECS) of vsr1-1/conglycinin seeds (white arrowheads
in [E]), whereas it was localized in PSVs of WT/conglycinin seeds (white
arrowheads in [D]). Bars ¼ 1 mm.
Vacuolar Sorting Mutant gfs 599
An Easy Method for Isolating Vacuolar
Sorting–Deficient Mutants (gfs)
We recently identified another factor, MAG1, for storage protein
sorting to PSVs. We found that the mag1 mutant missorts
storage proteins by secreting them out of the Arabidopsis seed
cells and that MAG1 is a homolog of the yeast retromer com-
ponent VPS29 (Shimada et al., 2006). To determine whether
GFP-CT24 was secreted in mag1/vps29 seeds, we crossed
each of two mag1/vps29 alleles (mag1-1/vps29-1 and mag1-2/
vps29-2) with WT/GFP-CT24 to produce mag1-1/vps29-1/GFP-
CT24 and mag1-2/vps29-2/GFP-CT24, respectively. Figure 5A
shows that some seeds of the F2 population fluoresced green.
Highly fluorescent seeds of mag1-1/vps29-1/GFP-CT24 and
Figure 3. A GFP Fusion Protein with a Vacuolar Targeting Signal Is
Secreted out of the Cells When Expressed in vsr1 Seeds but Is Localized
in PSVs When Expressed in Wild-Type Seeds.
(A) and (B) GFP-CT24 was expressed in the wild type (WT/GFP-CT24;
[A]) and the vsr1-1 mutant (vsr1-1/GFP-CT24; [B]). The seeds were
inspected with a confocal laser scanning microscope to obtain GFP
fluorescent images showing the localization of GFP-CT24 (GFP) and
autofluorescent image of PSVs. Both images were then merged. GFP
fluorescence was detected in PSVs of WT/GFP-CT24 seeds (A), whereas
a higher intensity of GFP fluorescence was detected in the extracellular
space of vsr1-1/GFP-CT24 seeds (B). Bars ¼ 5 mm.
(C) Immunoelectron micrograph of vsr1-1/GFP-CT24 seeds with anti-
GFP antibody, showing that the GFP fusion was localized in the extra-
cellular space (ECS; arrowheads). Bar ¼ 2 mm.
(D) Enlarged image of an immunoelectron micrograph of vsr1-1/GFP-
CT24 seeds with anti-GFP antibody, showing the distribution of the GFP
fusion in the extracellular space (ECS; arrowhead). Bar ¼ 1 mm.
(E) Immunoelectron micrograph of vsr1-1/GFP-CT24 seeds with anti-
12S globulin antibody, showing the secretion of 12S globulin in the vsr1-1
seeds (ECS; arrowhead). Bar ¼ 1 mm.
Figure 4. vsr1 Mutants That Express GFP-CT24 Produce Green Fluo-
rescent Seeds by Accumulating a Much Greater Amount of GFP-CT24 in
Seeds Than Do Wild-Type Plants That Express GFP-CT24.
(A) Expression levels of GFP-CT24 in the transformed seeds. Seeds of
nontransformant (lane 1), WT/GFP-CT24 (lane 2), vsr1-1/GFP-CT24 (lane
3), and vsr1-2/GFP-CT24 (lane 4) plants were subjected to immunoblot
analysis with either anti-VSR1 antibody or anti-GFP antibody. Molecular
masses are given at left.
(B) to (F) Seeds of wild-type (B), WT/GFP-CT24 (C), vsr1-1/GFP-CT24
(D), and vsr1-2/GFP-CT24 (E) plants were inspected with a binocular
microscope (top panels) and a fluorescence microscope (bottom
panels). For comparison of the intensities of fluorescence, the photo-
graphs were taken under the same conditions. Both vsr1-1/GFP-CT24
and vsr1-2/GFP-CT24 produced green fluorescent seeds.
600 The Plant Cell
mag1-2/vps29-2/GFP-CT24 exhibited GFP fluorescence in the
extracellular space (Figure 5B), as did vsr1/GFP-CT24, suggest-
ing that most of the GFP-CT24 molecules were secreted out of
mag1/vps29 seed cells. We selected highly green fluorescent
seeds (Figure 5C, lanes 3 to 5 and 9 to 11) and faintly fluorescent
seeds (Figure 5C, lanes 6 to 8 and 12 to 14) from the F2
population and subjected them to immunoblot analysis with
either anti-GFP antibody or anti-12S globulin antibody. The
intensity of GFP fluorescence of the seeds correlated not only
with the level of GFP accumulation but also with 12S globulin
precursor accumulation (Figure 5C). This finding suggested that
the intensity of green fluorescence of the seeds is a useful marker
for isolating vacuolar sorting mutant seeds.
Only two factors, VSR1 and MAG1/VPS29, have been shown
in planta to function in vacuolar sorting of storage proteins to
PSVs. The results in Figures 4 and 5 show that both vsr1 and
mag1/vps29 secreted a vacuole-targeted GFP, resulting in green
fluorescent seeds. This finding suggests that vacuolar sorting–
deficient mutants would produce green fluorescent seeds if
vacuole-targeted GFP was expressed in seeds. This provided an
idea for an efficient method for screening vacuolar sorting–
deficient mutants. Namely, a mutant seed can be easily identified
Figure 5. mag1/vps29 Mutants That Express GFP-CT24 Also Produce Green Fluorescent Seeds.
(A) mag1-1/vps29-1 was crossed with WT/GFP-CT24. The seeds of the F2 population were inspected with a binocular microscope (top panel) and a
fluorescence microscope (bottom panel). mag1/vps29/GFP-CT24 produced green fluorescent seeds (arrowheads).
(B) Highly fluorescent seeds of mag1-1/vps29-1/GFP-CT24 and mag1-2/vps29-2/GFP-CT24 were inspected with a confocal laser scanning
microscope. GFP fluorescence was detected out of the seed cells, showing missorting of GFP-CT24. The GFP fluorescence image was not merged
with the autofluorescent image (red) of PSVs. Each image of green and red fluorescence is shown in Supplemental Figure 1 online. Bars ¼ 10 mm.
(C) Selected highly green fluorescent seeds (lanes 3 to 5 and 9 to 11) and faintly fluorescent seeds (lanes 6 to 8 and 12 to 14) from the F2 population were
subjected to immunoblot analysis with anti-GFP and anti-12S globulin antibodies. The intensity of the GFP fluorescence of seeds correlated not only
with the accumulation levels of GFP but also with the accumulation of the 12S globulin precursor. p12S, 12S globulin precursor; 12S-a, a-subunit of
mature 12S globulin. Molecular masses are given at left.
Vacuolar Sorting Mutant gfs 601
by the strong green fluorescence that is caused by missorting of
GFP-CT24.
Based on this idea, we mutagenized the seeds of WT/GFP-
CT24 with ethyl methanesulfonate to obtain M2 seed populations.
Green fluorescent seeds were easily selected from a thousand of
the mutagenized M2 seeds with a fluorescence stereomicro-
scope, as shown in Figure 6A. We obtained >100 mutant lines
from ;3,000,000 lines of M2 seeds and designated them gfs
mutants. In this study, we describe 10 of these gfs mutants.
Four gfs1 Mutants Are vsr1 Alleles
Four of the mutants were named gfs1 because they were allelic
(discussed below). Figure 6B shows confocal laser scanning
microscopic images of seeds of three of them (gfs1-1, gfs1-2,
and gfs1-3). As expected, GFP fluorescence was observed out of
the cells in the seeds of these gfs mutants. An immunoblot
revealed that storage protein precursors, pro12S globulin and
pro2S albumin, were abnormally accumulated in the seeds of
these four allelic lines (gfs1-1 to gfs1-4), as in vsr1 seeds (Figure
6C). The different accumulations of pro2S albumin might be
attributable to different growing conditions of the plants and/or
different mutation sites. This result suggested that gfs1 mutants
have a defect in sorting of the storage proteins into PSVs.
The gene responsible for gfs1 mutants was roughly mapped to
a region close to the VSR1 gene (At3g52850). Immunoblot
analysis showed that VSR1 was not accumulated in the gfs1
seeds, as in the vsr1 seeds (Figure 7A). The faint signal on the blot
of these mutants might be attributable to the cross-reactivity of
the antibody with other homologs (VSR2 to VSR7). These results
suggested that the gfs1 mutants were allelic to vsr1. For a
complementation test, gfs1-1 was crossed with vsr1-2. Fourteen
F2 seeds were subjected to immunoblotting with anti-12S glob-
ulin antibody (Figure 7B). All of the F2 seeds accumulated 12S
globulin precursor, as did vsr1-2 and gfs1-1, suggesting that the
gfs1-1 mutant has a defect in VSR1. A complementation test of
gfs1-2 gave the same result (data not shown). Finally, we found
that each of the gfs1 mutants had a nucleotide substitution in the
VSR1 gene (At3g52850), as shown in Figure 7C. gfs1-1 had a
nucleotide substitution from GGA (Gly-214) to GAA (Glu). This
result suggests that the Gly-214 of gfs1-1 is essential for VSR1
function. gfs1-2 had a nucleotide substitution from CTTGga to
CTTGaa at the splicing donor site of the second intron. Both
gfs1-3 and gfs1-4 had the nucleotide substitutions from TGG
(Trp-105) to TGA (stop) and from TGG (Trp-297) to TGA (stop),
respectively. These results show that selecting green fluorescent
seeds among a large number of the seeds succeeded in isolating
vsr1 alleles.
Mapping of Other Mutations to Different Loci
We also analyzed six gfs mutants (gfs2, gfs3, gfs4, gfs5, gfs9, and
gfs10). All of them exhibited GFP fluorescence in the extracellular
matrix (Figure 8A), suggesting that they secrete vacuole-targeted
GFP out of the seed cells. The accumulation levels of GFP in
seeds of all of these mutants (data not shown) were similar to
those of vsr1/gfs1 (Figure 4A) and vps29/GFP-CT24 (Figure 5C).
An immunoblot revealed that these mutants abnormally
Figure 6. Selected gfs Mutants Have a Defect in Vacuolar Sorting of
Storage Proteins into PSVs.
(A) Seeds of the transformant (WT/GFP-CT24) were mutagenized with
ethyl methanesulfonate. The M2 seed populations were inspected with a
light microscope (left panel) and a fluorescence microscope (right panel)
to select green fluorescent seeds. Finally, ;100 fluorescent M2 lines
were isolated from 3,000,000 M2 seeds and were designated gfs
mutants.
(B) Seeds of the selected mutant lines were inspected with a confocal
laser scanning microscope. GFP fluorescence was detected out of the
seed cells of all gfs1 (gfs1-1, gfs1-2, and gfs1-3) mutants, showing
missorting of GFP-CT24. The GFP fluorescence image was not merged
with the autofluorescent image (red) of PSVs. Each image of green and
red fluorescence is shown in Supplemental Figure 1 online. Bars ¼10 mm.
(C) Seeds of wild-type (lane 1), vsr1-1 (lane 2), vsr1-2 (lane 3), gfs1-1 (lane
4), gfs1-2 (lane 5), and gfs1-3 (lane 6) plants were subjected to immu-
noblot analysis with anti-12S globulin and anti-2S albumin antibodies to
determine the behavior of endogenous storage proteins in gfs1 seeds. All
gfs1 mutants abnormally accumulated precursors of storage proteins
in seeds, as did the vsr1 mutant. p12S, 12S globulin precursors; 12S-a,
a-subunit of mature 12S globulin; p2S, 2S albumin precursors. Molecular
masses are given at left.
602 The Plant Cell
accumulated storage protein precursor of 12S globulin in their
seeds (Figure 8B), suggesting that they have a defect in vacuolar
sorting of storage proteins into PSVs. A rough map-based cloning
analysis mapped the gfs mutations to different loci (Figure 8C).
In gfs2, further mapping and DNA sequencing revealed a
single base pair mutation from G to A in the At2g26890 gene of
gfs2-1, which might cause a nonsense mutation from TGG (Trp-
33) to TGA (stop codon) (Figure 9A). To confirm that At2g26890 is
the gene responsible for the phenotype conferred by gfs2, we
obtained a knockout mutant of the At2g26890 gene
(SALK_132563). The mutant, which had T-DNA inserted into an
exon of the At2g26890 gene, was named gfs2-4 (Figure 9A).
gfs2-4 accumulated precursors of storage proteins, both 12S
globulin and 2S albumin (data not shown). We crossed gfs2-4
with WT/GFP-CT24 to generate gfs2-4/GFP-CT24. Some of the
F2 seeds were fluoresced (Figure 9B), and gfs2-4/GFP-CT24
exhibited GFP fluorescence in the extracellular space (Figure
9C). This result indicates that the mutation in the At2g26890 gene
is responsible for the phenotype conferred by gfs2.
Figure 7. Four gfs1 Mutant Lines Have a Mutation in the VSR1 Gene.
(A) Seeds of wild-type, vsr1-1, vsr1-2, gfs1-1, gfs1-2, and gfs1-3 plants
were subjected to immunoblot analysis with anti-VSR1 antibody. Faint
signals detected on the blot of vsr1 indicate cross-reactivity of the anti-
VSR1 antibody with the other VSR homologs in seeds.
(B) We crossed gfs1-1 with vsr1-2 and obtained F2 lines. Seeds of 14 F2
lines (one grain per lane) were subjected to immunoblot analysis with
anti-12S globulin antibody. All of the F2 seeds abnormally accumulated
12S globulin precursor, as did vsr1-2 and gfs1-1.
(C) Map-based cloning reveled that four gfs1 mutants had a mutation in
the VSR1 gene. Each point mutation site of gfs1-1 to gfs1-4 is shown in a
scheme of the VSR1 gene. Closed boxes indicate coding regions, and
solid lines indicate introns. The T-DNA insertion sites of vsr1-1 and vsr1-2
are also shown.
Figure 8. Six Different Mutants (gfs2, gfs3, gfs4, gfs5, gfs9, and gfs10)
Other Than vsr1/gfs1 Alleles Were Isolated.
(A) Seeds of six gfs mutants (gfs2, gfs3, gfs4, gfs5, gfs9, and gfs10) were
inspected with a confocal laser scanning microscope. GFP fluorescence
was detected out of the cells of these mutant seeds, showing missorting
of GFP-CT24. The GFP fluorescence image was not merged with the
autofluorescent image (red) of PSVs. Each image of green and red
fluorescence is shown in Supplemental Figure 1 online. Bars ¼ 10 mm.
(B) Seeds of wild-type, gfs2, gfs3, gfs4, gfs5, gfs6, gfs9, and gfs10 plants
were subjected to immunoblot analysis with anti-12S globulin antibody
to determine the behavior of endogenous seed storage proteins in these
seeds. These gfs mutants abnormally accumulated 12S globulin in
precursors in seeds. p12S, 12S globulin precursors; 12S-a, a-subunit
of mature 12S globulin. Molecular masses are given at left.
(C) Mutations of gfs1, gfs2, gfs3, gfs4, gfs5, gfs9, and gfs10 were
mapped to different loci from each other: gfs1 to At3g52850 (VSR1), gfs2
to At2g26890 (KAM2/GRV2), and gfs10 to At4g35870.
Vacuolar Sorting Mutant gfs 603
The At2g26890 gene was reported to be responsible for the
gravitropism defective2 (grv2) mutant, which is deficient in shoot
gravitropism and phototropism (Silady et al., 2004). We recently
found that the At2g26890 gene is also responsible for an
Arabidopsis mutant, katamari2 (kam2), that had deformed endo-
somes involving other endosomal compartments and formed
abnormally large aggregates with various organelles in the cells
(Tamura et al., 2007). kam2 alleles missorted seed storage
proteins by secreting them from cells (Tamura et al., 2007),
indicating that GFS2/KAM2 is involved in the vacuolar sorting of
storage proteins into PSVs. The GFS2/KAM2 gene encodes a
homolog of a DnaJ domain–containing RME8 (for receptor-
mediated endocytosis8) that is required for endocytosis and the
organization of endosomes in Caenorhabditis elegans (Zhang
et al., 2001), Drosophila melanogaster (Chang et al., 2004), and
human (Girard et al., 2005). KAM2 protein localized on punctate
structures, which did not merge with any markers for the Golgi,
the trans-Golgi network, endosomes, or prevacuolar compart-
ments, suggesting that GFS2/KAM2 is localized in an unchar-
acterized endosomal compartment (Tamura et al., 2007). It is
possible that DnaJ domain–containing GFS2/KAM2/GRV2 is
involved in the folding of protein(s) needed for the proper forma-
tion of endosomes that mediate protein trafficking to PSVs and
the biogenesis of endosomal compartments, including PSVs.
In gfs10, fine-scale mapping and DNA sequencing revealed a
single base pair mutation from G to A in the At4g35870 gene of
gfs10-1, which might cause a nonsense mutation from TGG (Trp-
397) to TGA (stop codon) (Figure 10A). We obtained a T-DNA–
tagged knockout mutant of the At4g35870 gene (SALK_139226)
and named it gfs10-2 (Figure 10A). An immunoblot showed that
gfs10-2 accumulated 12S globulin precursors in seeds, as did
gfs10-1 (Figure 10C). We crossed gfs10-2 with gfs10-1 that
expressed GFP-CT24 to generate F1 seeds. The expression of
GFP in the F1 seeds indicated that the two lines were success-
fully crossed. Each F1 seed grain accumulated 12S globulin
precursor as did both parent seeds (Figure 10D) and exhibited
GFP fluorescence in the extracellular matrix of the seed cells
(Figure 10E). These results indicate that the mutation in the
At4g35870 gene is responsible for the phenotype conferred by
gfs10. The ultrastructure of gfs10-1 seeds revealed that the
extracellular space was enlarged and filled with electron-dense
materials, possibly as a result of the secretion of storage proteins
(Figure 10F). The phenotype is very similar to those of the
vacuolar sorting mutants vsr1 (Shimada et al., 2003a), mag1/
vps29 (Shimada et al., 2006), and gfs2/kam2 (Tamura et al.,
2007). The At4g35870 gene, which is an intronless gene, en-
codes a novel membrane protein of unknown function that has
10 putative transmembrane domains (Figure 10B). The protein is
specific to higher plants. Thus, plants appear to have a novel and
plant-specific mechanism for the vacuolar sorting of storage
proteins. Further identification of the responsible genes of the
other gfs mutants should provide valuable insights into the
complex molecular mechanisms underlying the vacuolar sorting
of storage proteins.
DISCUSSION
A VSR Is Involved in Vacuolar Sorting of Storage Proteins
into PSVs by Recognizing Their Vacuolar Targeting Signals
It has long been thought that VSR is a sorting receptor for lytic
vacuoles but not a sorting receptor for PSVs (Hinz et al., 1999; Li
et al., 2002), although a VSR homolog (PV72) was isolated from
PAC vesicles of maturing pumpkin (Cucurbita maxima) seeds
(Shimada et al., 1997). However, we reported the involvement of
VSR1 in the proper sorting of storage protein precursors into PSV
cells (Shimada et al., 2003a). Additionally, in this study, we
isolated four vsr1 alleles by applying our green fluorescent seed
method (Figure 7) and found that all of the vsr1 alleles missorted
storage protein precursors in their seeds (Figure 6). It is now clear
that VSR1 is involved in the sorting of storage proteins into PSVs.
Arabidopsis has seven members (VSR1 to VSR7) of the VSR
family. Previously, we isolated seven mutants, each being defi-
cient in a different VSR gene, and found that only vsr1 accumu-
lated storage protein precursors in the seeds, whereas vsr2 to
vsr7 accumulated no precursors in the seeds (Shimada et al.,
2003a). By screening >3,000,000 seeds, we isolated four alleles
of vsr1 but no alleles of vsr2 to vsr7. These results suggest that
VSR1 functions primarily in vacuolar sorting of storage proteins in
Figure 9. A gfs2 Mutant Has a Mutation in the GRV2 Gene (At2g26890),
Which Encodes a Homolog of Mammalian RME8.
(A) A point mutation site of gfs2-1 is shown in a scheme of the GRV2
gene. Closed boxes indicate coding regions, and solid lines indicate
introns. A T-DNA–tagged allele (gfs2-4; SALK_132563) is also shown.
(B) We crossed the T-DNA–tagged mutant gfs2-4 with the transformant
WT/GFP-CT24. The F2 seed population was inspected with a binocular
microscope (top panel) and a fluorescence microscope (bottom panel).
gfs2-4/GFP-CT24 produced green fluorescent seeds (arrowheads).
(C) The gfs2-4/GFP-CT24 seeds were inspected with a confocal laser-
scanning microscope. GFP fluorescence was detected out of the cells of
these mutant seeds, showing missorting of GFP-CT24. The GFP fluo-
rescence image was not merged with the autofluorescent image (red) of
PSVs. Each image of green and red fluorescence is shown in Supple-
mental Figure 1 online. Bar ¼ 10 mm.
604 The Plant Cell
maturing seeds. This finding is consistent with VSR1 being the
most abundant of the seven VSRs in seeds.
Examination of several plants revealed >10 vacuolar targeting
signals for different vacuolar proteins, about half of which are
storage proteins of various plants (Vitale and Hinz, 2005). It is
possible that VSR functions in the sorting of storage proteins into
PSVs by recognizing the signals in various maturing seeds. This
notion is supported by the findings that VSR1 is responsible for
the sorting of 2S albumin and 12S globulin into PSVs of
Arabidopsis (Shimada et al., 2003a) and that VSR-like receptors
are involved in the sorting of 2S albumin and ricin into PSVs in
castor bean (Jolliffe et al., 2004).
What about the third major storage protein, 7S globulin? The
C-terminal peptide CT10 of b-conglycinin, which belongs to the
7S globulin family, was reported to act as a necessary and
sufficient signal for vacuolar targeting, because (1) deletion of the
signal from b-conglycinin resulted in secretion of the protein and
(2) addition of the signal to a GFP reporter resulted in delivery of
the GFP into PSVs (Nishizawa et al., 2003). Here, we demon-
strate that the vacuolar targeting signal was recognized by VSR1
for vacuolar sorting of b-conglycinin in Arabidopsis seeds.
As described above, in vivo and in vitro analyses have shown
that VSRs have an ability to bind to the vacuolar targeting signals
of various storage proteins. It remains to be determined how
VSRs can recognize a variety of signal sequences, because
these targeting signals have no consensus sequences.
Multiple Transport Pathways to PSVs
All of the gfs mutants that we isolated accumulated both pre-
cursor and mature forms of storage proteins (Figures 6 and 8), as
did vsr1 (Shimada et al., 2003a) and mag1/vps29 (Shimada et al.,
2006). We also found that another vacuolar sorting mutant,
mag2, accumulated both precursor and mature forms of storage
proteins in seeds. MAG2 might bind to ER-localized t-SNAREs
and function in the transport of storage protein precursors
between the ER and the Golgi complex in plants (Li et al.,
2006). These mutants normally deliver a portion of storage
proteins into PSVs, even though they have a defect in vacuolar
sorting of storage proteins: vsr1, mag1/vps29, gfs2/kam2/grv2,
and gfs10 secreted part of the storage proteins, whereas mag2
developed precursor-accumulating bodies within the seeds
(Li et al., 2006). The incomplete secretion of storage proteins
in these mutants might not be attributable to the redundancy
of each responsible gene, because MAG1/VPS29, MAG2,
GFS2/KAM2, and GFS10 do not have closely related homologs
in the Arabidopsis genome. Vacuolar sorting mutants that
accumulate all storage proteins as only precursor forms have
not been isolated. Such mutants may be lethal because they
could be affected in all mechanisms of vacuolar sorting.
Another possibility is that plants use multiple transport pathways
to PSVs.
One of the pathways is aggregation sorting, which is an
efficient way of transporting large amounts of protein. Two
compartments have been shown to mediate aggregation sorting:
ER-derived PAC vesicles in pumpkin (Hara-Nishimura et al.,
1993, 1998; Mitsuhashi et al., 2001), castor bean (Hiraiwa et al.,
1993), soybean (Mori et al., 2004), and rice (Oryza sativa)
Figure 10. A gfs10 Mutant Has a Mutation in the At4g35870 Gene,
Which Encodes a Novel Membrane Protein.
(A) A point mutation site of gfs10-1 is shown in a scheme of the
At4g35870 gene, which is an intronless gene. A T-DNA–tagged mutant
allele (gfs10-2; SALK_139226) is also shown.
(B) Scheme of GFS10 with putative 10 transmembrane domains (green
boxes).
(C) Seeds of gfs10 were subjected to immunoblot analysis with anti-12S
globulin. Two gfs10 alleles abnormally accumulated 12S globulin pre-
cursors. p12S, 12S globulin precursors; 12S-a, a-subunit of mature 12S
globulin. Molecular masses are given at left.
(D) We crossed gfs10-2 with gfs10-1 that expressed GFP-CT24. Six F1
seeds was subjected to immunoblot analysis with either anti-12S glob-
ulin antibody or anti-GFP antibody. The accumulation of GFP in each F1
seed grain indicated that crossing was successful. Each F1 seed
accumulated 12S globulin precursors, as did parent seeds.
(E) An F1 seed was inspected with a confocal laser scanning micro-
scope. GFP fluorescence was detected out of the cells, showing
missorting of GFP-CT24. The GFP fluorescence image was not merged
with the autofluorescent image (red) of PSVs. Each image of green and
red fluorescence is shown in Supplemental Figure 1 online. Bar¼ 10 mm.
(F) Electron micrograph of a gfs1-1 seed. The extracellular space of the
gfs1-1 seeds was abnormally enlarged and filled with electron-dense
materials (arrowheads). Bar ¼ 5 mm.
Vacuolar Sorting Mutant gfs 605
(Takahashi et al., 2005) and Golgi-derived dense vesicles in pea
(Hinz et al., 1999) and common bean (Chrispeels, 1983). PAC-like
vesicles were induced in transgenic soybean seeds that cosup-
pressed the a-subunit of b-conglycinin (Kinney et al., 2001) and
in transgenic Arabidopsis leaves that expressed a fusion protein
composed of a small subunit of 2S albumin and phosphinothricin
acetyltransferase (Hayashi et al., 1999). PAC vesicles can be
regarded as prevacuolar compartments or endosomes in ma-
turing seeds of various plant species.
We proposed a hypothetical model of the multiple pathways
that is composed of three steps. Storage proteins that are
actively synthesized form aggregates within the ER. As a result,
aggregates and free molecules of storage proteins exist together
in the ER lumen. This means that a storage protein species takes
two forms: an aggregate form and a free molecule form. Aggre-
gates develop to generate the PAC vesicles (first step), whereas
free molecules leave the ER for the Golgi complex, where they are
trapped by a VSR to be recruited to PAC vesicles (second step).
PAC vesicles are incorporated directly into PSVs (third step).
VSR1 and MAG1/VPS29 are involved in the second step. MAG2
might function in the exit from the ER in the first or second step.
Considering a subcellular localization of GFS2/KAM2/GRV2
in uncharacterized compartments (Tamura et al., 2007), a novel
pathway should be involved in vacuolar sorting. Further charac-
terization and subcellular localization of GFS10 will provide a clue
to the complex mechanism of vacuolar sorting.
A Powerful Tool for the Screening of Mutants Deficient
in Vacuolar Sorting
To understand the molecular mechanism underlying the multiple
pathways, forward genetics might be a useful approach. In this
study, we developed a simple method (the green fluorescent
seed method) to isolate vacuolar sorting mutants using GFP
fluorescence as an indicator. Using this method, we succeeded
in isolating four vsr1 alleles. We also identified GFS2/KAM2/
GRV2 and GFS10 as factors of the sorting of storage proteins to
PSVs. This technique must be one of the easiest and most
efficient for hunting vacuolar sorting mutants.
We found that mag1/vps29 plants that expressed GFP-CT24
exhibited green fluorescent seeds. MAG1/VPS29 is encoded by
a single-copy gene and is expressed ubiquitously in various
organs, including leaves, roots, flowers, and developing seeds
(Schmid et al., 2005). Thus, the MAG1/VPS29 gene may play an
indispensable role in various tissues of Arabidopsis. MAG1/
VPS29 might be required for vacuolar protein sorting into not only
PSVs but also lytic vacuoles. gfs2/kam2/grv2 mutants exhibit a
defect in the endomembrane organization of vegetative tissues
(Tamura et al., 2007), and both gfs4 and gfs5 mutants exhibit a
defect in the development of vegetative tissues (data not shown).
Considering these results, vacuolar sorting mutants for both
PSVs and lytic vacuoles could be isolated using our method.
The green fluorescent seed method makes it possible to
efficiently screen a large number of seeds and to easily isolate
mutant seeds. Selecting 3,000,000 seeds gave us >100 gfs
mutants, some of which might be allelic. Identification of the
responsible genes of these gfs mutants should provide a valu-
able insight into the complex molecular mechanisms in the
multiple pathways unique to higher plants, as do yeast sec and
vps mutants (Bonangelino et al., 2002). Further analysis of these
mutants will reveal not only the vacuolar sorting mechanism of
PSVs but also that of lytic vacuoles.
METHODS
Plant Materials and Growth Conditions
We used wild-type plants of Arabidopsis thaliana (ecotypes Columbia and
Wassilewskijia-2) and the Arabidopsis mutants vsr1-1 and vsr1-2 (Shimada
et al., 2003a) and mag1-1/vps29-1 and mag1-2/vps2-2 (Shimada et al.,
2006). We also used transgenic plants that were generated previously
(Nishizawa et al., 2003): Arabidopsis (ecotype Columbia) that expressed
soybean (Glycine max) b-conglycinin a9-subunit (WT/conglycinin) and
Arabidopsis (ecotype Columbia) that expressed GFP-CT24 (WT/GFP-
CT24). We crossed vsr1-1 and vsr1-2 with WT/conglycinin and WT/GFP-
CT24 to generate the transgenic plants vsr1-1/conglycinin and vsr1-2/
conglycinin and the transgenic plants vsr1-1/GFP-CT24 and vsr1-2/GFP-
CT24, respectively. We crossed vsr1-2 with gfs1-1 for complementation
tests. We crossed mag1-1/vps29-1 and mag1-2/vps29-2 with WT/GFP-
CT24 to generate two transgenic plants (mag1-1/vps29-1/GFP-CT24 and
mag1-2/vps29-2/GFP-CT24, respectively). We obtained T-DNA insertion
mutants (SALK_132563 and SALK_139226) from the ABRC at Ohio State
University. We crossed the SALK_132563 line (gfs2-4) with the trans-
formant WT/GFP-CT24 to produce gfs2-4/GFP-CT24. We further
crossed gfs10-1 with the SALK_139226 line (gfs10-2) to produce F1
seeds. Surface-sterilized seeds were sown onto 0.5% Gellan Gum
(Wako) with Murashige and Skoog medium supplemented with 1%
sucrose and were grown at 228C under continuous light.
Affinity Column Chromatography
We chemically synthesized the C-terminal peptide (CT10; PLSSILRAFY)
of soybean b-conglycinin a9-subunit and immobilized the peptide (5 mg)
to Aminolink coupling gel (2 mL of a bed volume; Pierce). Affinity column
chromatography was performed essentially as described (Watanabe
et al., 2002). Microsomes were prepared from young Arabidopsis siliques
and were applied to the column that had been equilibrated with 20 mM
HEPES-NaOH, pH 7.0, 150 mM NaCl, 1% CHAPS, and 1 mM CaCl2. The
bound proteins were eluted with either the CT10 peptide or 5 mM EGTA.
Each fraction was subjected to immunoblotting with anti-VSR1 antibody.
Specific Antibodies and Immunoblot Analysis
Immunoblot analysis of Arabidopsis seeds was performed essentially as
described previously (Shimada et al., 2002), except that dilutions of the
antibodies were as follows: anti-VSR1 (1:5000) (Yamada et al., 2005),
anti-b-conglycinin (1:5000) (Nishizawa et al., 2003), anti-2S albumin
(1:10,000), anti-12S globulin a-subunit (1:10,000) (Shimada et al.,
2003b), and anti-GFP (1:5000) (Tamura et al., 2003).
Immunoelectron Microscopy
Dry seeds were cut into slices with a razor blade and then fixed with 4%
(w/v) paraformaldehyde, 1% (v/v) glutaraldehyde, 0.05 M cacodylate
buffer, pH 7.4, and 10% DMSO. Procedures for immunoelectron micro-
scopic analysis were essentially as described previously (Hara-Nishimura
et al., 1993). Ultrathin sections were treated with antibodies against Arabi-
dopsis 2S albumin (dilution, 1:50), Arabidopsis 12S globulin a-subunit
(1:50), b-conglycinin (1:50), and GFP (1:50). Sections were examined with
a transmission electron microscope (model 120WX; JEOL) at 80 kV.
606 The Plant Cell
Confocal Laser Scanning Microscopy
Dry seeds were pressed in glycerol between glass slides and cover strips
to push out cotyledons. The cotyledons were inspected with a confocal
laser scanning microscope (LSM510 META; Carl Zeiss) using the 488-nm
line of a 40-mW Ar/Kr laser and an emission filter (BP 530-505) for GFP
fluorescence and the 544-nm line of a 1-mW He/Ne laser and an emission
filter (LP 560) for autofluorescence of PSVs with a 40 3 0.75 numerical
aperture dry objective. All images were obtained in multitrack mode and
analyzed using LSM image-examiner software (Carl Zeiss). The data were
exported as eight-bit TIFF files and processed using Adobe Photoshop
5.5 (Adobe Systems).
Isolation of the gfs Mutants
Seeds of the transformed plant that expressed GFP-CT24 (WT/GFP-
CT24) were mutagenized by soaking them for 16 h in 0.2 or 0.25% (v/v)
methanesulfonic acid ethyl ester (Sigma-Aldrich) and then washed for 11
h in running water to obtain M1 seeds. The M1 seeds were grown after
self-fertilization to generate the M2 seed populations. These seeds were
inspected with a fluorescence stereomicroscope (MZ-APO; Leica Micro-
systems), and images were acquired using a digital camera (DS-5ML1;
Nikon).
Map-Based Cloning
Each homozygous mutant (background, ecotype Columbia) was crossed
with Landsberg erecta wild-type plants to generate a mapping popula-
tion. Genomic DNA was isolated from leaves of each mutant that was
selected from the F2 progeny. The polymorphism between Columbia and
Landsberg erecta was analyzed using a combination of cleaved amplified
polymorphic sequences and simple sequence length polymorphism
markers with data obtained from The Arabidopsis Information Resource
(http://www.Arabidopsis.org).
According to our green fluorescent seed method, the mutant seeds can
be quickly and easily selected from the F2 progeny. For rough mapping,
we selected ;10 to 20 recombinants of the F2 progeny of each gfs
mutant. As a result, all four gfs1 mutations were roughly mapped near the
VSR1 gene on chromosome 3 and the other mutations (gfs2, gfs3, gfs4,
gfs5, gfs9, and gfs10) were roughly mapped as shown in Figure 8C. For
identification of the responsible genes of gfs1 mutants, we subjected the
seeds of the four gfs mutants to immunoblot analysis with anti-VSR1
antibody to see no accumulation of VSR1, crossed each gfs1 mutant with
vsr1-2 to confirm the responsible gene to be VSR1, and sequenced the
VSR1 gene of each gfs1 mutant to determine the mutation site of the
gene, as shown in Figure 7. We further performed fine-scale mapping for
gfs2 and gfs10 mutants. We selected 120 mutants from the F2 progeny for
gfs2 and 100 mutants from the F2 progeny for gfs10 by the green
fluorescent seed method and prepared genomic DNAs from the leaves of
each mutant for fine-scale mapping. We sequenced the respective genes
to determine the mutation sites of gfs2-1 and gfs10-1. Finally, to deter-
mine the responsible genes of gfs2-1 and gfs10-1, we obtained T-DNA–
tagged lines that had a defect of each gene to examine the phenotype. For
gfs2, we had six mutant alleles, all of which exhibited the vacuolar sorting
defect (Tamura et al., 2007). For further confirmation of the responsible
gene of gfs10-1, we crossed gfs10-1 with gfs10-2, as shown in Figure 10D.
Nucleotide sequences were determined from both strands using an ABI
Prism Big Dye Terminator Cycle sequence reaction kit (Applied Biosystems)
and a DNA sequencer (Prism 3100; Applied Biosystems).
Accession Numbers
Arabidopsis Genome Initiative locus identifiers for the genes mentioned in
this article are as follows: GFS1/VSR1 (At3g52850), MAG1/VPS29
(At3g7810), GFS2/KAM2/GRV2 (At2g26890), and GFS10 (At4g35870).
Supplemental Data
The following material is available in the online version of this article.
Supplemental Figure 1. Green Fluorescent and Red Fluorescent
Images of Seeds Are Shown Separately for All Transformants in This
Study.
ACKNOWLEDGMENTS
This work was supported by the Core Research for Evolutional Science
and Technology program of the Japan Science and Technology Cor-
poration and by Grants-in-Aid for Scientific Research (Grants 16044224,
16085203, and 17107002) and for 21st Century Centers of Excellence
Research, Kyoto University (Grant A14), from the Ministry of Education,
Culture, Sports, Science, and Technology of Japan. We are grateful to
the ABRC at Ohio State University for providing the seeds of Arabi-
dopsis T-DNA insertion mutants (SALK_132563 and SALK_139226).
Received July 18, 2006; revised December 19, 2006; accepted January
22, 2007; published February 9, 2007.
REFERENCES
Ahmed, S.U., Bar-Peled, M., and Raikhel, N.V. (1997). Cloning and
subcellular location of an Arabidopsis receptor-like protein that shares
common features with protein sorting receptors of eukaryotic cells.
Plant Physiol. 114: 325–336.
Ahmed, S.U., Rojo, E., Kovalentina, V., Venkataraman, S.,
Dombrowski, J.E., Matsuoka, K., and Raikhel, N.V. (2000). The
plant vacuolar sorting receptor AtELP is involved in transport of NH2-
terminal propeptide-containing vacuolar proteins in Arabidopsis thali-
ana. J. Cell Biol. 149: 1335–1344.
Bonangelino, C.J., Chavevz, E.M., and Bonifacino, J.S. (2002). Ge-
nomic screen for vacuolar protein sorting genes in Saccharomyces
cerevisiae. Mol. Biol. Cell 13: 2486–2501.
Chang, H.C., Hull, M., and Mellman, I. (2004). The J-domain protein
Rme-8 interacts with Hsc70 to control clathrin-dependent endocyto-
sis in Drosophila. J. Cell Biol. 164: 1055–1064.
Chrispeels, M.J. (1983). The Golgi apparatus mediates the transport of
phytohemagglutinin to the protein bodies in bean cotyledons. Planta
158: 140–151.
Cooper, A.A., and Stevens, T.H. (1996). Vps10p cycles between
the late-Golgi and prevacuolar compartments in its function as the
sorting receptor for multiple yeast vacuolar hydrolases. J. Cell Biol.
133: 529–541.
daSilva, L.L.P., Foresti, O., and Denecke, J. (2006). Targeting of the
plant vacuolar sorting receptor BP80 is dependent on multiple sorting
signals in the cytosolic tail. Plant Cell 18: 1477–1497.
daSilva, L.L.P., Taylor, J.P., Hadlington, J.L., Hanton, S.L., Snowden,
C.J., Fox, S.J., Foresti, O., Brandizzi, F., and Denecke, J. (2005).
Receptor salvage from the prevacuolar compartment is essential for
efficient vacuolar protein targeting. Plant Cell 17: 132–148.
Deloche, O., Yeung, B.G., Payne, G.S., and Schekman, R. (2001).
Vps10p transport from the trans-Golgi network to the endosome is
mediated by clathrin-coated vesicles. Mol. Biol. Cell 12: 475–485.
Frigerio, L., de Virgilio, M., Prada, A., Faoro, F., and Vitale, A. (1998).
Sorting of phaseolin to the vacuole is saturable and requires a short
C-terminal peptide. Plant Cell 10: 1031–1042.
Frigerio, L., Jolliffe, N.A., Cola, A.D., Felipe, D.H., Paris, N., Neuhaus,
J.-M., Lord, J.M., Ceriotti, A., and Roberts, L.M. (2001a). The
Vacuolar Sorting Mutant gfs 607
internal propeptide of the ricin precursor carries a sequence-specific
determinant for vacuolar sorting. Plant Physiol. 126: 167–175.
Frigerio, L., Pastres, A., Prada, A., and Vitale, A. (2001b). Influence of
KDEL on the fate of trimeric or assembly-defective phaseolin: selec-
tive use of an alternative route to vacuoles. Plant Cell 13: 1109–1126.
Girard, M., Poupon, V., Blondeau, F., and McPherson, P.S. (2005).
The DnaJ-domain protein RME-8 functions in endosomal trafficking.
J. Biol. Chem. 280: 40135–40143.
Hara-Nishimura, I., and Maeshima, M. (2000). Vacuolar processing
enzymes and aquaporins. In Vacuolar Compartments in Plants,
A.D.G. Robinson and J.C. Rogers, eds (London: Sheffield Academic
Press), pp. 20–42.
Hara-Nishimura, I., Matsushima, R., Shimada, T., and Nishimura, M.
(2004). Diversity and functions of ER-derived compartments in plants:
Are these compartments specific to plant cells? Plant Physiol. 136:
3435–3439.
Hara-Nishimura, I., Shimada, T., Hatano, K., Takeuchi, Y., and
Nishimura, M. (1998). Transport of storage proteins to protein-
storage vacuoles is mediated by large precursor-accumulating ves-
icles. Plant Cell 10: 825–836.
Hara-Nishimura, I., Takeuchi, Y., Inoue, K., and Nishimura, M.
(1993). Vesicle transport and processing of the precursor to 2S
albumin in pumpkin. Plant J. 4: 793–800.
Hayashi, M., Toriyama, K., Kondo, M., Hara-Nishimura, I., and
Nishimura, M. (1999). Accumulation of a fusion protein containing
2S albumin induces novel vesicles in vegetative cells of Arabidopsis.
Plant Cell Physiol. 40: 263–272.
Hinz, G., Hillmer, S., Baumer, M., and Hohl, I.I. (1999). Vacuolar
storage proteins and the putative vacuolar sorting receptor BP-80 exit
the Golgi apparatus of developing pea cotyledons in different trans-
port vesicles. Plant Cell 11: 1509–1524.
Hiraiwa, N., Takeuchi, Y., Nishimura, M., and Hara-Nishimura, I.
(1993). A vacuolar processing enzyme in maturing and germinating
seeds: Its distribution and associated changes during development.
Plant Cell Physiol. 34: 1197–1204.
Honing, S., Sosa, M., Hille-Rehfeld, A., and von Figura, K. (1997). The
46-kDa mannose 6-phosphate receptor contains multiple binding
sites for clathrin adaptors. J. Biol. Chem. 272: 19884–19890.
Jin, J.B., Kim, Y.A., Kim, S.J., Lee, S.H., Kim, D.H., Cheong, G.W.,
and Hwang, I. (2001). A new dynamin-like protein, ADL6, is involved
in trafficking from the trans-Golgi network to the central vacuole in
Arabidopsis. Plant Cell 13: 1511–1526.
Jolliffe, N.A., Brown, J.C., Neumann, U., Vicre, M., Bachi, A., Hawes,
C., Ceriotti, A., Roberts, L.M., and Frigerio, L. (2004). Transport of
ricin and 2S albumin precursors to the storage vacuoles of Ricinus
communis endosperm involves the Golgi and VSR-like receptors.
Plant J. 39: 821–833.
Kinney, A.J., Jung, R., and Herman, E.M. (2001). Cosuppression of the
a-subunits of b-conglycinin in transgenic soybean seeds induces the
formation of endoplasmic reticulum-derived protein bodies. Plant Cell
13: 1165–1178.
Kirsch, T., Paris, N., Butler, J.M., Beevers, L., and Rogers, J.C.
(1994). Purification and initial characterization of a potential plant
vacuolar targeting receptor. Proc. Natl. Acad. Sci. USA 91: 3403–
3407.
Kirsch, T., Saalbach, G., Raikhel, N.V., and Beevers, L. (1996).
Interaction of a potential vacuolar targeting receptor with amino-
and carboxyl-terminal targeting determinants. Plant Physiol. 111:
469–474.
Klumperman, J., Kuliawat, R., Griffith, J.M., Geuze, H.J., and Arvan,
P. (1998). Mannose 6-phosphate receptors are sorted from immature
secretory granules via adaptor protein AP-1, clathrin, and syntaxin
6-positive vesicles. J. Cell Biol. 141: 359–371.
Laval, V., Masclaux, F., Serin, A., Carriere, M., Roldan, C., Devic, M.,
Pont-Lezica, R.F., and Galaud, J.P. (2003). Seed germination is
blocked in Arabidopsis putative vacuolar sorting receptor (atbp80)
antisense transformants. J. Exp. Bot. 54: 213–221.
Li, L., Shimada, T., Takahashi, H., Ueda, H., Fukao, Y., Kondo, M.,
Nishimura, M., and Hara-Nishimura, I. (2006). MAIGO2 is involved in
exit of seed storage proteins from the endoplasmic reticulum in
Arabidopsis thaliana. Plant Cell 18: 3535–3547.
Li, Y.B., Rogers, S.W., Tse, Y.C., Lo, S.W., Sun, S.S., Jauh, G.Y., and
Jiang, L. (2002). BP-80 and homologs are concentrated on post-
Golgi, probable lytic prevacuolar compartments. Plant Cell Physiol.
43: 726–742.
Mitsuhashi, N., Hayashi, Y., Koumoto, Y., Shimada, T., Fukasawa-
Akada, T., Nishimura, M., and Hara-Nishimura, I. (2001). A novel
membrane protein that is transported to protein-storage vacuoles via
precursor-accumulating vesicles. Plant Cell 13: 2361–2372.
Mori, T., Maruyama, N., Nishizawa, K., Higasa, T., Yagasaki, K.,
Ishimoto, M., and Utsumi, S. (2004). The composition of newly syn-
thesized proteins in the endoplasmic reticulum determines the trans-
port pathways of soybean seed storage proteins. Plant J. 40: 238–249.
Muntz, K. (1998). Deposition of storage proteins. Plant Mol. Biol. 38:
77–99.
Nishizawa, K., Maruyama, N., Satoh, R., Fuchikami, Y., Higasa, T.,
and Utsumi, S. (2003). A C-terminal sequence of soybean
b-conglycinin a9 subunit acts as a vacuolar sorting determinant in
seed cells. Plant J. 34: 647–659.
Nishizawa, K., Maruyama, N., Satoh, R., Higasa, T., and Utsumi, S.
(2004). A vacuolar sorting determinant of soybean b-conglycinin
b subunit resides in a C-terminal sequence. Plant Sci. 167: 937–947.
Park, M., Lee, D., Lee, G.J., and Hwang, I. (2005). AtRMR1 functions
as a cargo receptor for protein trafficking to the protein storage
vacuole. J. Cell Biol. 170: 757–767.
Rojo, E., Gillmor, C.S., Kovaleva, V., Somerville, C.R., and Raikhel,
N.V. (2001). VACUOLELESS1 is an essential gene required for vacuole
formation and morphogenesis in Arabidopsis. Dev. Cell 1: 303–310.
Saalbach, G., Rosso, M., and Schumann, U. (1996). The vacuolar
targeting signal of the 2S albumin from Brazil nut resides at the C
terminus and involves the C-terminal propeptide as an essential
element. Plant Physiol. 112: 975–985.
Sanderfoot, A.A., Ahmed, S.U., Marty-Mazars, D., Rapoport, I.,
Kirchhausen, T., Marty, F., and Raikhel, N.V. (1998). A putative
vacuolar cargo receptor partially colocalizes with AtPEP12p on a
prevacuolar compartment in Arabidopsis roots. Proc. Natl. Acad. Sci.
USA 95: 9920–9925.
Schmid, M., Davison, T.S., Henz, S.R., Pape, U.J., Demar, M.,
Vingron, M., Scholkopf, B., Weigel, D., and Lohmann, J.U.
(2005). A gene expression map of Arabidopsis thaliana development.
Nat. Genet. 37: 501–506.
Shimada, T., Fuji, K., Tamura, K., Kondo, M., Nishimura, M., and
Hara-Nishimura, I. (2003a). Vacuolar sorting receptor for seed stor-
age proteins in Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA 100:
16095–16100.
Shimada, T., Koumoto, Y., Li, L., Yamazaki, M., Kondo, M.,
Nishimura, M., and Hara-Nishimura, I. (2006). AtVPS29, a putative
component of a retromer complex, is required for the efficient sorting
of seed storage proteins. Plant Cell Physiol. 47: 1187–1194.
Shimada, T., Kuroyanagi, M., Nishimura, M., and Hara-Nishimura, I.
(1997). A pumpkin 72-kDa membrane protein of precursor accumu-
lating vesicles has characteristics of a vacuolar sorting receptor. Plant
Cell Physiol. 38: 1414–1420.
Shimada, T., Watanabe, E., Tamura, K., Hayashi, Y., Nishimura, M.,
and Hara-Nishimura, I. (2002). A vacuolar sorting receptor PV72 on
608 The Plant Cell
the membrane of vesicles that accumulate precursors of seed storage
proteins (PAC vesicles). Plant Cell Physiol. 43: 1086–1095.
Shimada, T., et al. (2003b). Vacuolar processing enzymes are essential
for proper processing of seed storage proteins in Arabidopsis thaliana.
J. Biol. Chem. 278: 32292–32299.
Silady, R.A., Kato, T., Lukowitz, W., Sieber, P., Tasaka, M., and
Somerville, C.R. (2004). The gravitropism defective 2 mutants of
Arabidopsis are deficient in a protein implicated in endocytosis in
Caenorhabditis elegans. Plant Physiol. 136: 3095–3103.
Takahashi, H., Saito, Y., Kitagawa, T., Morita, S., Masumura, T.,
and Tanaka, K. (2005). A novel vesicle derived directly from
endoplasmic reticulum is involved in the transport of vacuolar
storage proteins in rice endosperm. Plant Cell Physiol. 46:
245–249.
Tamura, K., Shimada, T., Ono, E., Tanaka, Y., Nagatani, A., Higashi,
S.-i., Watanabe, M., Nishimura, M., and Hara-Nishimura, I. (2003).
Why green fluorescent fusion proteins have not been observed in the
vacuoles of higher plants. Plant J. 35: 545–555.
Tamura, K., Takahashi, H., Kunieda, T., Fuji, K., Shimada, T., and
Hara-Nishimura, I. (2007). Arabidopsis KAM2/GRV2 is required for
proper endosome formation and functions in vacuolar sorting and
determination of the embryo growth axis. Plant Cell 19: 320–332.
Vitale, A., and Hinz, G. (2005). Sorting of proteins to storage vacuoles:
How many mechanisms? Trends Plant Sci. 10: 316–323.
Watanabe, E., Shimada, T., Kuroyanagi, M., Nishimura, M., and
Hara-Nishimura, I. (2002). Calcium-mediated association of a puta-
tive vacuolar sorting receptor PV72 with a propeptide of 2S albumin.
J. Biol. Chem. 277: 8708–8715.
Watanabe, E., Shimada, T., Tamura, K., Matsushima, R., Koumoto,
Y., Nishimura, M., and Hara-Nishimura, I. (2004). An ER-localized
form of PV72, a seed-specific vacuolar sorting receptor, interferes the
transport of an NPIR-containing proteinase in Arabidopsis leaves.
Plant Cell Physiol. 45: 9–17.
Yamada, K., Fuji, K., Shimada, T., Nishimura, M., and Hara-
Nishimura, I. (2005). Endosomal proteases facilitate the fusion of
endosomes with vacuoles at the final step of the endocytotic path-
way. Plant J. 41: 888–898.
Zhang, Y., Grant, B., and Hirsh, D. (2001). RME-8, a conserved
J-domain protein, is required for endocytosis in Caenorhabditis
elegans. Mol. Biol. Cell 12: 2011–2021.
Vacuolar Sorting Mutant gfs 609
DOI 10.1105/tpc.106.045997; originally published online February 9, 2007; 2007;19;597-609Plant Cell
Utsumi, Keito Nishizawa, Nobuyuki Maruyama and Ikuko Hara-NishimuraKentaro Fuji, Tomoo Shimada, Hideyuki Takahashi, Kentaro Tamura, Yasuko Koumoto, Shigeru
Secretion of Vacuole-Targeted Green Fluorescent Protein in Their Seeds) Can Be Identified Efficiently bygreen fluorescent seed Vacuolar Sorting Mutants (Arabidopsis
This information is current as of March 30, 2019
Supplemental Data /content/suppl/2007/01/22/tpc.106.045997.DC1.html
References /content/19/2/597.full.html#ref-list-1
This article cites 53 articles, 32 of which can be accessed free at:
Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X
eTOCs http://www.plantcell.org/cgi/alerts/ctmain
Sign up for eTOCs at:
CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain
Sign up for CiteTrack Alerts at:
Subscription Information http://www.aspb.org/publications/subscriptions.cfm
is available at:Plant Physiology and The Plant CellSubscription Information for
ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists