Heterodimerization and Endocytosis of Arabidopsis
Brassinosteroid Receptors BRI1 and AtSERK3 (BAK1)
Eugenia Russinova,aJan-Willem Borst,a,bMark Kwaaitaal,aAna Can ˜o-Delgado,cYanhai Yin,cJoanne Chory,c
and Sacco C. de Vriesa,1
aLaboratory of Biochemistry, Wageningen University, 6703 HA Wageningen, The Netherlands
bMicroSpectroscopy Center, Wageningen University, 6703 HA Wageningen, The Netherlands
cHoward Hughes Medical Institute and Plant Biology Laboratory, Salk Institute for Biological Studies, La Jolla, California 92037
In Arabidopsis thaliana brassinosteroid (BR), perception is mediated by two Leu-rich repeat receptor-like kinases,
BRASSINOSTEROID INSENSITIVE1 (BRI1) and BRI1-ASSOCIATED RECEPTOR KINASE1 (BAK1) (Arabidopsis SOMATIC
EMBRYOGENESIS RECEPTOR-like KINASE3 [AtSERK3]). Genetic, biochemical, and yeast (Saccharomyces cerevisiae)
interaction studies suggested that the BRI1-BAK1 receptor complex initiates BR signaling, but the role of the BAK1 receptor
is still not clear. Using transient expression in protoplasts of BRI1 and AtSERK3 fused to cyan and yellow fluorescent green
fluorescent protein variants allowed us to localize each receptor independently in vivo. We show that BRI1, but not
AtSERK3, homodimerizes in the plasma membrane, whereas BRI1 and AtSERK3 preferentially heterodimerize in the
endosomes. Coexpression of BRI1 and AtSERK3 results in a change of the steady state distribution of both receptors
because of accelerated endocytosis. Endocytic vesicles contain either BRI1 or AtSERK3 alone or both. We propose that the
AtSERK3 protein is involved in changing the equilibrium between plasma membrane–located BRI1 homodimers and
endocytosed BRI1-AtSERK3 heterodimers.
Brassinosteroid (BR) signaling is one of the best-studied signal
transduction pathways in plants. In contrast with animals where
steroid hormones are perceived by nuclear receptors, plants
employ plasma membrane receptors that include the BRASSI-
NOSTEROID INSENSITIVE1 (BRI1) protein. The Leu-rich repeat
function BR-insensitive Arabidopsis thaliana mutant that cannot
be rescued by exogenous application of BRs (Li and Chory,
1997). The BRI1 protein consists of an extracellular domain,
a single transmembrane domain, and a cytoplasmic Ser/Thr
kinase. The extracellular domain contains 25 LRRs and a 70–
amino acid island domain between the 21st and the 22nd LRR
that was found essential for BR binding (He et al., 2000; Wang
et al., 2001). A second LRR RLK, BRI1-ASSOCIATED RECEP-
TOR KINASE1 (BAK1), was identified in an activation-tagging
screen for bri1 suppressors (Li et al., 2002) and in a yeast
(Saccharomyces cerevisiae) two-hybrid screen for BRI1 kinase
domain interacting proteins (Nam and Li, 2002). BAK1 has
a shorter extracellular domain, with only five LRRs, and it lacks
the 70–amino acid island domain. BAK1 is identical to the
previously described Arabidopsis SOMATIC EMBRYOGENESIS
of related RLKs (Hecht et al., 2001). Genetic and molecular data
support the notion that BRI1 and BAK1 (AtSERK3) are part of the
same BR receptor complex, although two other BRI1 homologs
have been reported to function as BR receptors (Yin et al.,
2002b). First, bak1 null alleles are semidwarfs with reduced
BR insensitive phenotypes. Second, overexpression of BAK1
(AtSERK3) suppresses weak bri1 alleles with mutations in the
extracellular and in the kinase domain and gives rise to a BRI1-
overexpression phenotype in wild-type plants (reviewed in
Clouse, 2002; Li, 2003). Several downstream components in
BR signaling have also been identified. Negative and positive
regulators of BR signaling were identified as a GSK-3/Shaggy-
like kinase called Brassinosteroid Insensitive 2 (Li and Nam,
2002) and the nuclear-localized Ser/Thr phosphatase bri1 Sup-
pressor 1 family (Mora-Garcı ´a et al., 2004), respectively. Their
potential substrates are the nuclear proteins BRASSINAZOLE
RESISTANT1 (Wang et al., 2002) and bri1-EMS-SUPPRESSOR1
(BES1) (Yin et al., 2002a). Neither of these proteins has been
shown to be a direct target of the BRI1/BAK1 (AtSERK3)
heterodimer. BRI1 and BAK1 (AtSERK3) localize to the plasma
membrane (Friedrichsen et al., 2000; Li et al., 2002), whereas
the nucleus after BR application (Wang et al., 2002; Yin et al.,
2002a). How the BRI1-BAK1 (AtSERK3) receptor complex trans-
mits the BR signal is not known, although two recent models
suggest similarities to either animal Try kinase or transforming
growth factor-b (TGF-b) cell surface receptor activation (re-
viewed in Clouse, 2002; Li, 2003).
somes are considered to function as signaling compartments
1To whom correspondence should be addressed. E-mail sacco.
email@example.com; fax 31-317-484801.
The author responsible for distribution of materials integral to the
findings presented in this article in accordance with the policy described
in the Instructions for Authors (www.plantcell.org) is: Sacco C. de Vries
Article, publication date, and citation information can be found at
The Plant Cell, Vol. 16, 3216–3229, December 2004, www.plantcell.org ª 2004 American Society of Plant Biologists
in addition to a more general role in receptor recycling (reviewed
in Gonza ´les-Gaita ´n, 2003). During endocytosis, several interme-
diate compartments are distinguished. In the early sorting endo-
somes, a decision is made to direct endocytosed receptors
either toward late endosomes and to degradation in lysosomes
or to recycle the receptors back to the plasma membrane via
recycling endosomes (reviewed in Gruenberg, 2001). In plants,
endocytosis is poorly understood, but similar mechanisms are
proposed (Ueda et al., 2001). Recently, studies of the GDP/GTP
exchange factor for small G-proteins of the auxin responsive
factor class (ARF-GEF), GNOM (Geldner et al., 2003), and of
sterol trafficking in Arabidopsis (Grebe et al., 2003) demonstrate
the importance of endocytosis in plant development. Although
previously demonstrated in plants (Horn et al., 1989), virtually
nothing is known about the role of endocytosis in plant receptor-
mediated signaling. Previously, we showed that the AtSERK1
protein internalizes into early endosomes when coexpressed
with the PP2C type KINASE ASSOCIATED PROTEIN PHOS-
PHATASE (KAPP) in cowpea (Vigna unguiculata) protoplasts
tightly coupled to a mechanism of plant receptor internalization.
us to follow both receptors simultaneously in vivo and to apply
imaging techniques, such as fluorescence lifetime imaging mi-
croscopy (FLIM) to determine Fo ¨rster resonance energy transfer
(FRET), indicative of receptor heterodimerization. These tech-
niques are essential for imaging protein interactions in living plant
(Shah et al., 2001, 2002; Immink et al., 2002) and animal cells
(Sorkin et al., 2000; Haj et al., 2002).
Our results show that BRI1 and AtSERK3 are constantly
recycled via endosomes. When coexpressed, the two receptors
are sorted into different endosomal compartments containing
either BRI1 or AtSERK3 alone, or both. Interestingly, BRI1 and
AtSERK3 do not constitutively interact because they show
interaction mainly in the endosomes and in restricted areas on
the plasmamembrane. Thisresemblesthe endocyticpathway of
internalization of animal receptors and suggests a function of the
AtSERK3 protein in redistributing the BRI1 receptor.
AtSERK3 Accelerates BRI1 Endocytosis in Protoplasts
To determine the subcellular localization of BRI1 and AtSERK3
proteins in plant cells, we performed in vivo targeting experi-
ments in cowpea and Arabidopsis protoplasts derived from leaf
tissue. BRI1 and AtSERK3 proteins were tagged at the C
terminus with either CFP or YFP and then transiently expressed
in protoplasts under control of the constitutive 35S promoter of
Cauliflower mosaic virus. The localization of each fusion protein
was examined using confocal laser scanning microscopy
(CLSM). To show the morphology of the protoplast and the
localization of the fluorescent proteins for each confocal optical
section of a cell, CFP or YFP fluorescence is presented in
a separate channel (cyan and yellow, respectively), followed by
the chlorophyll autofluorescence in the red channel and by an
overlay image combining both.
The transient protein expression system in protoplasts has
been successfully used to study gene regulation, signal trans-
duction (reviewed in Sheen, 2001), protein targeting, and traf-
ficking in plant cells (Jin et al., 2001, 2003; Kim et al., 2001; Ueda
et al., 2001; Sohn et al., 2003; Park et al., 2004). In spite of many
advantages, protoplasts also have limitations. First, localization
of the transiently expressed proteins may not always reflect that
of the endogenous proteins in intact plant cells because of
overexpression. Second, protoplasts aredevoid of cell walls and
plasmodesmata and lack normal cell–cell communication and
polarity cues, so the transiently expressed proteins may behave
differently from the endogenous proteins in intact plant cells.
Thus, the results obtained using transient expression in proto-
plasts should be considered with those limitations in mind.
In single transfections, BRI1-CFP and AtSERK3-CFP fusion
proteins are localized to the plasma membrane as early as 3 h
after transfection, and this pattern was unchanged up to 16 h
(overnight) incubation of the protoplasts. In Figures 1A to 1D,
representative images are shown after 8 h of incubation. Re-
sidual AtSERK3-CFP fluorescence was also seen in the cyto-
plasm, similar to what was observed for AtSERK1 (Shah et al.,
2002; indicated by an arrowhead in Figure 1C), the previously
described close homolog of AtSERK3. When protoplasts were
optically cross-sectioned close to the periphery of the cell,
multiple fluorescently labeled vesicle-like compartments were
observed (examples are shown in Figures 1E and 1F for BRI1-
CFP and in Figures 1G and 1H for AtSERK3-YFP). Often those
organelles appeared to be at a site near the plasma membrane
(indicated by arrowheads in Figures 1E and 1G). The protein
synthesis inhibitor cycloheximide (CHX) was added to proto-
plasts expressing either BRI1-CFP or AtSERK3-YFP 3 h after
transfection, and the protoplasts were examined up to 5 h after
the treatment (Figures 1I and 1J and Figures 1K and 1L,
respectively). In CHX-treated protoplasts, no change in the
distribution of either BRI1-CFP or AtSERK3-YFP fluorescence
between the plasma membrane and the vesicle-like compart-
ments was seen when compared with untreated protoplasts
(data not shown) or to protoplasts expressing BRI1-YFP and
AtSERK3-YFP overnight. In general, the expression was higher
when the longer incubation time was used (cf. Figures 1I and 1K
with Figures 1E and 1G). The constant level of membrane-
localized receptors in the absence of protein synthesis suggests
that neither undergoes rapid internalization when expressed
When BRI1 and AtSERK3 were coexpressed for 3 h as CFP
and YFP fusions, respectively, followed by 5 h of CHX treatment,
there was almost compete depletion of both BRI1 and AtSERK3
fluorescent proteins from the plasma membrane (cf. Figures 1M
to 1P with Figures 1Q to 1T). This suggested that the presence of
AtSERK3 altered the membrane location of BRI1. This could be
because of accelerated endocytosis and/or inefficient recycling
back to the plasma membrane. An interesting observation in the
BRI1/AtSERK3 coexpression experiments was the lack of com-
plete colocalization between the two fluorescently tagged pro-
teins in the endosomes (Figures 1U to 1W). Surprisingly, we
Endocytosis of BRI1 and AtSERK3 3217
Figure 1. Localization of BRI1 and AtSERK3 Proteins in Cowpea and in Arabidopsis Protoplasts.
(A) to (H) Confocal images of cowpea protoplast transfected with single constructs, BRI1-CFP ([A] and [E]), AtSERK3-CFP, and AtSERK3-YFP in (C)
and (G), respectively, and incubated in protoplast medium for 8 h. Because no differences in the localization and the expression levels between the
CFP- and the YFP-tagged versions of the proteins were observed, we present the data with either the CFP or the YFP construct. BRI1 and AtSERK3
3218 The Plant Cell
distinguished three different types of vesicle-like compartments,
compartments that contained BRI1 and AtSERK3 together (in-
dicated by an arrowhead in Figure 1W) and compartments that
contained either AtSERK3 (cyan arrowhead in Figure 1U) or BRI1
alone (yellow arrowhead in Figure 1V). The ratio between the
numbers of each type of compartment differed among different
protoplasts. Similar observations were made when Arabidopsis
protoplasts were cotransfected with BRI1 and AtSERK3 fluores-
cent proteins (Figure 1Z). Because an endocytic compartment
will contain more than a single molecule of the fluorescently
labeled receptor, this excludes the possibility of a random event.
Thus, we conclude that the endosomal-sorting mechanism in
plant cells can distinguish among different plasma membrane
To verify that the vesicular origin of the BRI1 and AtSERK3
fluorescent proteins is the plasma membrane and does not
represent proteins in transit from the Golgi membranes, we used
rat sialyltransferase fused to YFP (STtmd-YFP) as a Golgi marker
for colocalization experiments with either BRI1-CFP (Figures 2A
to 2D)orAtSERK3-CFP–tagged proteins (Figures 2Ito2L).Itwas
previously shown that the STtmd-YFP localizes to the trans side
of the Golgi compartments in tobacco (Nicotiana tabacum) cells
(Boevink et al., 1998), cowpea protoplasts (Carette et al., 2000),
and in Arabidopsis (Grebe et al., 2003). No colocalization
between STtmd-YFP and either BRI1-CFP (Figure 2D) or
AtSERK3-CFP fluorescence (Figure 2L) was observed.
In Arabidopsis roots, Brefeldin A (BFA) inhibits post-Golgi
vesicle trafficking by targeting large ARF-GEFs, resulting in rapid
internalization of plasma membrane proteins in so-called BFA
compartments (Geldner et al., 2001) containing fused endo-
somes (Geldner et al., 2003). BFA added to the cowpea proto-
plasts coexpressing STtmd-YFP and either BRI1-CFP or
AtSERK3-CFP, destroyed the integrity of the Golgi vesicles and
causing accumulation of the STtmd-YFP fluorescence on the
arrowheads in Figures 2F and 2N). Similar effects of BFA in
cowpea protoplasts were previously observed using the Golgi
marker ERD2-YFP (Pouwels et al., 2002). No change in the
morphology and localization of BRI1 and AtSERK3 fluorescently
labeled vesicles was observed after BFA application (cf. Figures
2E and 2M with Figures 2A and 2I). This suggested that in
cowpea protoplasts derived from a mesophyll tissue, the in-
ternalization and recycling of BRI1 and AtSERK3 proteins are
To confirm the endosomal localization of the BRI1 and
AtSERK3 proteins, we examined the localization of BRI1-YFP
and AtSERK3-YFP fusion proteins simultaneously with the
lipophilic dye FM4-64 (Figure 3). FM4-64 is routinely used to
follow the endocytic pathway in yeast and in plant cells (Vida and
Emr, 1995; Ueda et al., 2001). As indicated by arrowheads in
Figures 3C and 3F, most of the BRI1- and AtSERK3-containing
vesicles are stained with FM4-64, suggesting that BRI1 and
AtSERK3 proteins are indeed recycled and internalized via an
and in Endocytic Compartments
To determine where BRI1 and AtSERK3 receptor homodimeri-
zation or heterodimerization occurs in cells, we applied FLIM.
Thistechnique captures FREToccurring between theCFP donor
and the YFP acceptor molecules when both are in close
proximity of ;5 nm. This distance is regarded to be indicative
of direct protein–protein interaction (Bastiaens and Pepperkok,
2000; Hink et al., 2002). Using a combination of a two-photon
laser-scanning microscope and a time-correlated single photon
counting acquisition card, a fluorescence intensity image of the
tagged receptors is obtained. Subsequently, the fluorescence
lifetime (t)of theCFP donormolecule iscalculated for everypixel
in the image according to a double exponential decay model.
Upon FRET, the t of the CFP donor molecules will decrease and
is visualized as a false color-coded image superimposed over
the visible light image. The mean fluorescence lifetime ranges
from dark blue (t ¼ 2.5 ns; no interaction) to dark orange or red
(t ¼ 1.9 ns; interaction) (Figure 4). FLIM measurements per-
formed with all receptor combinations used in this analysis are
summarized in Table 1.
We first investigated whether AtSERK3 and BRI1 were able
to form homodimers (Figures 4A and 4B and Figures 4C and
4D, respectively). In protoplasts expressing AtSERK3-CFP/
Figure 1. (continued).
localized to the plasma membrane as shown in (A) for BRI1-CFP and in (C) for AtSERK3-CFP (cyan). An arrowhead in (C) indicates some residual
AtSERK3-CFP fluorescence in the cytoplasm. BRI1 and AtSERK3 also localized to the multiple vesicles when protoplast transiently expressing BRI1-
CFP ([E], cyan) and AtSERK3-YFP ([G], yellow) were optically cross-sectioned through the periphery of the cell. Note the arrowheads pointing to
vesicles budding from the plasma membrane. The combined images of (A), (C), (E), and (G) with the chlorophyll autofluorescence (red) are shown in (B),
(D), (F), and (H), respectively.
(I) to (L) Confocal images of cowpea protoplasts transfected with BRI1-YFP ([I] and [J]) and AtSERK3-YFP ([K] and [L]) incubated for 3 h in protoplast
medium and then for 5 h in the presence of 50 mM CHX. The images combined with the chlorophyll autofluorescence (red) images are shown in (J) and
(M) to (T) Confocal images of cowpea protoplast cotransfected with AtSERK3-CFP and BRI1-YFP incubated for 3 h in protoplast medium and then for
is shown in (N) and (R). The chlorophyll autofluorescence is shown in (O) and (S) and the combined images in (P) and (T), respectively.
(U) to (W) Confocal images of cowpea protoplast cotransfected with AtSERK3-CFP and BRI1-YFP recorded 3 h after transfection. The AtSERK3-CFP
and the BRI1-YFP fluorescence localized in the endosomes is shown in (U) and (V), respectively, and the combined image in (W). Note the arrowheads
pointing to the three different types of endosomes present in the cell.
(Z) A confocal image of Arabidopsis protoplast cotransfected with BRI1-CFP and AtSERK3- YFP recorded 16 h after transfection. Bar ¼ 10 mM.
Endocytosis of BRI1 and AtSERK33219
AtSERK3-YFP constructs, the lifetime of AtSERK3-CFP at the
plasma membrane remained 2.49 ns (displayed as a homoge-
neous dark blue membrane in Figure 4B and corresponding
t values in Table 1). Occasionally, single small areas showing
1). By contrast, in BRI1-CFP/BRI1-YFP–expressing protoplasts,
a reduction in the fluorescence lifetime of the donor CFP
molecules (t ¼ 2.5 to 1.9 ns) was detected, suggesting that
BRI1 was able to form homodimers in the plasma membrane
(FRET efficiency 23%). The reduction in the fluorescence lifetime
reduction of the fluorescence lifetime of AtSERK1-CFP in the
presence of AtSERK1-YFP (Table 1), which was previously
shown to form homodimers in the plasma membrane (Shah
et al., 2001; M.A. Hink, K. Shah, E. Russinova, S.C. de Vries, and
A.J.W.G. Visser, unpublished results). The areas where BRI1/
BRI1 homodimers in the plasma membrane occur do not appear
indicated by arrowheads in Figure 4D). We conclude that the
living plant cells. We cannot exclude the possibility that in some
cases the lack of FRET is because of a competitive interaction
between endogenous and transiently expressed BR receptors.
Figure 2. Colocalization of BRI1 and AtSERK3 Proteins with the Golgi STtmd-YFP Marker in Cowpea Protoplasts.
Confocal images of protoplast cotransfected with BRI1-CFP/STtmd-YFP ([A] to [D]) and with AtSERK3-CFP/STtmd-YFP ([I] to [L]) recorded 3 h after
transfection. The BRI1-CFP and AtSERK3-CFP fluorescence localized to the endosomes is shown in (A) and (I) (cyan). The STtmd-YFP fluorescence
localized to the Golgi stacks is shown in (B) and (J) (yellow). The chlorophyll autofluorescence (red) is shown in (C) and (K) and the combined images in
(D) and (L). Protoplast cotransfected with BRI1-CFP/STtmd-YFP ([E] to [H]) and with AtSERK3-CFP/STtmd-YFP ([M] to [P]) were allowed to express
the proteins for 3 h and were then incubated in the presence of 20 mg/mL BFA for 30 min. The BRI1-CFP and AtSERK3-CFP fluorescence is shown in (E)
and (M), respectively. The STtmd-YFP fluorescence is shown in (F) and (N). Membranes resembling the endoplasmic reticulum are indicated by
arrowheads in (F) and (N). The chlorophyll autofluorescence is shown in (G) and (O) and the combined images in (H) and (P). Bars ¼ 10 mM.
3220 The Plant Cell
This depends on the abundance of the endogenous receptors in
the respective tissues. Although the BRI1 receptor is expressed
relatively high in mesophyll cells (Friedrichsen et al., 2000), FRET
was always observed between transiently coexpressed BRI1-
CFP and BRI1-YFP receptors. The abundance of AtSERK3
transcripts in mesophyll cells was low as predicted by RT-PCR
(C. Albrecht and S.C. de Vries, unpublished results), suggesting
that the lack of interaction between transiently coexpressed
AtSERK3-CFP and AtSERK3-YFP receptors is unlikely to be
caused by interaction with its native counterpart and most likely
reflects the monomeric state of AtSERK3.
To determine whether BRI1/AtSERK3 heterodimerization oc-
curred in vivo, we transfected protoplasts with either BRI1-CFP
and AtSERK3-YFP or AtSERK3-CFP and BRI1-YFP constructs.
Because the plasma membrane of the cells cotransfected with
BRI1 and AtSERK3 is rapidly depleted of the receptors, we first
analyzed those parts where the fluorescent proteins were still
present (Figures 4E and 4F). In such areas, the fluorescence
intensity image of a protoplast cotransfected with, for example,
AtSERK3-CFP and BRI1-YFP showed a uniform distribution of
the CFP fluorescence (the boxed plasma membrane area in
or very little reduction in the fluorescence lifetime (t ¼ 2.47 ns)
was observed (visualized by the dark blue color in Figure 4F).
Those values corresponded to the values of the fluorescence
CFP (t ¼ 2.47 ns) or AtSERK3-CFP (t ¼ 2.53 ns) constructs
alone (Table 1) and indicated no interaction occurring between
AtSERK3 and BRI1 in those areas. In addition, multiple small
(Table 1) of the donor (BRI1-CFP) molecules (visualized as green
to orange patches, and indicated by arrowheads, in Figure 4F)
were observed. The FRET efficiency in these areas was ;20%
(Table 1). The size and the frequency of these discrete areas
varied among different protoplasts.
We next determined the lifetime of the AtSERK3-CFP donor
molecules in the presence of BRI1-YFP in the endocytic com-
partments at a site near the plasma membrane (Figures 4G and
4H) or located in the cytoplasm (Figures 4I and 4J). Twin-like
FRET efficiency 20%). Comparable reductions in the fluores-
cence lifetime oftheAtSERK3-CFP were observedinmostof the
BRI1 and AtSERK3 containing endocytic compartments located
in the cytoplasm (see Figure 4J where spotted areas with
different fluorescence lifetimes are visualized in green and
orange and indicated by arrowheads). We conclude that BRI1-
AtSERK3 heterodimerization is nonuniformly distributed in the
plasma membrane and appears to coincide with developing
endocytic compartments that contain both receptors.
AtSERK3-Accelerated Endocytosis of BRI1 in Protoplasts
Represents an Activated BR Signaling
To determine if brassinolide (BL) treatment of the cowpea pro-
toplasts coexpressing BRI1 and AtSERK3 fluorescently tagged
proteins would alter the amount or the size of the membrane
areas in which FRET occurred, we performed FLIM experiments
on protoplasts coexpressing BRI1 and AtSERK3 in the presence
of 1 mM BL. No changes in the t values of the CFP donor
molecules (Table 1) or in the number and distribution of the
membrane areas showing FRET were observed (data not
shown). Because BL has been shown to specifically bind to
BRI1and increaseBRI1phosphorylation (Wang etal.,2001),one
possibility could be that BRI1 and AtSERK3 show ligand-
independent endocytosis. Alternatively, the protoplasts could
contain sufficient endogenous amounts of BL to activate the
introduced fluorescent receptors.
To determine whether our FLIM data represent elements of an
activated BR signaling pathway, we made use of the BES1
protein, one of the two proposed nuclear targets of BR signaling.
Unphosphorylated BES1 protein was reported to accumulate in
the nucleus in response to BL and to promote cell elongation
in dark-grown Arabidopsis hypocotyls (Yin et al., 2002a). We
transiently expressed the BES1-GFP fusion protein in proto-
plasts and examined the cells before and after BL application.
The GFP fusion with the gain-of-function mutant bes1-D, which
was previously shown to accumulate in the nucleus without BL
application (Yin et al., 2002a), was used as a positive control. In
cowpea protoplasts, both the BES1-GFP and the mutant bes1-
GFP fusions were observed in the nucleus as early as 3 h after
transfection (arrowheads in Figures 5A and 5B). BES1-GFP and
bes1-GFP fluorescent proteins were also seen in the cytoplasm
and in the plasma membrane (Figures 5A and 5B). Overnight
expression of either BES1-GFP or mutant bes1-GFP proteins
induced cell elongation and eventually the rupture of the proto-
plasts, indicating the presence of functional BES1 proteins in the
protoplast system (data not shown). Application of BL had no
effect on the localization of either BES1 or bes1-D fluorescent
proteins (data not shown). We used Arabidopsis protoplasts
Figure 3. FM4-64 Labeling.
Confocal images of cowpea protoplast expressing BRI1-YFP (A) and
AtSERK3-YFP ([D], yellow) were labeled with FM4-64 (red) and viewed
after 3 h for BRI1-YFP (B) and for AtSERK3-YFP (E). The superimposed
images of (A) and (B) resulted in (C), and the superimposed images of (D)
and (E) resulted in (F). Arrowheads point to the overlapping yellow and
red dots representing endosomes in (C) and (F). Bars ¼ 10 mM.
Endocytosis of BRI1 and AtSERK3 3221
isolated from mutants defective in the biosynthesis of BRs, such
as cbb1, det2-1, and dwf4 to transiently express BES1-GFP
protein (Figures 5C to 5E, respectively). In all mutant back-
grounds, BES1 protein was localized in the nucleus without BL
treatment. We conclude that contrary to the previous observa-
tion made by Yin et al. (2002a), in mesophyll cells BES1 nuclear
localization is not dependent on the BR application. Similar
observations were reported previously when the nuclear local-
2002). We next checked the phosphorylation state of BES1
protein without and with BL treatment in Arabidopsis pro-
toplasts. In the absence of ligand, BES1 was found in phos-
phorylated form after BL treatment was shifted into a
dephosphorylated form of faster electrophoretic mobility. It
was proposed that the dephosphorylated form of BES1 modu-
lates the transcription of BL-regulated genes (Yin et al., 2002a).
As shown in Figure 5F, BES1-GFP was detected as a fast and
a slow migrating band after SDS-PAGE and immunoblotting with
anti-BES1 antibody or anti-GFP antibody (data not shown)
representing phosphorylated and unphosphorylated forms (Yin
et al., 2002a). The presence of the two forms in our system was
not dependent on BL application (Figure 5F, cf. lane 1 with lane
2). Based on the presence of unphosphorylated BES1 protein in
the nucleus, we conclude that in protoplasts the BR signaling is
active. Therefore, our observations concerning BRI1/AtSERK3
endocytosis appear to represent active BR signaling.
BRI1 Is Endocytosed in Roots in Arabidopsis Plants
RNA gel blot experiments (Li and Chory, 1997) and real-time
quantitative RT-PCR (Shimada et al., 2003) have shown that in
Arabidopsis the BRI1 gene is expressed ubiquitously and dis-
plays very little organ specificity. BRI1-GFP fluorescence was
observed in cells of the hypocotyls, roots, and cotyledons of
seedlings stably expressing BRI1-GFP fusion protein from the
BRI1 promoter (Friedrichsen et al., 2000). Although not organ
Figure 4. FRET between BRI1 and AtSERK3 Imaged by FLIM.
(A) to (D) FLIM on cowpea protoplast transiently expressing AtSERK3-CFP/AtSERK3-YFP ([A] and [B]) and BRI1-CFP/BRI1-YFP ([C] and [D]) for 16 h.
This time point was selected because a sufficient amount of fluorescence is required to measure FRET. Intensity images representing a steady state of
the donor CFP fluorescence are presented in (A) for AtSERK3-CFP and in (C) for BRI1-CFP, respectively. The mean fluorescence lifetime values (t) were
calculated as described in Methods, and the lifetime distribution of the outlined regions in (A) and (C) are presented as enlarged pseudocolor images in
(B) and (D). Arrowheads point to the areas with a significant reduction of the lifetime (dark orange to green, t ¼ 1.9 to 2.0 ns). Note the color bar where
dark blue is used to display t ¼ 2.5 ns (no interaction) and the red to dark orange to display t ¼ 1.9 ns (interaction).
(E) to (J) FLIM on cowpea protoplast transiently coexpressing AtSERK3-CFP and BRI1-YFP proteins for 16 h. The intensity images of the donor CFP
fluorescence are presented in (E), (G), and (I). Enlargements of the outlined regions in (E), (G), and (I) are presented as pseudocolor images in (F), (H),
and (J), respectively. The arrowheads point to the areas with short lifetime, indicative for FRET.
3222 The Plant Cell
specific, BRI1-GFP expression was temporally regulated, and
adult tissues always displayed low levels of BRI1-GFP fluores-
fluorescence was reported to be exclusively plasma membrane
localized (Friedrichsen et al., 2000; Li et al., 2002; Nam and Li,
2002). Based on our observation that in protoplasts coexpres-
the complex, we further explored the possibility that a similar
phenomenon could be observed in planta. We therefore in-
troduced the BRI1-GFP fusion under the control of the BRI1
et al., 2000). Two independent lines containing a single copy
of the BRI1-GFP construct were selected and analyzed further.
Table 1. Fluorescence Lifetime Analysis and FRET Characterization in Cowpea Protoplasts
Proteins Cellular LocalizationLifetime t (ns) 6 SD
FRET Efficiency (E) (%)FRETþ
2.57 6 0.03
2.47 6 0.03
2.53 6 0.04
2.51 6 0.02
2.00 6 0.11
2.00 6 0.09
2.01 6 0.10
1.90 6 0.11
2.49 6 0.02
1.96 6 0.11
FRET efficiency (E) is determined as described in Methods. N, total number of protoplasts analyzed; FRETþ, number of protoplasts showing FRET;
FRET?, number of protoplasts not showing FRET; n, number of independent transfection experiments; PM, plasma membrane; –, FRET efficiency is
Figure 5. BES1 Expression in Protoplast.
(A) and (B) Confocal images of cowpea protoplasts transiently expressing BES1-GFP (A) and bes1-GFP ([B], green) 3 h after transfection. The nuclear-
localized fluorescent protein is indicated by arrowheads.
(C) to (E) Confocal images of Arabidopsis protoplasts derived from different BR biosynthetic mutants, cbb1 (C), det2-1 (D), and dwf4 (E), and transiently
expressing BES1-GFP. In all images, the chlorophyll autofluorescence is shown in red.
(F) Immunoblot analysis of the transiently expressed BES1-GFP protein in Arabidopsis protoplasts detected with anti-BES1 antibody. The blot shows
that fast migrating band of;65 kD representing the unphosphorylated BES1-GFP is present in Arabidopsis protoplasts without and after BL treatment
(lanes 1 and 2, respectively). Lane 3 contains protein extract from protoplasts cotransfected with BRI1-CFP and AtSERK3-YFP. An asterisk indicates
additional bands that correspond in size to the endogenous BES1 proteins.
Endocytosis of BRI1 and AtSERK3 3223
roots of 7-d-old light-grown seedlings. As previously described,
BRI1-GFP fluorescence was detected on the plasma membrane
in roots but surprisingly also in endocytic compartments (Figure
6B). This occurred first at the apical part of the root meristem
immediately adjacent to the quiescent center and to a lesser
extent in the lateral and the columella root cap cells (Figure 6A)
and in the elongation zone. In the mature parts of the root, BRI1-
GFP fluorescence was reduced but also seen in the plasma
membrane and in the endosomes (data not shown). This obser-
vation supports the hypothesis that the increased rate of BRI1
endocytosis is correlated with the places where active BR
signaling is occurring in roots.
FM4-64 staining was used to colabel the vesicles expressing
BRI1-GFP. As shown in Figures 6F to 6H, FM4-64 colocalizes
with most of the BRI1-GFP containing vesicles. We next ana-
lyzed the effect of BFA on BRI1-GFP endocytosis in Arabidopsis
roots. BFA application in the presence of CHX caused reversible
aggregation of the BRI1-GFP fluorescence in compartments
similar in morphology and localization to the BFA-induced
compartments of the pin-formed (PIN) proteins (Geldner et al.,
2001, 2003) (Figures 6D and 6E). These results suggest that
protoplasts also occurs in Arabidopsis root cells. To show that
the observed difference in BFA sensitivity between cowpea
mesophyll cells and Arabidopsis root cells is not an artifact of the
cowpea protoplasts, we generated protoplasts from 5- to 7-d-
old BRI1-GFP transgenic seedlings (Figures 6I and 6J). BFA
application of the protoplasts caused formation of BFA bodies
only in the chlorophyll-free cells derived from either root or
Figure 6. BRI1 Endocytosis in Arabidopsis Roots.
(A) and (B) Confocal images of root from 7-d-old ProBRI1-BRI1-GFP transgenic seedling. BRI1-GFP signal is detected internalized in the cells
immediately adjacent to the quiescent center ([A], arrowhead) and in the meristem zone as shown in (B).
(C) A confocal image of root meristem from bak1/serk3 mutant seedlings expressing ProBRI1-BRI1-GFP construct.
(D) and (E) Confocal images of wild-type BRI1-GFP expressing roots after BFA application in the presence of CHX (D) followed by 2 h washing out in (E).
Note the BFA-induced accumulation of BRI1-GFP indicated by an arrowhead.
(F) to (H) Confocal images of FM4-64–stained root meristem area. BRI1-GFP fluorescence (green) is shown in (F). FM4-64–stained cells (red) are shown
in (G). The combined images of (F) and (G) resulted in (H). Arrowheads mark vesicles colabeled by both BRI1-GFP and FM4-64.
(I) to (K) Confocal images of protoplasts derived from 7-d-old ProBRI1-BRI1-GFP transgenic seedling.
(I) Root protoplasts expressing BRI1-GFP.
(J) Protoplasts from the aerial part of the seedling expressing BRI1-GFP.
(K) Mixture of root and shoot protoplasts expressing BRI-GFP after BFA application. Note the formation of BFA compartments only in the chlorophyll-
free cells (arrowhead). Bar ¼ 10 mM.
3224 The Plant Cell
hypocotyl parts of the plant (arrowhead in Figure 6K). We never
observed formation of BFA compartments in the chlorophyll-
containing cells under the same conditions (Figure 6K), although
some peripheral aggregation of BRI1 fluorescent protein was
occasionally observed. This observation also suggests there are
differences in the response of different plant cells to application
To determine whether the rate of BRI1 endocytosis in Arabi-
dopsis root cells is dependent on or reduced in the absence of
the AtSERK3 protein, the BRI1-GFP transgenic line was crossed
to a serk3/bak1 null mutant. No change in the rate of BRI1-GFP
endocytosis in the serk3/bak1 background compared with the
wildtypewas observed(cf. Figure6Bwith6C).Weconclude that
the observed BFA-sensitive trafficking of BRI1 in Arabidopsis
roots reflects general recycling as was also observed for the PIN
BRI1 and AtSERK3 Dynamics in Protoplasts
When transiently expressed as translational CFP or YFP fusions
in cowpea protoplasts, our results indicate that BRI1 and
AtSERK3 are localized to the plasma membrane, similar to
what was observed in Arabidopsis roots for BRI1 (Friedrichsen
etal.,2000)and BAK1(Lietal.,2002;Namand Li,2002).Wealso
detected both proteins in small vesicle-like compartments in the
cytoplasm close to the plasma membrane. Results of several
experiments led to the conclusion that these vesicles represent
endosomes. First, the vesicles containing BRI1 and AtSERK3
colocalize with the fluorescent endocytic tracer FM4-64 that is
used to colabel early endosomes in yeast (Vida and Emr, 1995)
and in plant cells (Ueda et al., 2001). Second, the BRI1 and
AtSERK3 containing vesicles do not contain a trans-Golgi
marker. Third, in contrast with the trans-Golgi compartments
that were larger, less mobile, and highly sensitive to BFA, the
BRI1 and AtSERK3 containing vesicles were smaller, highly
motile, and BFA insensitive.
The turnover of BRI1 and AtSERK3 proteins in protoplasts
when coexpressed in protoplasts, BRI1 and AtSERK3 first
colocalized at the plasma membrane and then are rapidly in-
ternalized into endosomes. Most of the internalized BRI1 and
AtSERK3 proteins were not recycled back to the membrane and
were possibly targeted for degradation, resulting in a nearly
complete depletion of BRI1 and AtSERK3 fluorescent proteins
from the plasma membrane. We believe that the rapid internal-
ization of the coexpressed BRI1 and AtSERK3 proteins is not
a result of overexpression or heterodimerization with cowpea
orthologs because neither BRI1-CFP/YFP nor AtSERK3-CFP/
YFP coexpression resulted in an increased rate of protein in-
ternalization. Because no differences between cowpea and
Arabidopsis expression systems were observed, we conclude
that receptor internalization via endocytosis generally occurs in
Receptor recycling is well documented in animal cells. For
instance, all members of the epidermal growth factor receptor
(EGFR) family are predominantly localized in the plasma mem-
brane and even in the absence of ligands are slowly internalized
yet quickly recycled back to the membrane through endosomes.
Epidermal growth factor ligand binding and activation of the
receptor downregulation in the lysosomes (reviewed in Wiley,
2003). Evidence that similar mechanisms operate in plants came
from recent studies showing that membrane proteins, such as
the auxin efflux carrier PIN1 (Geldner et al., 2001), cell wall
pectins (Balus ˇka etal.,2002),and sterols (Grebeetal., 2003),are
actively recycled through endosomes in either Arabidopsis or
maize (Zea mays) roots. PIN1 recycling and proper targeting
required the BFA sensitive ARF-GEF, GNOM, because the BFA-
induced rapid internalization of PIN1 in BFA compartments was
because of blocking of the resecretion of the protein via GNOM
from the endosomes to the plasma membrane (Geldner et al.,
2003). We have observed differences between mesophyll and
root cells in the BFA sensitivity of the endosomal compartments
containing BRI1 and AtSERK3 proteins. So far the assembly of
the BFA compartments enriched in plasma membrane proteins,
sterols, or pectins was mainly investigated in roots. Although
highly speculative, the different BFA sensitivity of the endocytic
compartments in different cell types might reflect differences in
the actin cytoskeleton (reviewed in Geldner, 2004; Sˇamaj et al.,
2004). Differences in the distribution of the Golgi markers in
responses to BFA were previously reported between maize and
onion (Allium cepa) root cells (Satiat-Jeunemaitre and Haves,
1992) and tobacco mesophyll cells (Boevink et al., 1998), BY-2
cells (Ritzenthaler et al., 2002), or Arabidopsis mesophyll proto-
plasts (Kim et al., 2001). Ritzenthaler et al. (2002) proposed
a model in which the early effects of BFA depends on the
physiological status of the Golgi and can result in either fusion of
the Golgi with the endoplasmic reticulum or clustering of the
Golgi aroundperinuclear BFAcompartments asobservedin root
cells (Satiat-Jeunemaitre and Haves, 1992; Geldner et al., 2003).
Because BR signaling is active in young tissues of shoots and
roots (Shimada et al., 2003), it is very unlikely that differences in
the BFA sensitivity reflect differences in the BR signaling mech-
The mechanism of animal receptor internalization is well
documented but not completely understood. It is proposed
that EGFR internalization is mediated mainly by clathrin-coated
pits (Galperin and Sorkin, 2003), whereas an additional raft-
caveolar internalization pathway was recently demonstrated for
the ligand-receptor complexes pass through a series of endo-
somal compartments where they are sorted to different intracel-
lular destinations. It is generally accepted that the early
endosomes are the first sorting station of the internalized
receptors (reviewed in Gruenberg, 2001). From that point,
some receptors are recycled to the plasma membrane through
the recycling endosomes. Others are sent to the late endosomal
compartments and lysosomes for degradation, which is facili-
tated by ubiquitination of the receptors (reviewed in Katzmann
et al., 2002; Sorkin and Von Zastrow, 2002).
When BRI1 and AtSERK3 were coexpressed together, three
Endocytosis of BRI1 and AtSERK33225
and AtSERK3 together. This intriguing observation suggests that
BRI1 and AtSERK3 might undergo receptor-specific sorting. It is
or how the observed proportions among endosomes containing
individual monomeric, homodimeric, or heterodimeric receptors
were accomplished. It seems likely that the three different
endosomes arise directly from the plasma membrane. Studies
of endocytosis in plants are fairly limited, but the analysis of the
Arabidopsis genome point to conservation in most of the
endocytic machinery (reviewed in Ju ¨rgens and Geldner, 2002).
The high number of putative endosomal Rab GTPases and the
fact that protein recycling through the ARF-GEF GNOM is not
exclusive in Arabidopsis suggests the existence of a rather
complex endosomal system in plants. This may consist of
several functionally independent endosomes involved in distinct
recycling pathways (Rutherford and Moore, 2002; Ueda and
Nakano, 2002; Geldner et al., 2003).
Previously, it was shown that BRI1 and BAK1 proteins could
phosphorylate each other in vitro and in vivo and that bak1
mutants phenotypically resemble weak bri alleles (Li et al., 2002;
Nam and Li, 2002). In our protoplast system, we observed that
the interaction between BRI1 and AtSERK3 was restricted to
only a few small parts of the plasma membrane. This was
surprising because BRI1 and AtSERK3 were colocalized in the
entire plasma membrane. BRI1-AtSERK3 heterodimerization
was also consistently observed in vesicles that were at the site
near the plasma membrane and vesicles that were internalized
and located in the cytoplasm. We therefore propose that
heterodimerization between both receptors mainly occurs
upon the onset of endocytosis. Whether this can be equated
with transphosphorylation events between BRI1 and AtSERK3 is
atpresent notknown. Forinstance,for theEGF receptor ErbB1 it
has been shown that intermolecular phosphorylation can occur
between individual receptor molecules inthe absence ofa ligand
(Verveer et al., 2000).
We observed that in protoplasts, exogenous BRs were not
required for BRI1and AtSERK3 interaction. This could mean that
the observed endocytosis is a ligand-independent process. In
that case, it apparently does not require BL-dependent BRI1
However, the presence of unphosphorylated BES1-GFP protein
in protoplast nuclei similar to the BL-dependent system, de-
scribed by Yin et al. (2002a), suggests that the observed
endocytosis of BRI1 and AtSERK3 takes place in cells that
show active BL signaling responses.
In protoplasts, a part of the BRI1 receptor molecules are in the
homodimeric state. AtSERK1, a close homolog of AtSERK3, is
similarly able to homodimerize when expressed in protoplasts
(Shah et al.,2001). Bycontrast,AtSERK3 was completely unable
to homodimerize in protoplasts. AtSERK3 lacks the second Cys
pair flanking the LRR in the extracellular domain, and this pair
might be essential for intermolecular interactions and receptor
homodimerizaton (Die ´vart and Clark, 2003).
Although not excluded, BRI1 homodimerization has so far not
been demonstrated in plants, and bri1 mutant analysis sug-
gested that BRI1 might function in a heterodimer (Li et al., 2002;
Nam and Li, 2002). If a situation such as observed in our proto-
plasts reflects the situation in plants, a mosaic of homodimeric
and heterodimeric BRI receptors could be envisaged. In that
scenario, BRI1 can transduce a signal through either BRI1/BRI1
homodimers orthrough heterodimers with AtSERK3 (BAK1). The
weak phenotype of the serk3/bak1 null allele suggests that the
BRI11/AtSERK3 (BAK1) pair is also not exclusive, and BRI1 can
form heterodimers with other AtSERK3-like proteins. We can
tune BR signaling by shifting the equilibrium of the membrane-
localized BRI1 receptors toward the endosomes.
BRI1 Dynamics in Planta
in planta. In Arabidopsis roots, BRI1 undergoes endocytosis in
the primary meristem zone and thus correlates with an area
where BR-controlled cell growth occurs. BRI1 recycling is BFA
sensitive so, at least in part, resembles the ARF-GEF GNOM-
mediated endocytic pathway also used by the PIN1 protein.
Therefore, this pathway may function to maintain the required
level of several membrane proteins reminiscent of clathrin-
mediated receptor recycling in animal cells. Our observation
that BRI1 receptor internalization in Arabidopsis roots was not
affected in bak1/serk3 mutant plants is likely because of func-
tional redundancy. AtSERK3 belongs to a small family of at least
five closely related homologs in Arabidopsis (Hecht et al., 2001;
Peng and Li, 2003). Whereas a direct link between BR signaling
and other members of the SERK family has not yet been
demonstrated, FRETexperiments showedthat atleasttwoother
SERK homologs are able to form heterodimers with BRI1 in
protoplasts (E. Russinova, J.W. Borst, and S.C. de Vries, un-
Recently, endocytosis has been demonstrated to be not only
amechanismforreceptor downregulation butalsoaprerequisite
for signaling through animal receptor Try kinases and TGF-b
receptors (reviewed in Gonza ´les-Gaita ´n, 2003). In animal TGF-b
receptor signaling, it was recently demonstrated that ligand-
independent constitutive trafficking occurs via a different endo-
cytic pathway. One of these pathways follows the classic
clathrin-dependent pathway, while simultaneously a second
one is dependent on lipid rafts (Di Guglielmo et al., 2003). Of
particular interest in the TGF-b model is that ligands do not
regulate receptor trafficking but rather stabilize the heteromeric-
receptor complex involved in active signaling during subsequent
intracellular trafficking. If such a scenario would apply to the
BRI1-AtSERK3 heterodimer, the endosomes containing both
receptors as we observed in this work would be candidates for
endosomes directly involved in BR signaling.
Plant Material and Growth Conditions
Arabidopsis thaliana ecotype Columbia (Col) was used as the wild type.
Seeds were germinated either on half-strength MS medium (Duchefa,
grown under fluorescent light (16-h-light/8-h-dark cycles). Arabidopsis
BRI1-GFP–expressing plants were generated by introducing the ProBRI1-
BRI1-GFPconstruct (Friedrichsenet al.,2000) using thefloral dip method
3226 The Plant Cell
(Clough and Bent, 1998). Transgenic seedlings were selected on half-
strength MS medium containing 50 mg/L kanamycin. Seven-day-old
Arabidopsis seedlings from two independent homozygous ProBRI1-BRI1-
GFP lines (T3 generation) were used to determine the BRI1-GFP local-
ization in roots using CLSM. A ProBRI1-BRI1-GFP transgenic line was
crossed with a bak1/serk3 null mutant (SALK_034523.56.00.x) generated
by the SALK Institute (http://signal.salk.edu). Seeds from cbb1, det2-1,
and dwf4 Arabidopsis mutants were obtained from the the Arabidopsis
Construction of the CFP/YFP-Tagged Proteins
The full-length cDNA of BRI1 was PCR amplified from an EST (Asamizu
et al., 2000) obtained from Kazusa DNA Research Institute (Kisarazu City,
Japan) with primers BRI1-F, 59-CATGCCATGGATGAAGACTTTTTC-
AAGC-39, and BRI1-R, 59-CATGCCATGGCTAATTTTCCTTCAGGAA-39.
The BRI1 cDNA was then inserted in the NcoI site upstream of the
CFP/YFP tags of the pMON999 (Monsanto, St. Louis, MO) vectors to
generate BRI1-CFP and BRI1-YFP fusions, respectively. The AtSERK3
cDNA was PCR amplified with primers S3-F, 59-CATGCCATGGAAC-
GAAGATTAATGATC-39, and S3-R, 59-CATGCCATGGCTCTTGGACCC-
GAGGG-39, and subcloned in the NcoI site of the vectors pMON999-
CFP/YFP. All constructs were verified by sequencing. The binary
constructs Pro35S-BES1-GFP and Pro35S-bes1-GFP used for transient
assays in protoplasts were the same as described by Yin et al. (2002a).
The full-length AtSERK1-CFP and YFP fusions were described before
(Shah et al., 2001).
Transient Expression in Cowpea and Arabidopsis Protoplasts,
Protoplast Treatments, and FM4-64 Staining
Cowpea and Arabidopsis mesophyll protoplasts were prepared and
transfected as described previously by Shah et al. (2002), Sheen (2001)
After transfection, the protoplasts were incubated in protoplast medium
either overnight or for 3 h followed by addition of CHX (Sigma, St. Louis,
MO) in concentration 50 mM from 50 mM stock. The protoplasts were
used for observations until 5 h after adding the CHX. BFA (Sigma) was
added to the protoplasts or applied to BRI1-GFP–expressing Arabidop-
sis seedlings in concentration 20 mg/mL from 5 mg/mL in DMSO. The
protoplasts or the Arabidopsis roots were observed with CLSM in
intervals of 30 min after the BFA application. The BFA wash-off experi-
ments were performed as described by Geldner et al. (2001), and 24-
epibrassinolide (Sigma) was applied to the protoplasts in concentration
1 mM from 0.5 mM stock in 80% ethanol. The protoplasts were taken
for observation or FRET measurements after overnight or 1 to 2 h incuba-
tion with the 24-epibrassinolide. FM4-64 staining of the protoplasts and
the Arabidopsis roots was performed exactly as described by Shah et al.
(2002). The protoplasts were washed from the dye and incubated in
protoplast medium for 0.5, 1, 2, and 3 h.
Arabidopsis protoplaststransiently expressingBES1-GFP proteinfor3to
4 hwere used for immunoblotting analysisas previously described by Yin
et al. (2002a) using anti-BES1 antibody.
The CFP and YFP fluorescence and the FM4-64 in protoplasts and in
Arabidopsis seedlings were analyzed with the Confocal Laser Scanning
et al. (2002). In addition, the GFP was excited by the 488-nm laser line in
combination with the main dichroic 488, and GFP fluorescence was
detected by a band-pass 505- to 550-nm filter. A 403 oil immersion
objective (numerical aperture 1.3) was used for scanning. The pinhole
maximum) of 1 mm. Images and data captures were analyzed with Zeiss
LSM510 software (version 3.2).
FLIM was performed using a Bio-Rad Radiance 2100 MP system
(Hercules, CA) in combination with a Nikon TE 300 inverted microscope
(Tokyo, Japan). Two-photon excitation pulses were generated by a
Ti:Sapphire laser (Coherent Mira; Santa Clara, CA) that was pumped by
a 5-W Coherent Verdi laser. Pulse trains of 76 MHz (150 fs pulse duration,
860 nm center wavelength) were produced. The excitation light was
directly coupled into the microscope and focused into the sample using
a CFI Plan Apochromat 603 water immersion objective lens (numerical
aperture 1.2). Fluorescent light was detected using the nondescanned
single photon counting detection, which is the most sensitive solution for
two-photon imaging. For the FLIM experiment, the Hamamatsu R3809U
MCP PMT (Hamamatsu City, Japan) was used, which has a typical time
resolution ;50 ps. CFP emission was selected using a 480DF30 band-
pass filter. Images with a frame size of 64 3 64 pixels were acquired, and
s (Borst et al., 2003; Chen et al., 2003; Becker et al., 2004; Chen and
Periasamy, 2004). From the intensity images obtained, complete fluo-
rescence lifetime decays were calculated per pixel and fitted using
a double exponential decay model. The fluorescence lifetime of one
component was fixed to the value found for AtSERK1-CFP (2.5 ns) (Table
1). The FRET efficiency (E) was determined by E ¼ 1?tDA/tD, where tDis
the fluorescence lifetime of the donor in the absence of acceptor and tDA
that of the donor in the presence of acceptor at a distance (R). The
distance between the donor and the acceptor was determined from the
relation tDA¼ tD/(1 þ (R0/R)6, where R0is the Fo ¨rster radius, the distance
between the donor and acceptor at which 50% energy transfer takes
place (Elangovan et al., 2002).
We thank Joan Wellink and Jeroen Pouwels for providing us with the
STtmd-YFP Golgi marker, for growing the cowpea plants, and for
protoplast transfections; Antonie Visser and Mark Hink for help with
FRET measurements and during data analysis; and Niko Geldner and
Jir ˇı ´ Friml for helpful discussions and the critical reading of the manu-
script. J.C. is an investigator of the Howard Hughes Medical Institute.
This work was supported by Grant ERBIO4-CT96-0689 from the
European Union Biotechnology program, Grant QLG2-2000-00602
from the European Union Quality of Life and Management of Living
Recourses program, and Wageningen University, Department of Agro-
technology and Food Sciences to E.R., M.K., and S.C.D.; by the Human
Frontier Science Program Organization long-term fellowship to A.C-D.;
and grants from the USDA and the Human Frontier Science Program
Received June 23, 2004; accepted September 8, 2004.
Asamizu, E., Nakamura, Y., Sato, S., and Tabata, S. (2000). A large
scale analysis of cDNA in Arabidopsis thaliana: Generation of 12,028
Endocytosis of BRI1 and AtSERK33227
non-redundant expressed sequence tags from normalized and size-
selected cDNA libraries. DNA Res. 7, 175–180.
Balus ˇka, F., Hlavacka, A., Sˇamaj, J., Palme, K., Robinson, D.G.,
Matoh, T., McCurdy, D.W., Menzel, D., and Volkmann, D. (2002).
F-actin-dependent endocytosis of cell wall pectins in meristematic
root cells. Insights from Brefeldin A-induced compartments. Plant
Physiol. 130, 422–431.
Bastiaens, P.I.H., and Pepperkok, R. (2000). Observing proteins in their
natural habitat: The living cell. Trends Biochem. Sci. 25, 631–637.
Becker, W., Bergmann, A., Hink, M.A., Konig, K., Benndorf, K., and
Biscup, C. (2004). Fluorescence lifetime imaging by time-correlated
single-photon counting. Microsc. Res. Tech. 63, 58–66.
Birnbaum, K., Shasha, D.E., Wang, J.Y., Jung, J.W., Lambert, G.M.,
Galbraith, D.W., and Benfey, P.N. (2003). A gene expression map of
the Arabidopsis root. Science 302, 1956–1960.
Boevink, P., Oparka, K., Santa Cruz, S., Martin, B., Betteridge, A.,
and Hawes, C. (1998). Stacks on tracks: The plant Golgi apparatus
traffics on an actin/ER network. Plant J. 15, 441–447.
Borst, J.W., Hink, M., van Hoek, A., and Visser, A.J.W.G. (2003).
Multiphoton microspectroscopy in living plant cells. In Proceedings of
SPIE, Vol. 4963: Multiphoton Microscopy in the Biomedical Sciences
III, A. Periasamy and P.T. So, eds (Bellingham, WA: SPIE), pp.
Carette, J.E., Stuiver, M., Van Lent, J., Wellink, J., and Van Kammen,
A. (2000). Cowpea mosaic virus infection induces a massive pro-
liferation of endoplasmic reticulum but not Golgi membranes and is
dependent on de novo membrane synthesis. J. Virol. 74, 6556–6563.
Chen, Y., Mills, J.D., and Periasamy, A. (2003). Protein localization in
Chen, Y., and Periasamy, A. (2004). Characterization of two-photon
excitation fluorescence lifetime imaging microscopy for protein local-
ization. Microsc. Res. Tech. 63, 72–80.
Clough, S.J., and Bent, A.F. (1998). Floral dip: A simplified method for
Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant
J. 16, 735–743.
Clouse, S.D. (2002). Brassinosteroid signal transduction: Clarifying the
pathway from ligand perception to gene expression. Mol. Cell 10,
Die ´vart, A., and Clark, S.E. (2003). Using mutant alleles to determine
the structure and function of leucine-rich repeat receptor-like kinases.
Curr. Opin. Plant Biol. 6, 507–516.
Di Guglielmo, G.M., Le Roy, C., Goodfellow, A.F., and Wrana, J.L.
(2003). Distinct endocytic pathways regulate TGF-b receptor signal-
ling and turnover. Nat. Cell Biol. 5, 410–421.
Elangovan, M., Day, R.N., and Periasamy, A. (2002). Nanosecond
fluorescence resonance energy transfer-fluorescence lifetime imaging
microscopy to localize the protein interactions in a single living cell.
J. Microsc. 205, 3–14.
Friedrichsen, D.M., Joazeiro, C.A.P., Li, J., Hunter, T., and Chory, J.
(2000). Brassinosteroid-insensitive-1 is ubiquitously expressed leu-
cine-rich receptor serine/threonine kinase. Plant Physiol. 123, 1247–
Galperin, E., and Sorkin, A. (2003). Visualization of Rab5 activity in
living cells by FRET microscopy and influence of plasma-membrane-
targeted Rab5 on clathrin-dependent endocytosis. J. Cell Sci. 116,
Geldner, N. (2004). The plant endosomal system—Its structure and role
in signal transduction and plant development. Planta 219, 547–560.
Geldner, N., Anders, N., Wolters, H., Keicher, J., Kornberger, W.,
Muller, P., Delbarre, A., Ueda, T., Nakano, A., and Ju ¨rgens, G.
(2003). The Arabidopsis GNOM AFR-GEF mediates endosomal recy-
cling, auxin transport, and auxin-dependent plant growth. Cell 112,
Geldner, N., Friml, J., Stierhof, Y.-D., Jurgens, G., and Palme, K.
(2001). Auxin transport inhibitors block PIN1 cycling and vesicle
trafficking. Nature 413, 425–428.
Gonza ´les-Gaita ´n, M. (2003). Signal dispersal and transduction through
the endocytic pathway. Nat. Rev. Mol. Cell Biol. 4, 213–224.
Grebe, M., Xu, J., Mo ¨bius, W., Ueda, T., Nakano, A., Geuze, H.J.,
Rook, M.B., and Scheres, B. (2003). Arabidopsis sterol endocytosis
involves actin-mediated trafficking via ARA6-positive early endo-
somes. Curr. Biol. 13, 1378–1387.
Gruenberg, J. (2001). The endocytic pathway: A mosaic of domains.
Nat. Rev. Mol. Cell Biol. 2, 721–730.
Haj, F.G., Verveer, P.J., Squire, A., Neel, B.G., and Bastiaens, P.I.H.
(2002). Imaging sites of receptor dephosphorylation by PTP1B on the
surface of the endoplasmic reticulum. Science 295, 1708–1711.
He, Z., Wang, Z.-Y., Li, J., Zhu, Q., Lamb, C., Ronald, P., and Chory,
J. (2000). Perception of brassinosteroids by the extracellular domain
of the receptor kinase BRI1. Science 288, 2360–2363.
Hecht, V., Velle-Calzada, J.-P., Hartog, M.V., Schmidt, E.D.L.,
Boutilier, K., Grossniklaus, U., and de Vries, S.C. (2001). The
Arabidopsis SOMATIC EMBRYOGENESIS RECEPTOR KINASE 1
gene is expressed in developing ovules and embryos and enhances
embryogenic competence in culture. Plant Physiol. 127, 803–816.
Hink, M.A., Bisseling, T., and Visser, A.J.W.G. (2002). Imaging protein-
protein interactions in living cells. Plant Mol. Biol. 50, 871–883.
Horn, M.A., Heinstein, P.F., and Low, P.S. (1989). Receptor-mediated
endocytosis in plant cells. Plant Cell 1, 1003–1009.
Immink, R.G.H., Gadella, T.W.J., Ferrario, S., Busscher, M., and
Angenent, G.C. (2002). Analysis of MADS box protein-protein inter-
actions in living plant cells. Proc. Natl. Acad. Sci. USA 99, 2416–2421.
Jin, J.B., Bae, H., Kim, S.J., Jin, Y.H., Goh, C.-H., Kim, D.H., Lee, Y.J.,
Tse, Y.C., Jiang, L., and Hwang, I. (2003). The Arabidopsis dynamin-
like proteins ADL1C and ADL1E play a critical role in mitochondria
morphogenesis. Plant Cell 15, 2357–2369.
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, 1511–1525.
Ju ¨rgens, G., and Geldner, N. (2002). Protein secretion in plants: From
the trans-Golgi network to the outer space. Traffic 3, 605–613.
Katzmann, D.J., Odorizzi, G., and Emr, S.D. (2002). Receptor down-
regulation and multivesicular-body sorting. Nat. Rev. Mol. Cell Biol. 3,
Kim, D.H., Eu, Y.-J., Yoo, C.M., Kim, Y.-W., Pih, K.T., Jin, J.B., Kim,
S.J., Stenmark, H., and Hwang, I. (2001). Trafficking of phosphati-
dylinositol 3-phosphate from the trans-Golgi network to the lumen of
the central vacuole in plant cells. Plant Cell 13, 287–301.
Li, J. (2003). Brassinosteroids signal through two receptor-like kinases.
Curr. Opin. Plant Biol. 6, 494–499.
Li, J., and Chory, J. (1997). A putative leucine-rich repeat recep-
tor kinase involved in brassinosteroid signal transduction. Cell 90,
Li, J., and Nam, K.H. (2002). Regulation of brassinosteroid singling by
a GSK3/SHAGGY-like kinase. Science 295, 1299–1301.
Li, J., Wen, J., Lease, K.A., Doke, J.T., Tax, F.E., and Walker, J.C.
(2002). BAK1, an Arabidopsis LRR receptor-like protein kinase,
interacts with BRI1 and modulates brassinosteroid signaling. Cell
Mora-Garcı ´a, S., Vert, G., Yin, Y., Can ˜o-Delgado, A., Cheong, H., and
Chory, J. (2004). Nuclear protein phosphatases with Kelch-repeat
domains modulate the response to brassinosteroids in Arabidopsis.
Genes Dev. 18, 448–460.
Nam, K.H., and Li, J. (2002). BRI1/BAK1, a receptor kinase pair
mediating brassinosteroid signaling. Cell 110, 203–212.
3228The Plant Cell
Park, M., Kim, S.J., Vitale, A., and Hwang, I. (2004). Identification of Download full-text
the protein storage vacuole and protein targeting to the vacuole in leaf
cells of three plant species. Plant Physiol. 134, 625–639.
Peng, P., and Li, J. (2003). Brassinosteroid signal transduction: A mix of
conservation and novelty. J. Plant Growth Regul. 22, 298–312.
Pouwels, J., Van Der Krogt, G.N.M., Van Lent, J., Bisseling, T., and
Wellink, J. (2002). The cytoskeleton and the secretory pathway are
not involved in targeting the cowpea mosaic virus movement protein
to the cell periphery. Virology 297, 48–56.
Ritzenthaler, C., Nebenfu ¨hr, A., Movafedhi, A., Stussi-Garaud, C.,
Behnia, L., Pimpl, P., Staehelin, L.A., and Robinson, D.G. (2002).
Reevaluation of the effects of Brefeldin A on plant cells using tobacco
Bright Yellow 2 cells expressing Golgi-targeted green fluorescent
protein and COPI antisera. Plant Cell 14, 237–261.
Rutherford, S., and Moore, I. (2002). The Arabidopsis Rab GTPase
family: Another enigma variation. Curr. Opin. Plant Biol. 5, 518–528.
Sˇamaj, J., Balus ˇka, F., Voigt, B., Schlicht, M., Volkmann, D., and
Menzel, D. (2004). Endocytosis, actin cytoskeleton, and signaling.
Plant Physiol. 135, 1150–1161.
Satiat-Jeunemaitre, B., and Haves, C. (1992). Redistribution of a Golgi
gycoprotein in plant cells treated with Brefeldin A. J. Cell Sci. 103,
Shah, K., Gadella, T.W.J., van Erp, H., Hecht, V., and de Vries, S.C.
(2001). Subcellular localization and dimerization of the A. thaliana
somatic embryogenesis receptor kinase 1 protein. J. Mol. Biol. 309,
Shah, K., Russinova, E., Gadella, T.W.J., Jr., Willemse, J., and de
Vries, S.C. (2002). The Arabidopsis kinase-associated protein phos-
phatase controls internalization of the somatic embryogenesis re-
ceptor kinase 1. Genes Dev. 16, 1707–1720.
Sheen, J. (2001). Signal transduction in maize and Arabidopsis meso-
phyll protoplasts. Plant Physiol. 127, 1466–1475.
Shimada, Y., Goda, H., Nakamura, A., Takatsuto, S., Fujioka, S., and
Yoshida, S. (2003). Organ-specific expression of brassinosteroid-
biosynthetic genes and distribution of endogenous brassinosteroids
in Arabidopsis. Plant Physiol. 131, 287–297.
Sohn, E.J., Kim, E.S., Zhao, M., Kim, S.J., Kim, H., Kim, Y.-W., Lee,
Y.J., Hillmer, S., Sohn, U., Jiang, L., and Hwang, I. (2003). Rha1, an
Arabidopsis Rab5 homolog, plays a critical role in the vacuolar
trafficking of soluble cargo proteins. Plant Cell 15, 1057–1070.
Sorkin, A., McClure, M., Huang, F., and Carter, R. (2000). Interaction
of EGF receptor and grb2 in living cells visualized by fluorescence
resonance energy transfer (FRET) microscopy. Curr. Biol. 10, 1395–
Sorkin, A., and Von Zastrow, M. (2002). Signal transduction and
endocytosis: Close encounters of many kinds. Nat. Rev. Mol. Cell
Biol. 3, 600–614.
Ueda, T., and Nakano, A. (2002). Vesicular traffic: An integral part of
plant life. Curr. Opin. Plant Biol. 5, 513–517.
Ueda, T., Yamaguchi, M., Uchimiya, H., and Nakano, A. (2001). Ara6,
a plant-unique novel type Rab GTPase, functions in the endocytic
pathway of Arabidopsis thaliana. EMBO J. 20, 4730–4741.
Verveer, P.J., Wouters, F.S., Reynolds, A.R., and Bastiaens, P.I.H.
(2000). Quantitative imaging of lateral ErbB1 receptor signal propa-
gation in the plasma membrane. Science 290, 1567–1570.
Vida, T.A., and Emr, S.D. (1995). A new vital stain for visualizing
vacuolar membrane dynamics and endocytosis in yeast. J. Cell Biol.
Wang, Z.-Y., Nakano, T., Gendron, J., He, J., Chen, M., Vafeados, D.,
Yang, Y., Fujioka, S., Yoshida, S., Asami, T., and Chory, J. (2002).
Nuclear-localized BZR1 mediates brassinosteroid-induced growth
and feedback suppression of brassinosteroid biosynthesis. Dev. Cell
Wang, Z.-Y., Seto, H., Fujioka, S., Yoshida, S., and Chory, J. (2001).
BRI1 is a critical component of a plasma-membrane receptor for plant
steroids. Nature 410, 380–383.
Wiley, H.S. (2003). Trafficking of the ErbB receptors and its influence on
signaling. Exp. Cell Res. 284, 78–88.
Yin, Y., Wang, Z.-Y., Mora-Garcia, S., Li, J., Yoshida, S., Asami, T.,
and Chory, J. (2002a). BES1 accumulates in the nucleus in response
to brassinosteroids to regulate gene expression and promote stem
elongation. Cell 109, 181–191.
Yin, Y., Wu, D., and Chory, J. (2002b). Plant receptor kinases: Systemin
receptor identified. Proc. Natl. Acad. Sci. USA 99, 9090–9092.
Zhao, J., Peng, P., Schmitz, R.J., Decker, A.D., Tax, F.E., and Li, J.
(2002). Two putative BIN2 substrates are nuclear components of
brassinosteroids signaling. Plant Physiol. 130, 1221–1229.
Endocytosis of BRI1 and AtSERK33229