Full text of DOI:10.1006/jmbi.2001.4706

doi:10.1006/jmbi.2001.4706 available online at http://www.idealibrary.com on
J. Mol. Biol. (2001) 309, 641±655
Subcellular Localization and Oligomerization of the
Arabidopsis thaliana Somatic Embryogenesis
Receptor Kinase 1 Protein
Khalid Shah, Theodorus W. J. Gadella Jr, Harrie van Erp, ValeÂrie Hecht
and Sacco C. de Vries*
Laboratory of Molecular
Biology, Department of Plant
Sciences, Wageningen
University and Research
Centre, Wageningen, The
Netherlands
The Arabidopsis thaliana somatic embryogenesis receptor kinase 1 (AtSERK1)
gene is expressed in developing ovules and early embryos. AtSERK1 is
also transiently expressed during somatic embryogenesis. The predicted
AtSERK1 protein contains an extracellular domain with a leucine zipper
motif followed by ®ve leucine-rich repeats, a proline-rich region, a single
transmembrane region and an intracellular kinase domain. The AtSERK1
cDNA was fused to two different variants of green ¯uorescent protein
(GFP), a yellow-emitting GFP (YFP) and a cyan-emitting GFP (CFP), and
transiently expressed in both plant protoplasts and insect cells. Using
confocal laser scanning microscopy it was determined that the AtSERK1-
YFP fusion protein is targeted to plasma membranes in both plant and
animal cells. The extracellular leucine-rich repeats, and in particular the
N-linked oligosaccharides that are present on them appear to be essential
for correct localization of the AtSERK1-YFP protein. The potential for
dimerization of the AtSERK1 protein was investigated by measuring the
YFP/CFP ¯uorescence emission ratio using ¯uorescence spectral imaging
microscopy. This ratio will increase due to ¯uorescence resonance energy
transfer if the AtSERK1-CFP and AtSERK1-YFP fusion proteins interact.
In 15 % of the cells the YFP/CFP emission ratio for plasma membrane
localized AtSERK1 proteins was enhanced. Yeast-protein interaction
experiments con®rmed the possibility for AtSERK1 homodimerization.
Elimination of the extracellular leucine zipper domain reduced the YFP/
CFP emission ratio to control levels indicating that without the leucine
zipper domain AtSERK1 is monomeric.
# 2001 Academic Press
Keywords: AtSERK1; green ¯uorescent protein; glycosylation;
*Corresponding author
oligomerization; ¯uorescence resonance energy transfer
Introduction
Abbreviations used: AtSERK1, Arabidopsis thaliana
somatic embryogenesis receptor kinase 1; LZ, leucine
zipper; LRR, leucine-rich repeat; GFP, green ¯uorescent
protein; YFP, yellow-emitting GFP; CFP, cyan-emitting
GFP; CLSM, confocal laser scanning microscopy; FSPIM,
¯uorescence spectral imaging microscopy; FRET,
¯uorescence resonance energy transfer; RLK, receptor-
like kinase; TGF-b, transforming growth factor b; EGF,
epidermal growth factor; SRK, S-locus receptor kinase;
CLV, CLAVATA; BRI1, Brassinosteroid insensitive 1;
SERK, somatic embryogenesis receptor kinase gene; IL1,
interleukin 1; FCS, ¯uoresence correlation spectroscopy;
AD, activation domain, BD, binding domain; 2,4-D, 2,4-
dichlorophenaxy acetic acid; EGFR, epidermal growth
factor receptor; PDGFR, platelet derived growth factor
receptor.
Receptor kinases have been shown to play a key
role in developmental processes such as cell pro-
liferation, migration and differentiation. Receptor
kinase-mediated signal transduction pathways are
highly conserved in both vertebrates and
invertebrates.1,2 In plants, similar receptors, desig-
nated as receptor-like kinases (RLKs) also appear
to play a crucial role in development.3
The classical model for receptor kinase-activation
involves ligand binding-induced dimerization of
the receptor, resulting in autophosphorylation of
both partners in the dimer, mainly on speci®c tyro-
sine, serine or threonine residues. The best known
examples of this mode of action are the transform-
ing growth factor (TGF-b) receptor and the epider-
E-mail address of the corresponding author:
Sacco.deVries@mac.mb.wau.nl
0022-2836/01/030641±15 $35.00/0
# 2001 Academic Press

642
AtSERK1 Localization and Oligomerization
mal growth factor (EGF) receptor.4 Ligand-induced
dimerization of cell surface receptors can bring
into proximity downstream proteins associated
with these receptors followed by transphosphory-
lation of such target proteins. Animal receptor
kinases can dimerize either by ligand-dependent
non-covalent dimerization5 or by ligand-indepen-
dent covalent dimerization.6 Dimerization can
involve two identical receptors (homodimeriza-
tion); different members of the same receptor
family or a receptor and an accessory protein
(heterodimerization).7 Receptor activation through
heterodimerization is the most common,4 but cer-
tain receptors such as those for growth hormones
employ homodimerization.
The evidence for monomer to dimer transition as
a mechanism for receptor activation derives from
experiments demonstrating that ligand binding
leads to receptor dimerization.4 Second, arti®cially
induced dimerization or naturally occurring
mutations leading to dimerization both mimic sig-
naling in the absence of the physiological ligand.8,9
Finally, oligomerization of intracellular regions of
receptors using cell permeable synthetic ligands
can lead to dimerization.10 However, one of the
most extensively studied eukaryotic receptors, the
insulin receptor, exists as an inactive (ab) dimer
maintained by intersubunit disul®de cross bridge-
s.11
The mechanisms by which plant receptors trans-
duce signals across the cell surface are largely
unknown but plant receptors may also dimerize.12
Studies on the action of the S-locus receptor kinase
(SRK), involved in the self-incompatibility response
in Brassica, have shown that recombinant SRK
autophosphorylates constitutively and exists as
dimers in nonpollinated stigmas.13 The CLAVATA
(CLV) genes control cell proliferation in the Arabi-
dopsis shoot and in¯orescence meristems. It has
been shown that the CLV3 peptide acts as a ligand
in the CLV1/CLV2 receptor complex.14±16 Recent
chimeric studies on RLKs, using the extracellular
domain of brassinosteroid insensitive1 (BRI1) pro-
tein, involved in the perception of brassinosteroids
and the intracellular domain of the Xa21 protein,
involved in pathogen resistance from rice have
revealed that the use of chimeric receptors may
also highly useful in plants.17
Transmembrane receptor proteins, like other cell
surface proteins undergo post-translational modi®-
cations such as disul®de bonding, glycosylation,
palmitoylation and proteolytic cleavage. Many of
these modi®cations play decisive roles in protein
maturation and/or intracellular traf®cking.18 In
mammalian cells, the correct folding and oligomer-
ization of complex, multi-domain secretory pro-
teins is controlled by N-glycosylation.19 For
example, the human insulin receptor has several
glycosylation sites close to the N terminus of the
receptor. Mutational analysis of these sites showed
that glycosylation was essential for correct proces-
sing and targeting of the protein to the cell
surface.20 In plant RLKs such studies have not
been conducted, but the predicted protein
sequence of plant RLKs shows that similar
domains close to the N terminus contain N-linked
glycosylation sites.
Somatic embryogenesis is the process whereby
somatic cells can develop into plants via character-
istic morphological stages. In cultured Daucus
carota suspension cells, the SERK (somatic embryo-
genesis receptor kinase) gene, a leucine-rich repeat
(LRR)RLK, was found to be a marker for single
suspension cells capable of forming embryos.21
During somatic and zygotic embryogenesis SERK
expression continues up to the early globular stage
and is absent in later stages of embryo develop-
ment. The most closely related Arabidopsis thaliana
SERK1 gene is expressed in developing ovules,
early embryos and in vascular tissues of seedlings
(V. Hecht et al., unpublished results). The predicted
AtSERK1 protein contains an N-terminal leucine
zipper (LZ), ®ve LRRs and a proline-rich region in
its predicted extracellular domain. This is followed
by a single transmembrane domain that is attached
to an intracellular kinase domain with all 11 con-
served subdomains as found in serine/threonine
kinases (V. Hecht et al., unpublished results).
Based on the jelly ®sh green ¯uorescent protein
(GFP) and confocal microscopy, spatiotemporal
dynamics of GFP-fusion proteins in intact cells can
be investigated.22±24 We have used AtSERK1-GFP
fusion proteins to determine the subcellular localiz-
ation of this receptor. In addition, dimerization of
the chimeric proteins was investigated making use
of ¯uorescence resonance energy transfer (FRET)
by employing spectroscopic variants of GFP. FRET
is a quantum-mechanical process by which the
excitation energy is transferred from a donor ¯uor-
ochrome to an appropriate acceptor ¯uorochrome.
FRET only occurs when: (i) the donor ¯uorophore
emission spectrum overlaps with the absorption
spectrum of the acceptor, (ii) the transition dipole
moments of the donor and the acceptor are not
perpendicularly oriented, and (iii) the distance
between the donor and the acceptor is less than
1.5Â the FoÈrster radius (R0) for energy transfer.25,26
Different engineered spectral variants of GFP, exhi-
biting yellow or cyan ¯uorescence, (YFP and CFP,
respectively) provide an opportunity to label the
same protein with two different ¯uorophores in
the same living cell.27 More importantly, the emis-
sion and the excitation spectrum of these two var-
iants of GFP are suitable for FRET. This allows to
probe proximity of proteins in intact living cells.28
The R0 for CFP/YFP FRET is 5.5 nm implying that
FRET between CFP and YFP will only be detected
when the proximity of the ¯uorophores is less than
10 nm. FRET is manifested in different ways: a
decrease in donor ¯uorescence quantum yield, a
decreased donor ¯uorescence lifetime, an increased
stability of the donor towards chemical photo-
bleaching29 and an increased sensitized acceptor
¯uorescence emission, if the acceptor is
a
¯uorophore.30,31 FRET microscopy28 has been suc-
cessfully employed to investigate the assembly of

AtSERK1 Localization and Oligomerization
643
the T-cell antigen receptor complex,32,33 the inter-
leukin-1 (IL-1) receptor complex after IL-1
tor is a true plasma membrane receptor with
cysteine-rich repeats in the extracellular domain.
The correspondence in the location of ¯uorescence
between the AtSERK1 and the chimeric EGFR-
AtSERK1 receptor strongly suggests that AtSERK1
is a plasma membrane-localized protein.
binding34
oligomerization.35,36
and
to
study
EGF
receptor
The results presented here show that the
AtSERK1 protein is localized in the plasma mem-
brane, mainly in the form of monomers. Only a
small percentage of the AtSERK1 protein may exist
as oligomers (e.g. homodimers), suggesting that
the AtSERK1 protein belongs to a class of receptors
that can undergo ligand-dependent non-covalent
homodimerization. Employing truncated AtSERK1
proteins, it was found that the LRR domain
is involved in correct targeting of the receptor
while the LZ domain is involved in receptor
dimerization.
To con®rm the AtSERK1 protein localization in
the plasma membrane, the AtSERK1ins-YFP-CFP
fusion construct (Figure 1(y)) was expressed in
Spodoptera frugiperda (Sf21) insect cells in order to
perform ¯uorescence correlation spectroscopy
(FCS). A line-scan along the optical axis (z-axis)
was made to determine the distribution of ¯uor-
escent proteins in the cell. Detection of ¯uorescent
fusion proteins by the sensitive, single molecule
detection, FCS technique is more convenient in
insect cells than in plant protoplasts because of the
absence of background chlorophyll auto-
¯uorescence. The AtSERK1ins-CFP-YFP expressing
cells showed the same low intensity of 20 kHz
(photoelectrons emitted per second) in the nuclear
region as in the non-infected cells. There was
a three- to fourfold (161 kHz) increase in the
¯uorescence intensity in the plasma membrane
region of AtSERK1ins-CFP-YFP expressing cells
(Figure 3(f)) as compared to highest ¯uorescence
intensity of around 40 kHz in the plasma mem-
brane of the non-infected Sf21 cells (Figure 3(e)).
This ®nding indicates that the AtSERK1ins-CFP-
YFP proteins are also correctly localized in the
plasma membrane in insect cells. In cells expres-
sing the fusion proteins to a very high level, also
endomembranes were labelled, possibly re¯ecting
proteins in transit or undergoing endocytosis (data
not shown).
To determine whether the extracellular LZ
domain and the different LRRs of the AtSERK1
protein have a role in targeting of the AtSERK1
receptor we transfected protoplasts with a series of
constructs in which the LZ domain or one or sev-
eral LRRs were removed (see Figure 1(b), (f), (h),
(j)-(m) for an overview of the constructs used). Pro-
toplasts expressing the AtSERK1ÁLZ-YFP protein
showed the same surface ¯uorescence as shown by
protoplasts expressing AtSERK1-YFP (Figure 4(a),
compare with Figure 3(a)). No effect of the removal
of the ®rst LRR was evident after transfecting
AtSERK1ÁLRR1-YFP into protoplasts (Figure 4(b)).
There was some decrease in surface ¯uorescence
when AtSERK1ÁLRR1,2-YFP (Figure 4(c)) or
AtSERK1ÁLRR1,2,3-YFP (Figure 4(d)) were trans-
fected into protoplasts. In addition to the plasma
membrane, the truncated AtSERK1 proteins were
also seen in subcellular organelles. There was a
considerable decrease in surface ¯uorescence when
AtSERK1ÁLRR1,2,3,4-YFP was transfected into proto-
plasts (Figure 4(e)). Most of the AtSERK1 protein
was seen to be concentrated in subcellular
organelles which might represent clogging in ER
or Golgi-derived vesicles or sequestering of the
AtSERK1 protein in the lysosomal compartments
of the protoplasts (Figure 4(e)). Upon transfection
of AtSERK1ÁLRR-YFP into protoplasts, almost no
Results
AtSERK1-YFP/CFP fusion proteins
Here, several AtSERK1-YFP/CFP constructs
were used. These constructs are shown in detail in
Figure 1. In order to determine the localization of
the AtSERK1 protein, several AtSERK1-YFP con-
structs (Figure 1(b), (d), (f) and (h)) were trans-
fected into cowpea mesophyll protoplasts. After
incubation for 12 hours, soluble and pellet fractions
were isolated, separated by denaturing SDS-PAGE
and immunoblotted using an antiserum raised
against GFP (Figure 2). The results show that
AtSERK1-YFP wild-type, AtSERK1ÁLZ-YFP and
AtSERK1ÁLRR-YFP proteins are all found exclu-
sively in the pellet fraction. Upon deletion of the
entire extracellular domain and the transmembrane
domain (Figure 1(d)), about 40 % of the protein
becomes soluble, while the remainder is present in
the pellet fraction (Figure 2). We conclude that the
AtSERK1-YFP fusion proteins are correctly
expressed and that the majority are present in the
insoluble pellet fractions that may represent the
membrane fraction.
Localization and targeting of AtSERK1 protein
To determine whether AtSERK1-YFP fusion pro-
teins were localized in the plasma membrane,
viable transfected protoplasts were analyzed by
confocal laser scanning microscopy (CSLM) 12
hours after transfection. Protoplasts transfected
with an AtSERK1-YFP construct (Figure 1(b))
clearly showed ¯uorescence at the surface of the
protoplasts (Figure 3(a)). Endomembranes were
labeled to a lesser extent, possibly re¯ecting pro-
teins in transit. In contrast, cytoplasmic expression
was observed when the pMON999-YFP vector
alone (Figure 3(c)) or the AtSERK1kin-YFP construct
(Figure 1(d)) was transfected into protoplasts
(Figure 3(b)). The EGFRex-AtSERK1kin-YFP con-
struct (Figure 1(n)), was also expressed in proto-
plasts and the chimeric protein showed the same
predominant surface localization (Figure 3(d)) as
that of the AtSERK1-YFP protein. The EGF recep-

644
AtSERK1 Localization and Oligomerization
Figure 1 (legend shown opposite)
surface ¯uorescence was observed (Figure 4(f)).
Most of the fusion protein was seen internally as in
Figure 4(e). We conclude that the LZ domain of
AtSERK1 is not involved in protein targeting but
that at least four LRRs of AtSERK1 contain speci®c
information for correct targeting of the AtSERK1
protein.
The AtSERK1 protein contains seven predicted
N-glycosylation sites, ®ve of which are present in
the LRRs of the extracellular domain (Figure 1(a)).
To determine whether these sites are used in the
proper targeting of the AtSERK1 protein, we trea-
ted the protoplasts with 20 mg ml tunicamycin
1
three hours after transfection with AtSERK1-YFP.
Tunicamycin prevents the addition of N-acetylglu-
cosamine to dolichol phosphate, the ®rst step in
the formation of the core oligosaccharide deployed
in N-linked glycosylation. The transfected proto-
plasts were visualized with CSLM 12 hours after
the transfection. The localization of AtSERK1-YFP
in tunicamycin-treated protoplasts (Figure 5(b))
showed reduced membrane targeting of the

AtSERK1 Localization and Oligomerization
645
camycin also showed intracellular clogging in
vesicles (results not shown) comparable to tunica-
mycin-treated plant protoplasts. Proteins from
tunicamycin-treated and untreated AtSERKins-CFP-
YFP expressing insect cells were isolated, separated
by denaturing SDS-PAGE and immunoblotted
using an antiserum against GFP. There was a
reduction of the molecular weight of the AtSER-
K1ins-CFP-YFP fusion protein from 96 kD to 84 kD
in the tunicamycin treated cell extract when com-
pared to the untreated AtSERK1ins-YFP-CFP
infected Sf21 cells (Figure 5(b)). These results
suggest that AtSERK1 is glycosylated in both insect
cells and cowpea protoplasts and that N-linked
glycosylation may be important for targeting the
AtSERK1 protein to the plasma membrane.
Homodimerization of AtSERK1 protein:
interaction in yeast
To determine if the AtSERK1 receptor was
capable of forming homodimers, the AtSERK1
intracellular kinase domain (AtSERK1(266-625)), the
Figure 2. Immunoblot of AtSERK1 fusion proteins in
transfected protoplasts. The protoplasts were harvested
after 12 hours of transfection with AtSERK1-YFP trans-
genes, lysed and separated on a reducing SDS-10 %
PAGE. Proteins transferred to nitrocellulose paper were
immunoblotted with anti-GFP antibody. S and P denote
the soluble cytosolic fractions and the insoluble pellet
fractions respectively. Lanes 1 and 2: AtSERK1-YFP;
extracellular domain (AtSERK1(26-234)
) and the
AtSERK1 protein without signal sequence
(AtSERK1(31-625)) were fused to the B42 activation
domain (AD) in pJG4-5 and the lexA DNA-binding
domain (BD) in pEG202 (Figure (1)(v, x)). There
was no interaction between the two intracellular
kinase domains of AtSERK1 (AtSERK1(266-625) in
Figure 6) whereas the AtSERK1 extracellular do-
mains do interact with each other (AtSERK1(26-234)
in Figure 6) as indicated by the b-galactosidase
activity and the growth without leucine. The inter-
action, between two full-length AtSERK1 proteins
lacking the N-terminal signal sequence was slightly
reduced compared to the interaction shown by the
extracellular domains only (AtSERK1(31-625) in
Figure 6). This is most likely due to aberrant fold-
ing of the full-length membrane receptor protein in
the yeast nucleus. From these data we conclude
that the AtSERK1 extracellular domains are
lanes
3 and 4: AtSERK1ÁLZ-YFP; lane 5 and 6:
AtSERK1ÁLRR-YFP; lane 7 and 8: AtSERK1kin-YFP. The
molecular mass of the proteins is indicated in kDa.
AtSERK1 protein and increased intracellular clog-
ging similar to that observed with the truncated
AtSERK1ÁLRR-YFP protein (Figure 4(e) and (f))
when compared with untreated protoplasts
(Figure 5(a)). Pretreatment of AtSERKins-CFP-YFP
1
expressing insect cells with 20 mg ml
of tuni-
Figure 1. AtSERK1 constructs used in this study. Different domains of (a) AtSERK1 and (b)-(q) AtSERK1 fusion
constructs used for expression in plant protoplast, (r)-(x) yeast two hybrid studies and (y) expression in Sf21 insect
cells are shown. The AtSERK1-YFP/CFP transgenes were constructed in the PMON999 vector under control of the
35 S promoter. (b) and (c) AtSERK1-YFP/CFP contains cDNA encoding for the full-length AtSERK1 as a C-terminal
fusion with the YFP/CFP genes. (d) and (e) AtSERK1kin-YFP/CFP encodes only the intracellular kinase domain of
AtSERK1 fused to YFP/CFP. The truncated constructs, (f) and (g) AtSERK1ÁLZ-YFP/CFP and (h) and (i)
AtSERK1ÁLRR-YFP/CFP were constructed by removing the leucine zipper and the ®ve extracellular LRRs respect-
ively from the coding sequence of ArSERK1. A series of LRR truncated constructs was made by removing LRRs start-
ing from the N-terminal end and fusing the rest of the coding sequence of AtSERK1 to the YFP. This resulted in the
1
construction of (j) AtSERK1ÁLRR -YFP, (k) AtSERK1ÁLRR1,2-YFP, (l) AtSERK1ÁLRR1,2,3-YFP, and (m)
AtSERK1ÁLRR1,2,3,4-YFP. (n) and (o) EGFR3x-AtSERK1kin-YFP/CFP chimeras were constructed by fusing the extracellu-
lar domain of EGFR and the intracellular kinase domain of AtSERK1-YFP/CFP. A single mutation in the AtSERK1
extracellular domain just above the transmembrane domain resulted in (p) and (q) AtSERK1Y230C-YFP/CFP con-
structs. An AtSERK1 (amino acid residues 31 to 625) construct lacking only the N-terminal signal sequence, and
AtSERK1 constructs consisting of the extra-cellular domain or the intracellular domain only were fused to the acti-
vation domain (AD: pJG 4-5) or to the LexA DNA-binding protein (BD: pEG202) resulted in (r) and (s) AtSERK1(31-
625)-BD/AD, (t) and (u) AtSERK1(26-234)-BD/AD, (v) and (x) AtSERK1(266-625)-BD/AD. For expression in insect cells the
full length AtSERK1 fused to YFP and the full-length AtSERK1 fused to CFP were cloned in front to the viral p10
and polyhedrin promoters respectively resulting in the (y) AtSERKins-CFP/YFP.

646
AtSERK1 Localization and Oligomerization
Figure 4. Confocal ¯uorescence images of protoplasts
transfected with LZ and the series of LRR truncated
constructs: (a) AtSERK1ÁLZ-YFP, (b) AtSERK1ÁLRR1-
YFP, (c) AtSERK1ÁLRR1,2-YFP, (d) AtSERK1ÁLRR1,2,3-YFP
and (e) AtSERK1ÁLRR1,2,3,4-YFP and (f) AtSERK1ÁLRR-
YFP expression is shown. In green the YFP ¯uorescence
is shown and in red the chlorophyll ¯uorescence is
shown.
Figure 3. Confocal ¯uorescence images of protoplasts
transfected with. (a) AtSERK1-YFP, (b) PMON999-YFP,
(c) AtSERK1kin-YFP, (d) EGFRex-AtSERK1kin-YFP. In
green the YFP ¯uorescence is shown and in red the
chlorophyll ¯uorescence is shown. (e) and (f) Fluor-
escence correlation spectroscopy (FCS) of AtSERK1-YFP-
CFP fusion proteins in insect cells. The pro®les show the
count rate along the optical z-axis of the (e) non-infected
cells and the (f) AtSERK1-YFP expressing Sf21 cells. N
and P indicate the ¯uorescence recorded in the nucleus
and the plasma membrane, respectively.
as a detection system. Spectral images were taken
from plasma membrane-regions of protoplasts
coexpressing AtSERK1-YFP and AtSERK1-CFP and
a ¯uorescence emission spectrum corrected for
background ¯uorescence was then generated. In
case of FRET, the CFP ¯uorescence will be
quenched and the YFP ¯uorescence would be
increased (sensitized). Assuming approximately
equal expression levels of both AtSERK1-CFP
(Figure 1(c)) and -YFP (Figure 1(d)), we consider a
ratio of ¯uorescence intensity at 525 nm over
475 nm (designated as the YFP/CFP emission
ratio) of 1.3 or higher as evidence of FRET and
hence of AtSERK1-AtSERK1 interactions such as
oligomerization or homodimerization. It should be
emphasized that with the FSPIM technique it can-
not be resolved whether the nonmonomeric
AtSERK1-receptors are dimeric, trimeric or present
as higher oligomeric structures.
capable of homodimerization while the kinase
domains are unlikely to interact physically.
Oligomerization of the AtSERK1 protein:
fluorescence energy resonance transfer
between CFP and YFP
Because protein-protein interactions demon-
strated in the yeast system may not be representa-
tive of interactions in living plant cells, nor for
interactions in membranes, we used cowpea proto-
plasts coexpressing AtSERK1-CFP and AtSERK1-
YFP fusion proteins. Possible AtSERK1 dimeriza-
tion was then studied by measuring FRET between
the CFP (donor) and the YFP (acceptor) ¯uoro-
phores. All experiments were based on ¯uor-
escence spectral imaging microscopy (FSPIM)
A single protoplast coexpressing AtSERK1-CFP
and AtSERK1-YFP fusion proteins is shown in
Figure 7(a). Five areas of which spectral images are
recorded are indicated by rectangles. The YFP/
CFP emission ratios deduced from these spectra is
close to 1.0 in four of the ®ve measurements and

AtSERK1 Localization and Oligomerization
647
Figure 5. The effect of tunicamycin on protoplasts and
Sf21 cells expressing AtSERK1-YFP. (a) Protoplasts
transfected with AtSERK1-YFP and visualized after 12
hours of transfection. (b) Protoplasts treated with 20 mg
ml of tunicamycin three hours after transfection with
AtSERK1-YFP and visualized after 12 hours of transfec-
tion. (c) Immunoblot analysis of the Sf21 insect cells,
Figure 6. AtSERK1-AtSERK1 interactions in yeast
using lexA two-hybrid vectors. The AtSERK1(31-625)
,
AtSERK1(26-234), AtSERK1(266-625) fused to the activation
domain (AD: pJG 4-5) or with the LexA DNA binding
protein (BD: pEG202) were transformed into yeast strain
EGY 48 and b-galactosidase activity and growth on leu-
cine was determined.
1
1
untreated (lane 1) and treated with either 10 mg ml
1
(lane 2) or 20 mg ml (lane 3) of tunicamycin prior to
AtSERK1ins-CFP-YFP infection, harvested after 48 hours,
lysed and separated on reducing SDS-8 % PAGE. Pro-
teins were transferred to nitrocellulose paper and immu-
noblotted with anti-GFP antibody. The molecular mass
of the proteins is indicated in kDa.
measurements. Similar results were obtained when
single spectral measurements were performed on
60 protoplasts coexpressing AtSERK1kin-CFP and
AtSERK1kin-YFP (Figure 7(e)). Collectively, these
results show that in the protoplasts analyzed, most
of the AtSERK1 receptor molecule (ꢀ85 %) are in
the monomeric state while up to 15 % may be oli-
gomerized. No interaction was found between the
kinase domains expressed in the cytoplasm, which
con®rms the yeast two-hybrid experiments.
The YFP/CFP emission ratios as shown in
Figure 7 may suffer from several potential artifacts,
an unbalanced expression ratio of the donor and
the acceptor ¯uorophores or the absorption of the
CFP ¯uorescence by chlorophyll in plant cells. To
exclude the ®rst possibility, we performed more
than ten independent transfection experiments in
which we coexpressed AtSERK1-CFP and
AtSERK1-YFP fusion proteins. In each of these
transfections 80 % of the protoplasts exhibited a
constant YFP/CFP emission ratio indicating a con-
stant expression ratio. In addition, the variation in
the YFP/CFP emission ratio within the same cell
(in 15 % of the cells) indicates that another
phenomenon than an unbalanced expression ratio
must be responsible for enhanced YFP/CFP emis-
sion ratio (i.e. FRET). To avoid the second artifact,
FSPIM measurements were performed exclusively
over areas lacking chloroplasts. We therefore con-
sider it likely that the results showing an increased
donor/acceptor ratio are due to FRET between the
CFP and the YFP molecules fused to the AtSERK1
1.35 in only one measurement (Figure 7(c)).
Repeating this experiment in ten different proto-
plasts and taking ®ve measurements from each
protoplast revealed that the YFP/CFP emission
ratio is close to 1.0 in about 80 % of the spectral
measurements and 1.35 in 15 % of the measure-
ments. This indicates that receptor oligomerization
is not uniformly distributed within the membrane
of a single protoplast. The percentage of YFP/CFP
emission ratios deduced from single spectral
measurements on 60 protoplasts co-expressing
AtSERK1-CFP and AtSERK1-YFP are shown in
Figure 7(e). About 15 % of the measurements show
an increased YFP/CFP emission ratio of 1.35,
which is similar to the results obtained from a
single protoplast (Figure 7(c)). The protoplasts co-
expressing the intracellular AtSERK1kin-CFP
(Figure 1(e)) and AtSERK1kin-YFP (Figure 1(d) and
Figure 7(b)) showed no or little increase in YFP
(acceptor) ¯uorescence (Figure 7(d)). The YFP/CFP
emission ratio was close to 1.0 in all ®ve measure-
ments was recorded from the different points
(shown by rectangles) in the protoplast
(Figure 7(b)). Repeating this experiment in ten
different protoplasts and taking ®ve measurements
from each protoplast revealed a YFP/CFP emission
ratio close to 1.0 in about 95 % of the spectral

648
AtSERK1 Localization and Oligomerization
Figure 7. FSPIM analysis of ¯uorescent AtSERK1
fusion proteins. (a) Confocal image of a protoplast
cotransfected with AtSERK1-YFP/CFP showing typical
regions (rectangles) used for spectral measurements. (c)
Emission spectra of the AtSERK1-CFP/YFP proteins
obtained from the spectral images shown in (a). The x-
axis represents wavelength (nm) and the y-axis rep-
resents the CFP and YFP ¯uorescence intensities. (b)
and (d) The same for AtSERK1kin-CFP/YFP, expressing
protoplasts. (e) Comparison of the 525/475 nm ¯uor-
escence emission ratio (the YFP/CFP ratio), varying
form 1.00 to 1.35, of the 60 protoplasts cotransfected
with AtSERK1-CFP/YFP, and the protoplasts cotrans-
fected with AtSERK1kin-CFP/YFP.
Figure 8. FSPIM analysis of truncated ¯uorescent
AtSERK1 fusion proteins. (a) Confocal image of a proto-
plast cotransfected with AtSERK1ÁLZ-CFP/YFP show-
ing typical regions (rectangles) used for spectral
measurements. (c) Emission spectra of the AtSERK1ÁLZ-
CFP/YFP proteins obtained from spectral images shown
in (a). The x-axis represents wavelength (nm) and the y-
axis represents the CFP and YFP ¯uorescence intensities.
(b) Confocal image of a protoplast cotransfected with
AtSERK1ÁLRR-CFP/YFP showing typical plasmamem-
brane regions (rectangles) and intracellular regions
(small rectangles) used for spectral measurements and
(d) emission spectra of the AtSERK1ÁLRR-CFP/YFP pro-
teins obtained from plasma membrane regions, (e) Emis-
sion spectra of the AtSERK1ÁLRR-CFP/YFP proteins
obtained from intracellular regions. (f) Comparison of
the ¯uorescence emission ratio (YFP/CFP ratio) on the
plasma membrane, varying from 1.00 to 1.35, of the 60
protoplasts cotransfected with AtSERK1ÁLZ-CFP/YFP,
and the protoplasts cotransfected with AtSERK1ÁLRR-
CFP/YFP.
protein, indicating receptor-oligomerization (e.g.
homodimerization).
To verify whether AtSERK1 receptor oligomeri-
zation was dependent on the extracellular
domains, protoplasts were co-transfected with
AtSERK1ÁLZ-YFP (Figure 1(f)) and AtSERK1ÁLZ-
CFP (Figure 1(g)) (Figure 8(a)). Experiments were
performed in the same way as described for the
full-length AtSERK1 receptor as shown in Figure 7.
The spectra obtained from the area of the
protoplast expressing AtSERK1ÁLZ-YFP and

AtSERK1 Localization and Oligomerization
649
AtSERK1ÁLZ-CFP (Figure 8(a)) at the plasma
membrane (indicated by rectangles) showed that
there was no increase in YFP ¯uorescence in
almost all the measurements recorded (Figure 8(c)).
Repeating this experiment in ten different proto-
plasts taking ®ve measurements from each proto-
plast revealed that the intensity ratio is close to 1.0
in about 95 % of the spectral measurements. The
same results were obtained when single spectral
measurements were performed on 60 co-expressing
AtSERK1ÁLZ-CFP (Figure 1(i)) and AtSERK1ÁLZ-
YFP protoplasts (Figure 8(f)). Next, protoplasts co-
expressing AtSERK1ÁLRR-CFP and AtSERK1ÁLRR-
YFP (Figure 1(h) and Figure 8(b)) were analyzed.
The FSPIM measurements on protoplast expressing
these truncated AtSERK1 receptors were per-
formed on both types of receptor molecules, the
ones that were properly targeted to the plasma
membrane (shown by rectangles) and the ones that
remained clogged in the intracellular vesicles
(shown by small rectangles) (Figure 8(b)). The
spectra derived from the ®ve measurements on
areas indicated by small rectangles showed an
increased ¯uorescence ratio of 1.35 in three to four
out of ®ve measurements (Figure 8(d)). When this
experiment was repeated in ten different proto-
plasts, there was an increase YFP/CFP of 1.35 in
70 % of the measurements recorded. A YFP/CFP
¯uorescence intensity ratio close to 1.35 was seen
in 1 out of 5 measurements recorded from the
different points indicated by rectangles at the plas-
ma membrane of the protoplast. Repeating this
experiment in 10 different protoplasts showed that
about 15 % of the measurements had an increased
ratio of 1.35 (Figure 8(e)). When these experiments
were performed at the plasma membrane of 60
different protoplasts taking one single measure-
ment from each protoplast, the results coincided
with the measurements shown for each individual
protoplast (Figure 8(f)). The results show that there
was an increased YFP/CFP ¯uorescence intensity
ratio again in about 15 % of the coexpressing proto-
plasts suggesting that LRRs are not involved at all
in AtSERK1 dimerization at the plasma membrane.
To determine whether AtSERK1 receptor-dimeri-
zation at the plasma membrane could be increased
by an intermolecular disul®de bond, we intro-
duced a single unpaired cysteine in the extracellu-
lar domain in the region immediately adjacent
to the transmembrane domain (Y230C). Proto-
plasts cotransfected with AtSERK1Y230C-YFP and
AtSERK1Y230C-CFP constructs (Figure 1(p) and (q))
showed protein localization similar to that of
AtSERK1-YFP and AtSERK1-CFP (results not
shown). The spectra from these protoplasts were
obtained in the same way as for AtSERK1-YFP and
AtSERK1-CFP. In about 15 % of the measurements,
the YFP/CFP ratio increased to 1.35 (Figure 9).
Since this is similar to the ratio observed in full
length AtSERK1 proteins, the introduction of an
extra cysteine residue does not appear to result in
increased AtSERK1 receptor dimerization.
Figure 9. FSPIM analysis of truncated ¯uorescent
AtSERK1 fusion proteins. Emission spectra of the
AtSERK1Y230C-CFP/YFP proteins after excitation at
430 nm.
The synthetic auxin 2,4-dichlorophenoxy acetic
acid (2,4-D) is commonly used to initiate somatic
embryogenesis. Because of the correspondence
between the acquistion of embryogenic potential
and SERK expression21 we investigated if 2,4-D
itself can in¯uence AtSERK1 dimerization. Proto-
plasts were cotransfected with AtSERK-CFP and
AtSERK-YFP and were treated with 2,4-D. There
was no change in the YFP/CFP emission ratio at
the plasma membrane (data not shown).
Taken together, we conclude from these results
that the predominant form of AtSERK1 is mono-
meric and that up to 15 % of the AtSERK1-CFP/
YFP proteins may be oligomerized (possibly as
homodimers) at the plasma membrane of the pro-
toplasts. The oligomerization appears to require
only the extracellular LZ domain and not the LRRs
or the kinase domain of AtSERK1.
Discussion
The major goals of this work were to determine
the subcellular localization of the AtSERK1 protein
and to determine whether the AtSERK1 protein
was capable of homodimerization in vivo. The
results show that AtSERK1 is a plasma membrane
protein that occurs predominantly as a monomer.
In living plant cells, less than 15 % of the receptor
molecules may exist as a homodimer.
Based on the cDNA sequence, the AtSERK1 pro-
tein was predicted to be a membrane associated
serine/threonine receptor protein kinase (V. Hecht
et al., unpublished results). This prediction was
supported by several experiments. Western blot
analysis showed that AtSERK1-YFP fusion proteins
produced in plant protoplasts are present in a
40,000 g pelletable cell fraction whereas similar
YFP fusion proteins lacking the extracellular and
transmembrane domains remained in the soluble
fraction. Confocal microscopy with AtSERK1-YFP
expressed in plant protoplasts showed the presence
of the fusion proteins in the plasma membrane in a
pattern comparable with that of an EGFRex-
AtSERK1kin-YFP chimeric protein. Finally, FCS

650
AtSERK1 Localization and Oligomerization
indicated that the AtSERK1-YFP fusion protein
was located in the plasma membrane of insect
cells. Based on these data we propose that the
functional AtSERK1 receptor kinase is located in
the plasma membrane.
phorylation.8,40 Well-known examples are the
epidermal growth factor receptor (EGFR) and
platelet derived growth factor receptor (PDGFR)
that undergo dimerization in response to binding
of their respective ligands. However, the sensitivity
of the PDGFR41 and some other receptors like cal-
cium sensing receptor (CaR)42 to reducing agents
suggests that one type of intermolecular interaction
mediating dimerization is via intermolecular disul-
®de bonds.
The mechanism by which plant receptors trans-
duce signals across the cell surface is largely
unknown. There is evidence which suggests that
plant receptors like CLV1 and CLV2 also
dimerize12 and are assembled into a heterodimeric
complex to transduce extracellular signals.15 Recent
studies have shown that CLV1 and CLV3 form a
potential receptor/ligand pair and that CLV3
binds to the CLV1 and CLV2 heterodimer to form
a multimeric protein complex.16
Our experiments suggest that the AtSERK1
receptor can form oligomerized structures such as
homodimers. This is based both on yeast inter-
action experiments and expression of AtSERK1-
CFP/YFP fusion proteins in protoplasts. In the
latter studies, one potential problem may be that
AtSERK1 expression is raised far beyond normal
levels due to the use of the strong constitutive 35 S
promoter. However, because the majority of the
protoplasts expressing AtSERK1 fusion proteins do
not show FRET and therefore exist as monomers in
the membrane, this does not appear to be a wide
spread problem. It also appeared that there are
only certain regions on the plasma membrane
where AtSERK1 is not monomeric. This could indi-
cate that a minority of AtSERK1 receptor molecules
on the plasma membrane is in a predimerized
state. This observation is in line with EGF receptor
dimerization in mammalian cells. High af®nity
EGF receptors (12 %) dimerize highly in some
regions of the plasma membrane. Based on the
quantitative determination of FRET on the EGF
receptor in mammalian cells, it was suggested that
the high af®nity subclass of receptors are present
in a predimerized state in the absence of the
ligand.36 The predimerized AtSERK1 receptor mol-
ecules on the plasma membrane appear to be far
less abundant than those observed for the EGF
receptor owing to the fact that in only one out of
®ve measurements, there appeared to be a change
in the ¯uorescence intensity ratio. In addition, the
ratio change is on the lower side of FRET detect-
ability. This also indicates that the region of the
plasma membrane used for single FRET measure-
ments (indicated by rectangles in Figure 7 and
Figure 8) mainly contain AtSERK1 receptors that
are in the monomeric form. The small population
of non-monomeric AtSERK1 receptors are not
detected in the absence of the LZ domain in the
AtSERK1 receptor, implying that our FSPIM detec-
tion system is sensitive enough to detect
the change from a fully monomeric to a state
The correct assembly and localization into the
plasma membrane of
a glycosylated receptor
kinase such as AtSERK1 must involve a series of
events. The AtSERK1 predicted proprotein has a 29
amino acid signal sequence, immediately followed
by the LZ domain and the ®ve LRRs (V. Hecht
et al., unpublished results). Thus, it is likely that
the protein is synthesized on the ER membrane
and translocated into the ER lumen where N-glyco-
sylation of the LRRs takes place. The extracellular
domains of most of the characterized transmem-
brane receptors have been shown to be involved in
ligand binding, oligomerization and the sub-
sequent activation of the intracellular kinase
domain and the initiation of the downstream signal
transduction cascade.4 More recently, it is becom-
ing clear that the extracellular domains are also
involved in correct targeting of the receptors. For
instance, the correct folding, assembly and
expression of the human insulin receptor depends
on glycosylation of the extracellular domains.20
Inhibition of N-linked glycosylation by tunicamy-
cin leads to the intracellular accumulation of non-
glycosylated insulin proreceptors, that fail to fold
correctly.37 The LRRs of AtSERK1 contain ®ve
putative N-glycosylation sites. Two of these are
present in the second LRR, two in the fourth and
one in the ®fth LRR starting from the N terminus
of the AtSERK1 protein. Removal of at least two of
the LRRs resulted in considerable reduction of the
number of AtSERK1 protein molecules being
targeted to the plasma membrane. In line with the
role of N-linked glycosylation in insulin receptor
targeting, AtSERK1 mistargeting would be due
to the lack of glycosylation of the extracellular
LRRs. Support for this hypothesis comes from tuni-
camycin studies suggesting a role of the N-linked
glycosylation in AtSERK1 protein targeting to the
plasma membrane.
Several roles of addition of N-linked oligosac-
charides to nascent proteins have been demon-
strated; for example, N-linked oligosaccharides
attach to lectin-like molecular chaperones such as
calnexin and calreticulin,19 facilitating correct pro-
tein folding.38 N-linked oligosaccharides also play
a role in the ``quality control'' system of the ER
that ensures selective transportation of the prop-
erly folded proteins destined for the Golgi com-
plex.39 Removal of the N-linked oligosaccharides
either by chemical means or through removal of
LRRs may thus result in incorrectly folded, non-
glycosylated AtSERK1 proteins and these may be
trapped in the ER. This suggests that recognition
of properly glycosylated proteins by an ER-based
sorting machinery is an essential element of the
assembly and localization process.
Receptor dimerization is employed as a mechan-
ism for both receptor activation and autophos-

AtSERK1 Localization and Oligomerization
651
schematically in Figure 1. The AtSERK1 intracellular
kinase domain was similarly cloned into the NcoI site
of pMON999-CFP and pMON999YFP using primers
nAtSERKC 50 CCATCCGATGGGCCCACTAGATATTT
TCTT 30 and NcoI2068r. This resulted in the fusion con-
structs AtSERK1-YFP;AtSERK1-CFP (Figure 1(b) and (c))
and AtSERK1kin-YFP;AtSERK1kin-CFP (Figure 1(d) and
(e)). A 60-bp NocI-KpnI fragment DyiiKgPSp KpnI-NocI
fragment corresponding to the coding sequence of the
AtSERK1 cDNA without the leucine zipper were pre-
pared by PCR using primer pair NcoI215 and KpnI260 50
CAAATTAGCAGAAGCAAGCCAC 30 and primer pair
KpnI420 50 TGGGGAATGCAGAGTTATCTGGC 30 and
NcoI2068, respectively. Both were ligated into the NcoI
site of PMON999-CFP/YFP plasmids to create truncated
AtSERK1 fusion constructs without the leucine zipper
where monomeric and oligomeric AtSERK1
receptor coexist.
LZ domains have been shown to play a role in
dimerization of various nuclear and cytoplasmic
proteins such as Nek2,43 ZIP kinases,44 MLK-345
and TSC-22,46 respectively. There are also recent
reports that show the involvement of the LZ
domain in receptor dimerization.47 The presence of
cysteines in the N-terminal extracellular domains
of transmembrane receptors has been shown to
have a role in receptor dimerization by the for-
mation of disul®de bridges.48 In animals, the type
II transforming growth factor receptor (also a
receptor serine/threonine kinase) can form homo-
dimers in the absence of ligands,49,50 whereas
the insulin growth factor receptor is found as a
disul®de-linked tetramer.51 Since it is not known
whether ligands that bind to AtSERK1 are present
in the protoplast culture medium, we cannot
answer whether the putative AtSERK1 dimers
detected by FSPIM represent ligand-activated
dimers or preassociated dimers. The presence of
conserved cysteine pairs in the extracellular
domains of plant RLKs may imply that a popu-
lation of predimerized receptors on the cell surface
can exist.52 Among plant RLKs, CLV1 and CLV2
form a disul®de linked heterodimeric receptor
complex15 and also the ethylene receptor, ETR1, is
domain.
The
fusion
constructs
were
named
AtSERK1ÁLZ-CFP and AtSERK1ÁLZ-YFP (Figure 1(f)
and (g)). To create the truncated AtSERK1 fusion con-
structs (without the extracellular LRRs), the AtSERK1-
CFP and YFP constructs were digested with van91 and
religated,
resulting
in
AtSERK1ÁLRR-YFP
and
AtSERK1ÁLRR-CFP constructs (Figure 1(h) and (i)). A
series of constructs with one or more LRRs deleted were
prepared by PCR. DNA fragments of 1580-bp (primers
KpnI489 50 CGGGGTACCGAGCTTTACAGTAACAA-
CATAAC 30 and NcoI2068r), 1508 bp (primers KpnI561 50
CGGGGTACCGATCTTTACTTAAACAGCTTCT 30 and
NcoI2068r), 1364-bp (primers KpnI633 50 CGGGGTA
CCCGGCTTAACAACAACAGTCT 30 and NcoI2068r),
and 1292-bp (primers KpnI705 50 CGGGTACCGATC-
TATCAAATAACAGACTCT 30 and NcoI12068r) gener-
ated with 50 KpnI and 30 NcoI linkers were prepared by
PCR. They were ligated together with a 60-bp NcoI-KpnI
fragment corresponding to the signal sequence of
AtSERK1 cDNA into the NcoI site of pMON999-YFP
plasmids. This resulted in the series of truncated
a
disul®de linked dimer.53 The SRG protein
involved in the pollen recognition in Brassica
stigma is also found as a disul®de-linked dimer.54
The AtSERK1 protein has four cysteine residues in
the extracellular domain so the small amount of
AtSERK1 oligomers found might arise due to the
formation of disul®de bridges between the
cysteines in the LZ domain. Generation of stable
disul®de-bonded receptor-like-protein tyrosine
phosphatase-a, (RTPT-a) homodimers by introdu-
cing a single cysteine in the extracellular domain
immediately adjacent to the transmembrane
domain has been shown.55 The introduction of a
cysteine residue in a comparable region as RTPT-a
does however not change the oligomerization state
of the AtSERK1 receptor, so disul®de linkages may
be of less importance in AtSERK1 homodimeriza-
tion.
AtSERK1
fusion
constructs
designated
as
AtSERK1ÁLRR1-YFP (Figure 1(j)), AtSERK1ÁLRR1,2-YFP
(Figure 1(k)), AtSERK1ÁLRR1,2,3-YFP (Figure 1(l)) and
AtSERK1ÁLRR1,2,3,4-YFP (Figure 1(m)), respectively.
A 2.1-kb BglII/NcoI fragment corresponding to the
extracellular and transmembrane domain of human
EGFR was ampli®ed by PCR using primers EGFBglIIf
50 GGAAGATCTGCGATGCGACCCITCCGGG 30 and
EGFNcoIr 50 CATGCCATGGGGCGCTTCCGAACGAT-
GTGG 30 and ligated into the BglII/NcoI sites of
PMON999-CFP/YFP vectors. The resulting plasmids
were digested with NcoI and the NcoI/NcoI fragment
encoding the AtSERK1 intracellular kinase domain was
cloned in frame with the C-terminus of EGFR cDNA.
This resulted in EGFRex-AtSERK1kin-YFP/EGFRex-
AtSERK1kin-CFP chimeras (Figure 1(n) and (o)).
While identi®cation of the ligand for AtSERK1 is
clearly essential for determining its precise mode
of action, we propose that AtSERK1 exhibits
ligand-induced homodimerization which would
place it in the EGF class of receptors.
For expression in insect cells a Pdual Fast Bac vector,
that contains a polyhedrin and a p10 promoter was
used. The entire coding sequence of the AtSERK1-CFP
using primers KCNot1 50 ATAAGAATGCGGCCGCG-
GATCCTTACTTGTACAGCTCG 30 and KCSa1I 50
ACGCGTCGACAGAGCCATGGTGGAGTCGAG 30 and
the AtSERK1-YFP fusion using primers KYKpnI 50
CGGGGTACCGGATCCTTACTTGTACAGCTCG 30 and
KYSmaI 50 TCCCCCGGGAGAGCCATGGTGGAGTC-
GAG 30 were ampli®ed by PCR and cloned into the
SmaI/KpnI sites downstream of the polyhedrin promoter
and SalI/NotI sites downstream of the p10 promoter
respectively. This resulted in the AtSERK1ins-CFP-YFP
construct (Figure 1(g)). All the constructs were con®rmed
by sequence analysis.
Materials and Methods
AtSERK1-YFP/CFP tagged vectors
The entire open reading frame of the AtSERK1 cDNA
was ampli®ed by PCR from a AtSERK1 full-length
cDNA (acc. no: A67827) and cloned downstream of
the 35 S promoter into the NcoI site of pMON999-CFP
and pMON999-YFP using primers NcoI215f 50 CATGC-
CATGGTGGAGTCGAGTTATGTGG 30 and NcoI2068r 50
CATGCCATGGACCTTGGACCAGATAACTC 30. This
clone and all subsequently described clones are shown

652
AtSERK1 Localization and Oligomerization
Site-directed mutagenesis
were washed in TBS-T buffer (10 mM Tris-HCl (pH 7.5),
150 mM NaCl, 0.2 % (v/v) Tween-20) and blocked with
5 % (w/v) non-fat dried milk in TBS-T with for three
hours at room temperature, washed with TBS-T and
incubated with the anti-rabbit GFP antibody (Clontech;
1:100 ®nal dilution) for 1.5 hours at room temperature.
After three washes in TBS-T, membranes were incubated
with alkaline phosphatase conjugated anti-rabbit anti-
body (New England Biolabs; 1:1000 ®nal dilution).
Membranes were washed in TBS-T, incubated in
Western blue substrate for Alkaline Phosphatase detec-
tion (Promega) until staining was visible.
This fusion constructs AtSERK1-YFP and AtSERK1-
CFP were used for replacing tyrosine230 with single
unpaired cysteine230 in the extracellular domain
immediately adjacent to the transmembrane domain. Site
directed mutagenesis was performed according to manu-
facturer's instructions (Stratagene). This resulted in the
AtSERK1Y230C-YFP and AtSERK1Y230C-CFP (Figure 1(p)
and (q)) constructs. Clones incorporating the mutations
were con®rmed by DNA sequence determination.
Transient expression in protoplasts
Construction and transformation of yeast two
hybrid plasmid vectors
Cowpea mesophyll protoplasts were prepared essen-
tially as described by Rezelman et al.56 except that
10 mM CaCl2 was added to the enzyme solution (0.5 M
mannitol, 10 mM CaCl2, 0.1 % cellulase and 0.05 % pecti-
nase). Ten to ®fteen micrograms of each puri®ed plasmid
DNA in 20-75 ml of water was added to 0.5-1 Â 106 pro-
toplasts in 75-150 ml of ice cold solution C (0.6 M manni-
tol, 10 mM CaCl2 (pH 5.3-5.8), mixed by gentle shaking
and followed immediately by the addition of solution H
(40 % (w/v) PEG Mw 6000, 0.6 M mannitol, 0.1 M
Ca(NO3)2. The protoplast suspension was incubated for
5-25 seconds with gentle shaking followed by the
addition of 4.5 ml of solution M (0.5 M mannitol, 15 mM
MgCl2 and 0.1 % MES (pH 5.3-5.7) and incubated at
room temperature for 20 minutes. Protoplasts were then
washed 3 times with solution M and incubated for 8-24
hours as described by Rottier et al.57 For glycosylation
studies, protoplasts were treated with 20 mg ml 1 of tuni-
camycin three hours after transfection. Protoplasts were
mounted on microscopic slides and visualized under
¯uorescence microscope. For dimerization studies with
2,4-D, cotransfected protoplasts were treated with 2 mM
2,4-D 12 hours after transfection.
The coding sequence of AtSERK1 cDNA without sig-
nal sequence (spanning amino acid residues 31-625), the
kinase domain (amino acid residues 266 to 625) and
extracellular domain (amino acid residues 26 to 234)
were cloned into either PEG202 bait or PJG4-5 prey yeast
expression vectors to generate the translational fusions
of AtSERK1 domains with either the E. coli peptide (B42)
as activation domain (AD: pJG 4-5) or with the LexA
DNA binding domain (BD: pEG202) obtained from the
DuplexA two-hybrid system (Origene Technologies). All
constructs were veri®ed by sequence analysis and are
shown in schematic form in Figure 1. Yeast (Saccharo-
myces cerevisae) strain EGY 48 (lacZ) was transformed
with the plasmid constructs and an autoactivation assay
and a repression assay of the AtSERK1 fusion constructs
was performed according to the users manual to deter-
mine possible self-activation and proper entry of the
proteins into the nucleus, respectively. Yeast cells
cotransformed with both bait and prey constructs were
grown at 30 ꢁC for two days. Colonies obtained were
scored as positive when they could grow in absence of
leucine and show b-galactosidase activity within 72
hours at 30 ꢁC.
Expression in insect cells
Monolayers of Sf2l cells (1 Â 106 cells/35 mm Petri
dish) were cotransfected with 2 mg of recombinant
plasmid DNA (obtained by transforming DH10BAC
Escherichia coli cells (GIBCO-BRL)) with AtSERK1ins-CFP-
YFP Pdual FastBac plasmid DNA using a lipofectin
reagent (GIBCO-BRL) as described.58 After 72 hours of
incubation in Graces's medium (GIBCO-BRL) at 27 ꢁC,
the supernatant containing extracellular virus particles
was collected and Sf21 cell monolayers were infected.
For glycosylation studies in insect cells, the tunicamycin
treatment was performed on cells prior to the infection
with recombinant AtSERK1ins-CFP-YFP virus.
Confocal microscopy
Surface ¯uorescence and intracellular ¯uorescence
was analyzed by confocal laser scanning microscopy
(CLSM) using a Zeiss LSM 510 microscopy system (Carl-
Zeiss) based on an Axiovert inverted microscope
equipped with an Argon ion laser as excitation source.
Protoplasts were excited at 514 nm and YFP ¯uorescence
was separated from chlorophyll ¯uorescence by using a
Zeiss 630 dichroic beamsplitter. YFP emission was selec-
tively detected by using bandpass 535-590 nm ®lter, and
chlorophyll ¯uorescence was detected simultaneously in
another detection channel using a 650 nm longpass ®lter.
A 40Â oil immersion objective (numerical aperture 1.3)
was used for scanning protoplasts with step (pixel) size
of 0.2 mm2 in the xy plane and a step size of 0.4 mm2 for
serial optical sections in the z-axis. The pinhole setting
was 40 mm, which yielded a theoretical thickness (full
width at half maximum) of approximately 1 mm.
SDS PAGE and Western Blotting
Twelve hours (cowpea protoplasts) or two to three
days (insect cells) after transfection, cells were lysed in
ice cold HB buffer (50 mM Tris-HCl (pH 7.4), 10 mM
KAc, 1 mM EDTA) supplemented with single strength
proteinase inhibitor cocktail (Boehringer). The lysates
were passed through a 25 mm needle and centrifuged at
40,000 g for 30 minutes 4 ꢁC. Thirty microlitre of 2Â SDS
sample buffer (100 mM Tris HCl(pH 6.8), 4 % (w/v)
SDS, 0.2 % (w/v) bromophenol blue and 20 % glycerol)
and DTT was added to lysates as well as to pellets. After
boiling for three minutes, the proteins were separated by
SDS-PAGE and transferred to nitrocellulose membrane
(IMMOBILON) for three hours at 50 V. The membranes
Fluorescence correlation spectroscopy
Three days after infection with AtSERK1ins-CFP-YFP,
measurements on Sf21 cells were performed with a
Zeiss-Evotec ConfoCor1 system based on an Axiovert
100 inverted confocal microscope (Carl-Zeiss). The
sample was excited using an Argon ion laser (output
10 mW). The ®lter block consisted of a 514 (Á5) exci-

AtSERK1 Localization and Oligomerization
653
tation ®lter, a 530 (Á20) dichroic mirror and a 570
(Á40) emission ®lter. The light was focused into the
sample (stored in glass-bottomed 96-well plates [Poly-
®ltronics]) by a water-immersible objective lens (40 Â ,
N.A. 1.2). The emitted light passes a size-adjustable
pinhole (diameter 40 mm) to reject out-of-focus light,
and ®nally hit an avalanche photodiode. Pro®le scans
along the optical (z) axis through the cells were car-
ried out by piezo-driven movement of the objective
lens (step size 5 mm s 1). Pro®le scans through the
Sf21 insect cells incubated with Bodipy-labeled lipids
indicated the position of the plasma membrane and
the nucleus in the pro®le scan.
4. Heldin, C. H. (1995). Dimerization of cell surface
receptors in signal transduction. Cell, 80, 213-223.
5. Bishayee, S., Majumdar, S., Khire, J. & Das, M.
(1989). Ligand-induced dimerization of the platelet-
derived growth factor receptor: monomer-dimer
occurs independent of receptor phosphorylation.
J. Biol. Chem. 264, 11699-11705.
6. Westermark, B. & Heldin, C. H. (1989). Growth
tactors and their receptors. Curr. Opin. Cell Biol. 25,
1279-1285.
7. Lemmon, M. A. & Schlessinger, J. (1994). Regulation
of signal transduction and signal diversity by recep-
tor oligomerization. Trends Biochem. Sci. 19, 459-463.
8. Schlessinger, J. (1988). Signal transduction by allo-
steric receptor oligomerization. Trends Biochem. Sci.
13, 443-447.
9. Weiner, D. B., Kokai, Y., Wada, T., Cohen, J. A.,
Williams, W. V. & Greene, M. I. (1989). Linkage of
tyrosine kinase activity with transforming ability of
the p185neu oncoprotein. Oncogene, 4, 1175-1183.
10. Spencer, D. M., Wandless, T. J., Schreiber, S. L. &
Crabtree, G. R. (1993). Controlling signal transduc-
tion with synthetic ligands. Science, 262, 1019-1024.
11. Jiang, G. & Hunter, T. (1999). Receptor signalling:
when dimerization is not enough. Curr. Biol. 9, 568-
571.
Fluorescence resonance energy transfer between
AtSERK1-CFP and AtSERK1-YFP on membrane
associated AtSERK1 receptors
FRET between the ¯uorescently labelled AtSERK1
receptors in the plasma membrane was measured by
FSPIM.59 Spectral imaging was done using a Leica DMR
epi¯uorescence microscope equipped with a 250IS ima-
ging spectrograph (Chromex) coupled to a CH250 CCD
camera (Photometrics) incorporating a back-illuminated
SIT502 chip with 512 Â 512 24 mm square pixels. The
excitation light source was a 100 W-mercury arc lamp
coupled to an excitation ®lter wheel. Fluorescent spectral
images were acquired using a 20Â Plan Neo¯uar objec-
tive (NA 0.5), a bandpass excitation ®lter (Omega), a
430DCLP dichroic mirror (Omega) and a 455 longpass
emission ®lter (Schott). Spectral images were acquired
using a 150 groove/mm grating, set at a central wave-
length of 500 nm and a slit width of 200 mm correspond-
ing to 10 mm in the object plane. Typical exposure and
CCD integration time was two to ®ve seconds. Data pro-
cessing and background subtraction was performed as
described.60
12. Williams, R. W., Wilson, J. M. & Meyerowitz, E. M.
(1997). A possible role of kinase-associated protein
phosphatase in the Arabidopsis CLAVATA1 signaling
pathway. Proc. Natl Acad. Sci. USA, 94, 10467-10472.
13. Giranton, J. L., Dumas, C., Cock, J. M. & Gaude, T.
(2000). The integral membrane S-locus receptor
kinase of Brassica has serine/threonine kinase
activity in a membranous environment and spon-
taneously forms oligomers in planta. Proc. Natl Acad.
Sci. USA, 97, 3759-3764.
14. Fletcher, J. C., Brand, U., Running, M. P., Simon, R.
& Meyerowitz, E. M. (1999). Signaling of cell fate
decisions by CLAVATA3 in Arabidopsis shoot
meristems. Science, 283, 1911-1914.
15. Jeong, S., Trotochaud, A. E. & Clark, S. E. (1999).
The Arabidopsis CLAVATA2 gene encodes
a
Acknowledgments
receptor-like protein required for the stability of the
CLAVATA1 receptor-like kinase. Plant Cell, 11,
1925-1934.
We are grateful to Dr Mark Hink for his help with the
FCS and Dr P.M.P. van Bergen en Henegouwen for pro-
viding us with the cDNA of EGFR. We are very thankful
to the people from the Laboratory of Virology, especially
Magda Usmany for providing us with the assistance
with insect cell cultures and Dr Just Vlak for some useful
discussions. We especially thank Casper Vroemen for
many useful discussions and critical comments on the
manuscript. This work was supported by grants
ERBIO4-CT96-2282 and ERBIO4-CT96-0689 from the EU
Biotechnology program (K.S. and V.H.) and the Royal
Netherlands Academy of Arts and Sciences (T.W.J.G.).
16. Trotochaud, A. E., Jeong, S. & Clark, S. E. (2000).
CLAVATA3,
a multimeric ligand for the CLA-
VATA1 receptor-kinase. Science, 289, 613-617.
17. He, Z., Wang, Z., Li, J., Zhu, Q., Lamb, C., Ronald,
P. & Chory, J. (2000). Perception of brassinosteroids
by the extracellular domain of the receptor kinase
BRI1. Science, 288, 2360-2363.
18. Nagayama, Y., Namba, H., Yokoyama, N.,
Yamashita, S. & Niwa, M. (1998). Role of aspara-
gine-linked oligosaccharides in protein folding,
membrane targeting, and thyrotropin and autoanti-
body binding of the human thyrotropin receptor.
J. Biol. Chem. 273, 33423-33428.
19. Trombetta, E. S. & Helenius, A. (1998). Lectins as
chaperones in glycoprotein folding. Curr Opin.
Struct. Biol. 8, 587-592.
20. Elleman, T. C., Frenkel, M. J., Hoyne, P. A., McKern,
N. M., Cosgrove, L. & Hewish, D. R. et al. (2000).
Mutational analysis of the N-linked glycosylation
sites of the human insulin receptor. Biochem J. 347,
771-779.
References
1. Pawson, T. & Bernstein, A. (1990). Receptor tyrosine
kinases: genetic evidence for their role in Drosophila
and mouse development. Trends Genet. 6, 350-356.
2. Cantley, L. C., Auger, K. R., Carpenter, C.,
Duckworth, B., Graziani, A., Kapeller, R. & Soltof, S.
(1991). Oncogenes and signal transduction. Cell, 64,
281-302.
3. Lease, K., Ingham, E. & Walker, J. C. (1998). Chal-
lenges in understanding RLK function. Curr. Opin.
Plant Biol. 1, 388-392.
21. Schmidt, E. D. L., Guzzo, F., Toonen, M. A. J. & de
Vries, S. C. (1997). A leucine rich repeat containing
receptor-like kinase marks somatic plant cells com-

654
AtSERK1 Localization and Oligomerization
petent to form embryos. Development, 124, 2049-
2062.
39. High, S., Lecomte, F. J., Russell, S. J., Abell, B. M. &
Oliver, J. D. (2000). Glycoprotein folding in the
endoplasmic reticulum: a tale of three chaperones?
FEBS Letters, 476, 38-41.
40. Ullrich, A. & Schlessinger, J. (1990). Signal transduc-
tion by receptors with tyrosine kinase activity. Cell,
61, 203-212.
41. Li, W. & Schlessinger, J. (1991). Platelet derived
growth factor receptor (PDGFR)-induced dimeriza-
tion of PDGF receptor in living cells. Mol. Cell Biol.
11, 3756-3761.
42. Bai, M., Trivedi, S. & Brown, E. M. (1998). Dimeriza-
tion of the extracellular calcium-sensing receptor
(CaR) on the cell surface of CaR-transfected BEK293
cells. J. Biol. Chem. 273, 23605-23610.
22. Gerdes, H. H. & Kaether, C. (1996). Green ¯uor-
escent protein: applications in cell biology. FEBS
Letters, 389, 44-47.
23. Ludin, B., Doll, T., Meili, R., Kaech, S. & Matus, A.
(1996). Application of novel vectors for GFP-tagging
of proteins to study microtubule-associated proteins.
Gene, 173, 107-111.
24. Carey, K. L., Richards, S. A., Lounsbury, K. M. &
Macara, I. G. (1996). Evidence using a green ¯uor-
escent protein-glucocorticoid receptor chimera that
the Ran/TC4 GTPase mediates an essential function
independent of nuclear protein import. J. Cell Biol.
133, 985-996.
43. Fry, A. M., Arnaud, L. & Nigg, E. A. (1999). Activity
of the human centrosomal kinase, Nek2, depends on
an unusual leucine zipper dimerization motif. J. Biol.
Chem. 274, 16304-16310.
25. FoÈrster, T. (1948). Intermolecular energy migration
and ¯uorescence. Ann. Physik. 6, 55-75.
26. Johnson, I. D., Kang, H. C. & Haugland, R. P.
(1991). Fluorescent membrane probes incorporating
dipyrrometheneboron di¯uoride ¯uorophores. Anal.
Biochem. 198, 228-237.
27. Rizzuto, R., Brini, M., De Giorgi, F., Rossi, R., Heim,
R., Tsien, R. Y. & Pozzan, T. (1996). Double labelling
of subcellular structures with organelle-targeted GFP
mutants in vivo. Curr. Biol. 6, 183-188.
28. Gadella, T. W. J., Jr, van der Krogt, G. & Bisseling,
T. (1999). GFP-based FRET microscopy in living
plant cells. Trends Plant Sci. 4, 287-291.
44. Baxevanis, A. D. & Vinson, C. R. (1993). Interactions
of coiled coils in transcription factors: where is the
speci®city? Curr. Opin. Genet. Dev. 3, 278-285.
45. Leung, W. & Lassam, N. (1998). Dimerization via
tandem leucine zippers is essential for the activation
of the mitogen-activated protein kinase kinase
kinase, MLK-3. J. Biol. Chem. 273, 32408-32415.
46. Kester, H. A., Blanchetot, C., den Hertog, J., van der
Saag, P. T. & van der Burg, B. (1999). Transforming
growth factor-beta-stimulated clone-22 is a member
of a family of leucine zipper proteins that can homo
and hetero dimerize and has transcriptional repres-
sor activity. J. Biol. Chem. 274, 27439-27447.
29. Jovin, T. M. & Arndt-Jovin, D. J. (1989). Lumin-
escence digital imaging microscopy. Annu. Rev.
Biophys. Chem. 18, 271-308.
30. Wu, P.
& Brand, L. (1994). Resonance energy
47. Behncken, S. N., Billestrup, N., Brown, R., Amstrup,
J., Conway-Campbell, B. & Waters, M. J. (2000).
Growth hormone (GH)-independent dimerization of
GH receptor by a leucine zipper results in constitu-
tive activation. J. Biol. Chem. 275, 17000-17007.
48. Romano, C., Yang, W. L. & O'Malley, K. L. (1996).
Metabotropic glutamate receptor 5 is a disul®de-
linked dimer. J. Biol. Chem. 271, 28612-28616.
49. Chen, R. H. & Derynck, R. (1994). Homomeric inter-
actions between type II transforming growth factor-
beta receptors. J. Biol. Chem. 269, 22868-22874.
50. Henis, Y. I., Moustakas, A., Lin, H. Y. & Lodish,
H. F. (1994). The types II and III transforming
growth factor-beta receptors form homo-oligomers.
J. Cell Biol. 126, 139-154.
51. Cheatham, B. & Kahn, C. R. (1995). Insulin action
and the insulin signaling network. Endocrine Rev. 16,
117-142.
52. Trotochaud, A. E., Hao, T., Wu, G., Yang, Z. &
Clark, S. E. (1999). The CLAVATA1 receptor-like
kinase requires CLAVATA3 for its assembly into a
signaling complex that includes KAPP and a Rho-
related protein. Plant Cell, 11, 393-406.
transfer: methods and applications. Anal. Biochem.
218, 1-13.
31. Clegg, R. M. (1995). Fluorescence resonance energy
transfer. Curr. Opin. Biotechnol. 6, 103-110.
32. Chuck, R. S., Cantor, C. R. & Tse, D. B. (1990). CD4-
T-cell antigen receptor complexes on human leuke-
mia T cells. Proc. Natl Acad. Sci. USA, 87, 5021-5025.
33. Vignali, D. A., Carson, R. T., Chang, B., Mittler, R. S.
& Strominger, J. L. (1996). The two membrane proxi-
mal domains of CD4 interact with the T cell recep-
tor. J. Ex. Med. 183, 2097-2107.
34. Guo, C., Dower, S. K., Holowka, D. & Baird, B.
(1995). Fluorescence resonance energy transfer
reveals interleukin (IL)-1-dependent aggregation of
IL-1 type I receptors that correlates with receptor
activation. J. Biol. Chem. 270, 27562-27568.
35. Carraway, K. L. & Cerione, R. A. (1993). Inhibition
of epidermal growth factor receptor aggregation by
an antibody directed against the epidermal growth
factor receptor extracellular domain. J. Biol. Chem.
268, 23860-23867.
36. Gadella, T. W., Jr & Jovin, T. M. (1995). Oligomeri-
zation of epidermal growth factor receptors on A431
cells studied by time-resolved ¯uorescence imaging
microscopy. A stereochemical model for tyrosine
kinase receptor activation. J. Cell Biol. 129, 1543-
1558.
53. Schaller, G. E., Ladd, A. N., Lanahan, M. B.,
Spanbauer, J. M. & Bleecker, A. B. (1995). The ethyl-
ene response mediator ETR1 from Arabidopsis
forms a disul®de-linked dimer. J. Biol. Chem. 270,
12526-12530.
37. Olson, T. S., Bamberger, M. T. & Lane, M. D. (1988).
Post-translational changes in tertiary and quaternary
structure of the insulin proreceptor. Correlation with
acquisition of function. J. Biol. Chem. 263, 7342-7351.
38. Bass, J., Chiu, G., Argon, Y. & Steiner, D. F. (1998).
Folding of insulin receptor monomers is facilitated
by the molecular chaperones calnexin and calreticu-
lin and impaired by rapid dimerization. J. Cell Biol.
141, 637-646.
54. Doughty, J., Dixon, S., Hiscock, S. J., Willis, A. C.,
Parkin, I. A. & Dickinson, H. G. (1998). PCP-A1, a
defensin-like Brassica pollen coat protein that binds
the S locus glycoprotein, is the product of gameto-
phytic gene expression. Plant Cell, 10, 1333-1347.
55. Jiang, G., den Hertog, J., Su, J., Noel, J., Sap, J. &
Hunter, T. (1999). Dimerization inhibits the activity
of receptor-like protein-tyrosine phosphatase-alpha.
Nature, 401, 606-610.

AtSERK1 Localization and Oligomerization
655
56. Rezelman, G., van Kammen, A. & Wellink, J. (1989).
Expression of cowpea mosaic virus mRNA in cow-
pea protoplasts. J. Gen. Virol. 70, 3043-3050.
57. Rottier, P. J., Rezelman, G. & van Kammen, A.
(1979). The inhibition of cowpea mosaic virus repli-
cation by actinomycin D. Virology, 92, 299-309.
58. Groebe, D. R., Chung, A. E. & Ho, C. (1990).
Cationic lipid-mediated co-transfection of insect
cells. Nucl. Acids Res. 18, 4033-4036.
59. Gadella, T. W., Jr, Vereb, G., Jr, Hadri, A. E., Rohrig,
H., Schmidt, J. & John, M. et al. (1997). Microspectro-
scopic imaging of nodulation factor-binding sites
on living Vicia sativa roots using a novel bioactive
¯uorescent nodulation factor. Biophys. J. 72, 1986-
1996.
60. Goedhart, J. & Gadella, T. W. J., Jr (2000). Advanced
microspectroscopic methods for monitoring ¯uor-
escent molecules in living plant root hairs. In Cell
and Molecular Biology of Living Plant Root Hairs
(Ridge, R. W. & Emons, A. M. C., eds), pp. 65-93,
Springer Verlag, Berlin.
Edited by N.H. Chua
(Received 11 December 2000; received in revised form 16 April 2001; accepted 16 April 2001)