Studies on the role of the receptor protein motifs
possibly involved in electrostatic interactions on the
dopamine D
1
and D
2
receptor oligomerization
Sylwia Łukasiewicz
1
, Agata Faron-Go
´
recka
2
, Jerzy Dobrucki
3
, Agnieszka Polit
1
and Marta Dziedzicka-Wasylewska
1,2
1 Department of Physical Biochemistry, Jagiellonian University, Krako
´
w, Poland
2 Laboratory of Biochemical Pharmacology, Polish Academy of Sciences, Krako
´
w, Poland
3 Division of Cell Biophysics, Jagiellonian University, Krako
´
w, Poland
Various molecular techniques based on biophysical, bio-
chemical and pharmacological approaches have dem-
onstrated that G protein-coupled receptors (GPCRs),

2008)
doi:10.1111/j.1742-4658.2008.06822.x
We investigated the influence of an epitope from the third intracellular
loop (ic3) of the dopamine D
2
receptor, which contains adjacent arginine
residues (217RRRRKR222), and an acidic epitope from the C-terminus of
the dopamine D
1
receptor (404EE405) on the receptors’ localization and
their interaction. We studied receptor dimer formation using fluorescence
resonance energy transfer. Receptor proteins were tagged with fluorescence
proteins and expressed in HEK293 cells. The degree of D
1
–D
2
receptor
heterodimerization strongly depended on the number of Arg residues
replaced by Ala in the ic3 of D
2
R, which may suggest that the indicated
region of ic3 in D
2
R might be involved in interactions between two dopa-
mine receptors. In addition, the subcellular localization of these receptors
in cells expressing both receptors D
1
–cyan fluorescent protein, D
2
–yellow

of GPCRs point to a new level of molecular cross-talk
among signaling molecules [1,5,11,17].
Structural information about receptor dimer forma-
tion is currently limited, and the question of whether
receptors dimerize in a similar way or have their own
paths of dimerization remains open. In general, either
covalent or noncovalent interactions are involved in
this process; however, the latter seem to be more effec-
tive [18–21]. Either the transmembrane domains (TMs)
[22–31] of GPCRs or the N- [32–34] or C-tail [35,36]
could play a role in dimer formation. It has been
shown that cysteine residues located in the extracellu-
lar loops are essential for disulfide-linked m3 musca-
rinic receptor (M3R) dimer formation; however this
kind of interaction is not the only point of contact
[37]. For GABA
B
receptors (GBR), a coiled-coil inter-
action within the C-tail of GBR1 and GBR2 seems to
be involved in receptor heterodimerization. However,
this motif is not necessary, as deleting the C-tail does
not abolish dimerization. Also, hydrophobic interac-
tions within the TM of GPCRs are essential for forma-
tion and stabilization of the dimers and have been
detected for beta-adrenergic, dopamine, muscarinic
and angiotensin receptors [38–40].
In earlier studies, the role of certain amino acid resi-
dues in the formation of noncovalent complexes
between protein molecules was highlighted. Electro-
static interactions occur between an epitope containing

ological function. The view that these receptors may
also function as a physically linked unit is especially
important because recent data suggest that the D
1
and
D
2
receptors are co-expressed by a moderate to sub-
stantial proportion of striatal neurons [45,46]. Lee et al.
provided anatomical evidence suggesting significant col-
ocalization of D
1
and D
2
receptors in the caudate and
pyramidal cells in the rat frontal cortex [47]. Earlier
studies by Vincent et al. have also shown that the lami-
nar distribution of medial prefrontal cortex neurons
expressing both D
1
and D
2
receptors was similar to
that of the mesocortical dopamine afferents [48].
The dopamine D
2
receptor can form homodimers
[19]. Recently, we have shown that the D
2
receptor

and yellow fluorescent proteins (YFP; fluorescence
acceptor) and expressed in HEK293 cells. We find
FRET to be a very sensitive tool, and measurements
are especially useful to quantitatively monitor the
physical interactions between receptor proteins [51,52].
Results
Radioligand binding assay
As shown in Table 1, the binding parameters obtained
for dopamine D
1
receptor and its mutant indicate that
the K
d
values for these two receptors were similar;
however, the density of the D
1
MUT (404AA405) was
Table 1. Binding parameters for the dopamine receptors. For dopa-
mine D
2
receptor binding, the statistical significance was evaluated
using a one-way ANOVA, followed by a Dunnett’s test for post hoc
comparison. *P < 0.05. For dopamine D
1
receptor binding, the
statistical significance was evaluated using a Student’s t-test;
***P < 0.001.
Species
B
max

lower than that of wild-type D
1
R (Fig. 1A). Also, all
three genetic variants of dopamine D
2
R displayed sim-
ilar K
d
values, but the density of these receptors
strongly depended on the number of Arg residues still
present within the receptor sequence. The D
2
R1
(217AARRKR222) mutant displayed half of the B
max
value obtained for D
2
R, whereas the density of the
D
2
R2 (217AAAAKR222) mutant was much lower
(Fig. 1B). For the D
2
R3 (217AAAAAA222) variant,
no binding parameters could be obtained, which indi-
cates that there was no receptor protein in the cellular
membrane. This conclusion is further justified by
confocal microscopy analysis of receptor localization.
Analysis of the localization of dopamine D
1

MUT, D
2
R1, D
2
R2 and
D
2
R3 receptors in different combinations. Merged pic-
tures with apparent yellow signal indicating overlap of
green fluorescent signal (CFP channel) and red fluores-
cent signal (YFP channel) show colocalization.
As seen from the figures, these receptor proteins
were localized differentially in the cell. Cell edge sharp-
ness confirms that dopamine D
1
and D
1
MUT recep-
tors localize in the plasma membrane, in contrast to
the dopamine D
2
receptor and its genetic variants,
D
2
R1, D
2
R2, which were localized in the plasma mem-
brane and inside the cell. In the case of the dopa-
mine D
2

2
R3 variant, colo-
calization was observed in both the plasma membrane
and inside the cell. For a quantitative estimation of the
degree of colocalization between the two different pro-
teins of interest, Pearson’s correlation coefficients and
coefficients of determination were estimated (Fig. 2C).
In case of cells co-expressing dopamine D
1
and dopa-
mine D
2
receptor mutants, the degree of colocalization
decreased, which was correlated with number of
exchanged residues within the ic3 of D
2
receptor.
When cells were cotransfected with the same type of
receptors (D
1
MUT–CFP ⁄ D
1
–YFP, D
2
–CFP ⁄ D
2
R1–
YFP, D
2
–CFP ⁄ D

dopamine receptors, respectively. Data are
from a single experiment performed in triplicate and are representa-
tive of at least three independent experiments. Elimination of the
Arg-rich or di-Glu motif in D
2
RorD
1
R, respectively, does not alter
the ligand binding constant.
Dopamine D
1
and D
2
receptors dimerization S. Łukasiewicz et al.
762 FEBS Journal 276 (2009) 760–775 ª 2008 The Authors Journal compilation ª 2008 FEBS
very demonstrative and gives a quick answer to
whether there is any energy transfer in the examined
sample. Therefore, we used this type of measurement
to investigate interactions between the dopamine D
1
and D
2
receptors and their genetic variants. Fluores-
cence emission profiles for the HEK293 cell suspension
expressing fusion proteins in different combinations
(D
1
–CFP ⁄ D
2
–YFP, D

–YFP and D
2
–CFP ⁄ D
2
R3–YFP) were
compared using an excitation wavelength of 434 nm
(donor absorption).
The upper panel of Fig. 3 shows emission spectra of
HEK293 cell populations after cotransfection with
plasmids encoding genes for dopamine D
1
and D
2
receptor fusion proteins (D
1
–CFP and D
2
–YFP) in
comparison with emission spectra of the cell popula-
tions that co-express dopamine D
1
receptor fusion
protein (D
1
–CFP) and one of the genetic variants of
dopamine D
2
receptor fusion protein (D
2
R1, D

tor was present in the sample, there was no visible
energy transfer, despite the presence of both fluoro-
phores in the sample.
Figure 3C,D shows the emission profiles of cells
cotransfected with plasmids encoding genes for the
same type of dopamine receptor (D
1
or D
2
, respec-
tively), tagged with different fluorescence proteins,
A
B
C
Fig. 2. Expression of D
1
R and D
2
R and their mutants in HEK293 cells. (A) HEK293 cells were cotransfected with either D
1
–CFP or D
1
MUT–
CFP and either D
2
–YFP, D
2
R1–YFP, D
2
R2–YFP, D

2
–CFP and either D
2
–YFP, D
2
R1–YFP, D
2
R2–YFP or D
2
R3–YFP. Image overlays show extensive colocalization in
every case. (C) Bar graph of Pearson‘s correlation coefficient calculated for HEK293 cells cotransfected with different dopamine D
1
and D
2
receptor protein construct combination. Data are mean ± SE, and statistical significance was evaluated using Student’s t-test and Mann–
Whitney test. ***P < 0.001 for combinations D
1
with all variants of D
2
versus D
1
⁄ D
2
. Either D
2
⁄ D
2
R1, D
2
⁄ D

S. Łukasiewicz et al. Dopamine D
1
and D
2
receptors dimerization
FEBS Journal 276 (2009) 760–775 ª 2008 The Authors Journal compilation ª 2008 FEBS 763
compared with emission profiles of cells in which one
of the tagged receptors was its own mutant (D
1
–CFP ⁄
D
1
–YFP and D
1
MUT–CFP ⁄ D
1
–YFP or D
2
–CFP ⁄
D
2
–YFP and D
2
–CFP ⁄ D
2
R3–YFP). The lower panel
of Fig. 3 shows that both dopamine receptors, D
1
and
D

1
–YFP ⁄ a
i
–CFP or D
2
YFP ⁄ a
s
CFP,
despite the identical overexpression level of the
proteins in all studied combinations.
Fluorescence lifetime microscopy studies of
dopamine receptor dimerization
Time-correlated single-photon counting experiments
were performed on the inverted fluorescence micro-
scope. The FRET phenomenon was observed in a
single living cell transiently transfected with the
dopamine D
1
and D
2
receptors and their genetic
variants, tagged with fluorescent proteins. This kind of
measurement provides highly quantifiable data because
it is independent of any change in fluorophore concen-
tration or excitation intensity.
To determine FRET efficiency, precise measurement
of the donor fluorescence lifetime (CFP), in the pres-
ence and absence of the acceptor (YFP), is required.
A
C

1
–YFP (gray line) in comparison with D
1
–CFP and D
1
–YFP (black
line). (D) Cotransfection of HEK293 with D
2
–CFP and D
2
R3–YFP (gray line) in comparison with D
2
–CFP and D
2
–YFP (black line). CFP was
excited at 434 nm, and fluorescence was detected at 450–550 nm through a double monochromator. The spectral contributions arising from
light scattering and nonspecific fluorescence of cells and buffer were eliminated.
Dopamine D
1
and D
2
receptors dimerization S. Łukasiewicz et al.
764 FEBS Journal 276 (2009) 760–775 ª 2008 The Authors Journal compilation ª 2008 FEBS
Fluorescence decays were analyzed as both mono- and
multi-exponentials. Analysis of the reduced chi-squared
value and residual distribution led to the conclusion
that best fit parameters were obtained with two expo-
nentials. Adding a third exponential did not signifi-
cantly influence the parameters, and the fractional
contribution of the additional lifetime was close to

R1, D
2
R2 or D
2
R3) and also when
D
1
MUT was used instead of the dopamine D
1
recep-
tor. Transfer efficiency was equal to 2.1% (2.32 ns) for
A
B
C
D
Fig. 4. Representative fluorescence emission spectra of HEK293 cells cotransfected with either D
1
–YFP or D
2
–YFP and Ga–CFP fusion pro-
teins. (A) Negative FRET control, spectra from a 1 : 1 mixture of cells individually expressing the Ga
S
–CFP (black line) fusion protein (excited
at 434 nm) and the D
1
–YFP (gray line) fusion protein (excited at 475 nm). (B) Cotransfection of HEK293 cells with D
1
–YFP and Ga
S
–CFP

2
R1, further decreased to 1.26% (2.34 ns) for
D
1
⁄ D
2
R2, and finally reached the value of 0.44%
(2.36 ns) for D
1
⁄ D
2
R3.
The lowest E value, similar to that obtained for the
D
1
⁄ D
2
R3 combination, was observed for D
1
MUT ⁄
D
2
R3 and was equal to 0.4% (2.36 ns). A similar
result (0.8%; 2.35 ns) was obtained for cells co-express-
ing the dopamine D
1
receptor mutant (D
1
MUT) and
the wild-type dopamine D

2
⁄ D
2
R3, it equaled
3.4% (2.29 ns) versus 3.5% (2.28 ns) for D
2
⁄ D
2
combi-
nations.
The summary of TCSPC results is presented in
Tables 2 and 3. The error of the average fluorescence
lifetime is the standard error of mean obtained from
different cells and independent transfections (we
ignored standard deviations derived from fitting of
individual fluorescence decay because they were very
small).
Discussion
The data provided from numerous studies indicate that
oligomerization may play important roles in receptor
trafficking and ⁄ or signaling. In several cases, receptors
appear to fold into constitutive dimers early after bio-
synthesis, although ligand-promoted dimerization at
the cell surface has been also proposed [53]. Many
GPCRs have been shown to participate in homo- or
heterodimerization [54]. Using a biophysical approach,
we had previously shown that the D
2
and D
1

2
–YFP.
Species
Average lifetime (ns)
Transfer
efficiency
ÆEæ (%)Æs
D
æÆs
DA
æ
D
1
–CFP
a
2.37 ± 0.01
D
1
–CFP ⁄ D
2
–YFP
b
2.27 ± 0.02 4.01
D
1
–CFP ⁄ D
2
R1–YFP
c
2.32 ± 0.02 2.10*

1
receptor.
b
Measured in cell co-expressing dopamine D
1
and D
2
fusion proteins (D
1
–CFP and D
2
–YFP).
c
Measured in cell
co-expressing dopamine D
1
and D
2
fusion proteins (D
1
–CFP and
D
2
R1–YFP – genetic variant of dopamine D
2
receptor).
d
Measured
in cell co-expressing dopamine D
1

teins (D
1
MUT–CFP – genetic variant of dopamine D
1
receptor and
D
2
–YFP).
g
Measured in cell co-expressing dopamine D
1
and D
2
fusion proteins (D
1
MUT–CFP – genetic variant of dopa-
mine D
1
receptor and D
2
R3–YFP – genetic variant of dopamine D
2
receptor).
Dopamine D
1
and D
2
receptors dimerization S. Łukasiewicz et al.
766 FEBS Journal 276 (2009) 760–775 ª 2008 The Authors Journal compilation ª 2008 FEBS
experimental studies also suggested the participation of

this question. The receptor proteins under investigation
were tagged with fluorescent proteins and transfected
into HEK293 cells; their localization was then
observed with the use of a confocal microscope. The
degree of receptor dimerization was also judged by
changes in fluorescence lifetime, which we find to be
the most sensitive technique with which to measure
FRET [49].
The results presented here indicate that dopa-
mine D
1
and D
2
receptors form homo- and hetero-
dimers; results that are in agreement with previously
published data [19,49,55]. Measuring receptor dimer-
ization by monitoring changes in the fluorescence life-
time of probes linked to the receptors of interest seems
the best approach in this kind of the study. Although
the approach enables only qualitative estimation of
FRET phenomenon, steady-state fluorescence spectros-
copy measurements in suspension are also useful
because they are very demonstrative. In this study,
both approaches yield similar conclusions, although we
are aware that quantitative results can only be
obtained from fluorescence lifetime microscopy.
An often-discussed problem when using biophysical
techniques to study receptor oligomerization is that
these experiments predominantly involve heterologous
expression systems, which in most cases have been per-

2
receptors and their interactions
with the appropriate a subunits of G protein, further
confirm that the use of advanced fluorescence techni-
ques does indeed allow for the observation of true
interactions. The dopamine D
1
receptor did not inter-
act with Ga
i
, and the D
2
receptor did not interact with
Ga
s
, although the physical contact of these receptors
with their appropriate a subunit partners could indeed
have been observed, despite the identical level of over-
expression of the proteins in all studied combinations.
The experiments described above serve as a control
that must always be performed when using FRET to
determine if two proteins interact. That control is to
express (preferentially using the same expression con-
struct in all experiments) two noninteracting fusion
proteins that carry CFP and YFP in the same cell and
Table 3. Summary of energy transfer measurements obtained by
fluorescence lifetime microscopy in HEK293 cells. Excitation was
set up at 434 nm, and emission was observed through appropriate
interference filters, as described in Experimental procedures. The
standard errors of means (obtained from at least 15 single cells)

2
–CFP
d
2.37 ± 0.02
D
2
–CFP ⁄ D
2
–YFP
e
2.28 ± 0.02 3.50
D
2
–CFP ⁄ D
2
R3–YFP
f
2.29 ± 0.01 3.40
a
Measured in cell expressing CFP coupled to dopamine D
1
recep-
tor.
b
Measured in cell co-expressing two dopamine D
1
receptor
fusion proteins (D
1
–CFP and D

Measured in cell
co-expressing two dopamine D
2
receptor fusion proteins (D
2
–CFP
and D
2
R3–YFP – genetic variant of dopamine D
2
receptor).
S. Łukasiewicz et al. Dopamine D
1
and D
2
receptors dimerization
FEBS Journal 276 (2009) 760–775 ª 2008 The Authors Journal compilation ª 2008 FEBS 767
show that there was no FRET fluorescence after nor-
malizing and making corrections for cross-talk. In
experiments investigating receptor interactions, that
was the case; FRET was observed only when the
receptor was co-expressed with the appropriate a sub-
unit of the G protein and not in the other case.
Although there is discussion in the literature concern-
ing the possibilities of photoconversion of YFP into a
CFP-like species during acceptor photobleaching
FRET experiments, we, as well as others, can exclude
that such photoconversion interferes with FRET
measurements under standard conditions.
Two acidic residues in the C-terminal end of the D

in the FRET efficiency by $ 50 to 1.26%. When all res-
idues in the basic region of the D
2
receptor were
replaced, only a marginal level of energy transfer was
observed (0.44%). A similar effect on energy transfer
was observed after the replacement of two acidic Glu
residues in the C-tail of the D
1
receptor. The efficiency
of energy transfer was reduced to 0.8%. A possible
interpretation of the data suggests that the indicated
basic region of ic3 of the D
2
receptor and acidic region
of the C-tail of the D
1
receptor might be involved in
the interactions between the two dopamine receptors.
In addition, the subcellular localization of D
1
–CFP,
D
2
–YFP and all the mutants of both receptors was
examined in cells expressing one or both types of
receptors using confocal microscopy. In cotransfected
cells, both the D
1
and D

indicates that the significant decrease in energy transfer
observed between D
1
MUT and D
2
is the effect
of impaired heterodimerization of the dopamine
receptors.
Moreover, confocal microscopy experiments revealed
that modification of the Arg-rich region in the ic3 of
the D
2
receptor substantially changed its receptor traf-
ficking properties. The binding experiments also
pointed to a decrease in the density of the D
2
R vari-
ants in the cellular membrane; the number of D
2
receptor binding sites decreased with the number of
changed Arg residues in the ic3. When compared with
wild-type receptor, the binding of [
3
H]spiperone to
D
2
R1 and D
2
R2 showed a significant decrease in the
B

fact that wild-type D
2
–D
2
R3 homodimers are being
created during D
2
receptor biosynthesis, whereas that
process does not take place in the case of D
1
-D
2
R3
co-expression. It is probably the direct interactions
between the D
2
and the D
2
R3 receptor mutant that
reduced efficiency in the trafficking of the wild-type
receptor to the cell surface. These observations are
consistent with data showing that co-expression of a
C- or N-terminal-truncated D
2
receptor with the wild-
type receptor resulted in attenuation of binding and
reduced efficiency in the trafficking of the wild-type D
2
receptor [61].
The construction of genetic variants of the studied

the C-terminal acidic residues are by no means
involved in the regulation of D
1
receptor membrane
localization.
However, genetically manipulating the Arg-rich epi-
tope in the ic3 of the D
2
receptor induced alterations in
the cellular localization of the resulting receptor pro-
teins. If not for confocal microscopy, which allowed for
the visualization of receptor localization, the gradual
decrease in the degree of D
1
–D
2
receptor (and its vari-
ants) heterodimerization that was observed in FRET
experiments could have been interpreted as a direct
indication of the role of the Arg-rich epitope in the for-
mation of heterodimers, as had been done in case of
adenosine A
2A
–dopamine D
2
heterodimerization [43].
However, based on these data, we have to conclude
that the Arg-rich epitope in the ic3 loop of D
2
is also

endoplasmic reticulum [62,64,65]. Under normal condi-
tions, this motif is masked, and proteins are trans-
ported to the cell surface without significant
accumulation in the endoplasmic reticulum. If the Arg-
rich motif in D
2
R serves as a retention signal, then
replacing adjacent Arg residues should increase the
surface expression of D
2
R. We observed the opposite
effect; the Arg-rich sequence in the cytoplasmic ic3
loop of D
2
R does not act as an endoplasmic reticulum
retention signal. Misfolding of the D
2
R2 and D
2
R3
mutants could potentially be responsible for their accu-
mulation in the endoplasmic reticulum because only
protein that has assumed its native conformation is
available for recruitment into the transport vesicles
leaving the endoplasmic reticulum. Therefore, the Arg-
rich motif might be responsible for interactions with
cytoskeletal proteins. Binda et al. have shown that
cytoskeletal protein 4.1 N, a member of the 4.1 family,
facilitates the transport of D
2

well as its dimerization with other receptor partners, is
very important for understanding the rules that govern
receptor activity, both in physiological and patholo-
gical conditions.
Receptor dimerization, which is important for trans-
membrane signal generation [54], also plays a role in
intracellular trafficking of receptors and controlling
their folding status. As suggested by So et al., hetero-
oligomerization, by changing the exposure or masking
motifs responsible for endoplasmic reticulum retention
or export, may be a strong regulator of the cellular
distribution of receptors [14].
Incorrect membrane localization of D
2
R after modi-
fication within ic3 217–222 region (observed in the cells
co-expressing D
1
R and D
2
R3) can result from defec-
tive interactions with cytoskeletal proteins as well as
from impaired heterodimerization with D
1
R. When in
the cell both D
2
R3 mutant and D
2
R wild-type are

encoding the human dopamine D
1
, human dopamine D
2
receptors and a subunits of G proteins were obtained from
the UMRcDNA Resource Center (University of Missouri-
Rolla, MO, USA). The bacterial cell line Escherichia coli
DH5a (Dam+) was purchased from Novagen (Darmstadt,
Germany).
HEK293 cells were obtained from the American Type
Culture Collection (Manassas, VA, USA). All cell culture
materials were purchased from Gibco (Carlsbad, CA, USA)
and Sigma (Poznan
´
, Poland).
Construction of fusion proteins
The human dopamine D
1
and D
2
receptor genes were
cloned into the pcDNA3.1(+) plasmid and used as the
starting point to construct the fusion proteins. Molecules
were tagged with cDNA encoding enhanced cyan or yellow
fluorescent proteins (ECFP or EYFP) and used after
expression as the fluorescence donor or acceptor, respec-
tively. Henceforth the cyan (ECFP) and yellow (EYFP)
variants are called CFP and YFP, respectively.
The full-length cDNAs encoding the above-mentioned
proteins were PCR-amplified. The forward primer was uni-

mutated to an alanine residue.
The appropriate point mutations were produced accord-
ing to the QuikChange II Site-Directed Mutagenesis Kit
Manual (Stratagene, La Jolla, CA, USA). Dopamine D
1
and D
2
genes inserted into pECFP–N1 and pEYFP–N1
vectors, respectively, were used as the mold for the PCR-
Quik reaction. Incorporating the oligonucleotide primers,
each complementary to the opposite strand of the vector
and containing the desired mutations, generated a mutated
plasmid. The resulting product was treated with endonucle-
ase DpnI, specific for methylated and hemimethylated
DNA, in order to select synthesized DNA containing the
introduced mutations. E. coli DH5 a cells were transformed
with mutated plasmid. D
2
R1–pEYFP vector was used as
the mold for PCR-Quik in which D
2
R2–pEYFP was
obtained, which then served as the starting point to make
the D
2
R3–pEYFP construct.
Cell culture and transfection
HEK293 cells were grown in Dulbecco’s modified essential
medium, supplemented with 1% l-glutamine and 10%
heat-inactivated fetal bovine serum, at 37 °C in an atmo-

30 mm dishes at a density of 1 · 10
6
cells per dish for fluo-
rescence lifetime measurements and confocal imaging. They
were transfected with 12 lg of DNA per 100 mm dish and
2 lg of DNA per 30 mm dish. The ratio of DNA coding
donor to DNA coding acceptor was 1 : 1 or 1 : 2.
Membrane preparation and radioligand
binding assay
For binding experiments, the transfected HEK293 cells
were washed with NaCl ⁄ P
i
, scraped from the dish in
NaCl ⁄ P
i
, and centrifuged at 160 g for 5 min.
The pellet was frozen at )30 °C until use. Frozen pellets
were resuspended in binding buffer (50 mm Tris ⁄ HCl pH
7.4 containing 120 mm NaCl, 5 mm KCl, 4 mm MgCl
2
and
1mm EDTA) using an Ultra Turrax homogenizer. The
Dopamine D
1
and D
2
receptors dimerization S. Łukasiewicz et al.
770 FEBS Journal 276 (2009) 760–775 ª 2008 The Authors Journal compilation ª 2008 FEBS
homogenates were centrifuged twice at 30 000 g for 10 min.
[

3
H]spiperone ranging from 0.01 to 4 nm. Nonspecific
binding was assessed by the addition 10 lm cis-(Z)-flu-
pentixol (Lundbeck, Copenhagen, Denmark) for the dopa-
mine D
1
receptor or 50 lm butaclamol (Research
Biochemicals Inc., Natick, MA, USA) for the dopamine D
2
receptor. Tubes were incubated either for 90 min at room
temperature ([
3
H]SCH23390) or for 30 min at 37 °C
([
3
H]spiperone), then binding was terminated by rapid fil-
tration through glass fiber filters (GF ⁄ C, Whatman). The
filters were washed four times with 5 mL of ice-cold wash-
ing buffer (50 mm Tris ⁄ HCl pH 7.4), and the amount of
bound radioactivity was determined by liquid scintillation
counting (Beckman LS 650).
Radioligand binding parameters, K
d
and B
max
, were esti-
mated using the graphpad prism 2.0 curve-fitting program
(GraphPad Software, San Diego, CA, USA).
Fluorescence spectroscopy measurements
Spectrofluorymetric measurements of the cell suspensions

TCSPC measurements were performed using a Nicon
Eclipse TE-2000 inverted fluorescence microscope (Precoptic
Co., Warsaw, Poland). The specimen was excited with the
diode pulse laser (Horiba, Jobin Yvon IBH S.A.S.) at
434 nm with 1 MHz repetition. Fluorescence emission was
recorded by a picosecond detector, TBX-04 (Horiba, Jobin
Yvon IBH S.A.S.). The Jobin Yvon IBH data station and
the das 6 software were used for data acquisition and decay
analysis.
Two fluorescence lifetime standards, p-terphenyl and
erythrosine B, that have single exponential decays (p-ter-
phenyl in cyclohexane: 980 ps – sd 30 ps, erythrosin in
methanol: 470 ps – sd 20 ps and erythrosin in water 89 ps –
sd 3 ps) were used to test our lifetime instrumentation. The
obtained lifetimes agree very well with the ones reported by
Boens et al. [72].
Cells dedicated to TCSPC experiments were grown on cov-
erslips. The fluorescence decay from single cells transfected
with fusion protein constructs was measured using a ·60
objective and dichroic beam splitter at 455 nm, combined
with an emitter cut off filter > 475 nm. The excitation pulse
diode laser profile, required for deconvolution analysis, was
measured on the diluted glycogen using the fluor cube with
400 nm dichroic beam splitter only. All measurements were
performed at 37 °C. Cells were incubated in the same iso-
tonic buffer as used for fluorescence spectra measurements.
During each experiment, fluorescence decay from at least 10
cells on the coverslip was measured.
Each fluorescence decay measurement was analyzed with
the multiexponential model given by:

Á s
2
i
P
i
a
i
Á s
i
ð2Þ
The average efficiency of energy transfer ÆEæ was calcu-
lated from the average fluorescence lifetime of donor in the
absence Æs
D
æ or presence Æs
DA
æ of an acceptor from:
E
hi
¼ 1 À
s
DA
hi
s
D
hi
ð3Þ
Confocal microscopy
HEK293 cells grown on cover slips were transiently trans-
fected with the cDNA encoding the fluorescently labeled

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P
i
ðRi À RavÞ
2
Á
P
i
ðG
i
À GavÞ
2
r
ð4Þ
where Ri and Gi are the red and green intensities of voxel
I, respectively, and Rav and Gav the average values of Ri
and Gi, respectively.
It is used for describing the correlation of the intensity
distributions between red and green component of each
dual-channel image. Pearson’s correlation coefficients were
calculated from randomly selected parts of the image
(membrane signal) from individual cells cotransfected with
different construct combinations (wild-type or mutant fluor-
escently tagged D
1
and D
2
receptor protein). The average
intensity of the fluorescence signal was measured for every
image in a determined individual area of interest free of cell

gic receptors is controlled by heterodimerization with
alpha1B-adrenergic receptors. J Biol Chem 279, 15541–
15549.
5 Lopez-Gimenez JF, Canals M, Pediani JD & Milligan
G (2007) The alpha1b-adrenoceptor exists as a higher-
order oligomer: effective oligomerization is required for
receptor maturation, surface delivery, and function.
Mol Pharmacol 71, 1015–1029.
6 Margeta-Mitrovic M, Jan YN & Jan LY (2000) A traf-
ficking checkpoint controls GABA(B) receptor heterodi-
merization. Neuron 27, 97–106.
7 Minneman KP (2007) Heterodimerization and surface
localization of G protein coupled receptors. Biochem
Pharmacol 73, 1043–1050.
8 Salahpour A, Angers S, Mercier JF, Lagace
´
M, Marullo
S & Bouvier M (2004) Homodimerization of the beta2-
adrenergic receptor as a prerequisite for cell surface
targeting. J Biol Chem 279, 33390–33397.
9 Breit A, Lagace
´
M & Bouvier M (2004) Hetero-oligo-
merization between beta2- and beta3-adrenergic recep-
tors generates a beta-adrenergic signaling unit with
distinct functional properties. J Biol Chem 279, 28756–
28765.
10 Mellado M, Rodrı
´
guez-Frade JM, Vila-Coro AJ, Fern-

772 FEBS Journal 276 (2009) 760–775 ª 2008 The Authors Journal compilation ª 2008 FEBS
14 So CH, Varghese G, Curley KJ, Kong MM, Alijania-
ram M, Ji X, Nguyen T, O’Dowd BF & George SR
(2005) D1 and D2 dopamine receptors form heteroolig-
omers and cointernalize after selective activation of
either receptor. Mol Pharmacol 68, 568–578.
15 Terrillon S & Bouvier M (2004) Roles of G-protein-
coupled receptor dimerization. EMBO Rep 5, 30–34.
16 Cvejic S & Devi LA (1997) Dimerization of the delta
opioid receptor: implication for a role in receptor inter-
nalization. J Biol Chem 272, 26959–26964.
17 Breitwieser G (2004) G protein-coupled receptor oligo-
merization: implications for G protein activation and
cell signaling. Circ Res 94, 17–27.
18 Kroeger KM, Pfleger KD & Eidne KA (2003) G-pro-
tein coupled receptor oligomerization in neuroendocrine
pathways. Front Neuroendocrinol 24, 254–278.
19 Lee SP, O’Dowd BF, Rajarm RD, Nguyen T & George
SR (2003) D2 dopamine receptor homodimerization is
mediated by multiple sites of interaction, including an
intermolecular interaction involving transmembrane
domain 4. Biochemistry 42, 11023–11031.
20 Nemoto W & Toh H (2005) Prediction of interfaces for
oligomerizations of G-protein coupled receptors. Pro-
teins 58, 644–660.
21 Romano C, Yang WL & O’Malley KL (1996) Metabo-
tropic glutamate receptor 5 is a disulfide-linked dimer.
J Biol Chem 271, 28612–28616.
22 Berthouze M, Rivail L, Lucas A, Ayoub MA, Russo O,
Sicsic S, Fischmeister R, Berque-Bestel I, Jockers R &

DG, Barret C & Bouvier M (1996) A peptide derived
from a beta2-adrenergic receptor transmembrane
domain inhibits both receptor dimerization and activa-
tion. J Biol Chem 271, 16384–16392.
29 Hernanz-Falco
´
n P, Rodrı
´
guez-Frade JM, Serrano A,
Juan D, del Sol A, Soriano SF, Roncal F, Go
´
mez L,
Valencia A, Martı
´
nez-A C et al.
(2004) Identification of
amino acid residues crucial for chemokine receptor
dimerization. Nat Immunol 5, 216–223.
30 Gouldson PR, Higgs C, Smith RE, Dean MK, Gkoutos
GV & Reynolds CA (2000) Dimerization and domain
swapping in G-protein-coupled receptors: a computa-
tional study. Neuropsychopharmacology 23, 60–77.
31 Overton MC, Chinault SL & Blumer KJ (2003) Oligo-
merization, biogenesis, and signaling is promoted by a
glycophorin A-like dimerization motif in transmem-
brane domain 1 of a yeast G protein-coupled receptor.
J Biol Chem 278, 49369–49377.
32 AbdAlla S, Zaki E, Lother H & Quitterer U (1999)
Involvement of the amino terminus of the B(2) receptor
in agonist-induced receptor dimerization. J Biol Chem

40 Hebert TE & Bouvier M (1998) Structural and func-
tional aspects of G protein-coupled receptor oligomeri-
zation. Biochem Cell Biol 76, 1–11.
41 Woods AS & Ferre S (2005) Amazing stability of the
arginine–phosphate electrostatic interaction. J Proteome
Res 4, 1397–1402.
S. Łukasiewicz et al. Dopamine D
1
and D
2
receptors dimerization
FEBS Journal 276 (2009) 760–775 ª 2008 The Authors Journal compilation ª 2008 FEBS 773
42 Jackson SN, Wang HYJ, Yergey A & Woods AS
(2006) Phosphate stabilization of intermolecular interac-
tions. J Proteome Res 5, 122–126.
43 Ciruela F, Burguen
˜
o J, Casado V, Canals M, Marcelli-
no D, Goldberg SR, Bader M, Fuxe K, Agnati LF,
Lluis C et al. (2004) Combining mass spectrometry and
pull-down techniques for the study of receptor hetero-
merization. Direct epitope–epitope elostratic interac-
tions between adenosine A2A and dopamine D2
receptors. Anal Chem 76, 5354–5363.
44 Woods AS, Ciruela F, Fuxe K, Agnati LF, Lluis C,
Franco R & Ferre S (2005) Role of electrostatic interac-
tion in receptor-receptor heteromerization. J Mol
Neurosci 26, 125–132.
45 Aizman O, Brismar H, Uhlen P, Zettergren E, Levey
AI, Forssberg H, Greengard P & Aperia A (2000)

51 Elangovan M, Day RN & Periasamy A (2002) Nanosec-
ond fluorescence resonance energy transfer-fluorescence
lifetime imaging microscopy to localize the protein inter-
actions in a single living cell. J Microsc 205, 3–14.
52 Łukasiewicz S, Błasiak E, Faron-Go
´
recka A, Polit A,
TworzydłoM,Go
´
recki A, Wasylewski Z & Dziedzicka-
Wasylewska M (2007) Fluorescence studies of homooli-
gomerization of adenosine A
2A
and serotonin 5-HT
1A
receptors reveal the specificity of receptor interactions
in the plasma membrane. Pharmacol Rep 59, 379–392.
53 Angers S, Salahpour A & Bouvier M (2002) Dimeriza-
tion: an emerging concept for G protein-coupled recep-
tor ontogen and function. Annu Rev Pharmacol Toxicol
42, 409–435.
54 Milligan G (2007) G protein-coupled receptor dimerisa-
tion: molecular basis and relevance to function. BBA
1768, 825–835.
55 O’Dowd BF, Ji X, Alijaniaram M, Rajaram RD, Kong
MM, Rashid A, Nguyen T & George SR (2005) Dopa-
mine receptor oligomerization visualized in living cells.
J Biol Chem 80, 37225–37235.
56 Rashid AJ, So CH, Kong MMC, Furtak T, El-Ghundi
M, Cheng R, O’Dowd BF & George SR (2007) D1–D2

63 Michelsen K, Yuan H & Schwappach B (2005) Hide
and run. Arginine-based endoplasmic-reticulum-sorting
motifs in the assembly of heteromultimeric mebrane
proteins. EMBO Rep 6, 717–722.
64 Ren Z, Riley NJ, Garcia EP, Sanders JM, Swanson GT
& Marshall J (2003) Multiple trafficking signals regulate
kainate receptor KA2 subunit surface expression.
J Neurosci 23, 6608–6616.
65 Hardt B, Kalz-Fu
¨
ller B, Aparicio R, Vo
¨
lker C & Bause
E (2003) (Arg)
3
within the N-termianl domain of gluo-
cosidase I contains ER targeting information but is not
required absolutely for ER localization. Glycobiology
13, 159–168.
66 Binda AV, Kabbani N, Lin R & Levenson R (2002) D2
and D3 receptor cell surface localization mediated by
interaction with protein 4.1N. Mol Pharmacol 62, 507–
513.
67 Yamamoto S, Kawamura K & James TN (1998)
Intracellualr distribution of adenylate cyclase in human
cardiocytes determined by electron microscopic cyto-
chemistry. Microsc Res Tech 40, 479–487.
Dopamine D
1
and D

receptors dimerization
FEBS Journal 276 (2009) 760–775 ª 2008 The Authors Journal compilation ª 2008 FEBS 775