Journal of Immunological Methods 370 (2011) 14–23

Contents lists available at ScienceDirect

Journal of Immunological Methods

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j i m

Research paper

Novel antigen design for the generation of antibodies to

G-protein-coupled receptors
K. Larsson a, C. Hofström b, C. Lindskog c, M. Hansson d, P. Angelidou d, T. Hökfelt e, M. Uhlén a,
H. Wernérus d, T. Gräslund b, S. Hober a,⁎
a
Division of Proteomics, School of Biotechnology, KTH/AlbaNova University Center, Stockholm, Sweden
b
Division of Molecular Biotechnology, School of Biotechnology, KTH/AlbaNova University Center, Stockholm, Sweden
c
Department of Genetics and Pathology, The Rudbeck Laboratory, Uppsala University, Uppsala, Sweden
d
Atlas Antibodies, AlbaNova University Center, Stockholm, Sweden
e
Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden

a r t i c l e i n f o a b s t r a c t

Article history: Antibodies are important tools for the study of G-protein-coupled receptors, key proteins in
Received 27 December 2010 cellular signaling. Due to their large hydrophobic membrane spanning regions and often very
Received in revised form 5 May 2011 short loops exposed on the surface of the cells, generation of antibodies able to recognize the
Accepted 5 May 2011
receptors in the endogenous environment has been difficult. Here, we describe an antigen-
Available online 12 May 2011
design method where the extracellular loops and N-terminus are combined to a single antigen for
generation of antibodies specific to three selected GPCRs: NPY5R, B2ARN and GLP1R. The design
Keywords: strategy enabled straightforward antigen production and antibody generation. Binding of the
GPCR
antibodies to intact receptors was analyzed using flow cytometry and immunofluorescence based
Antibody
confocal microscopy on A-431 cells overexpressing the respective GPCR. The antibody–antigen
Antigen design
interactions were characterized using epitope mapping, and the antibodies were applied in
immunohistochemical staining of human tissues. Most of the antibodies showed specific binding
to their respective overexpressing cell line but not to the non-transfected cells, thus indicating
binding to their respective target receptor. The epitope mapping showed that sub-populations
within the purified antibody pool recognized different regions of the antigen. Hence, the genetic
combination of several different epitopes enables efficient generation of specific antibodies with
potential use in several applications for the study of endogenous receptors.
© 2011 Elsevier B.V. All rights reserved.

1. Introduction with a diverse selection of ligands, such as neurotransmitters,

G-protein-coupled receptors (GPCRs) belong to a protein
family characterized by seven transmembrane helices con-
nected by loops, an extracellular N-terminal and an intracellular
C-terminal tail (Fig. 1A). This family of cellular receptor proteins
is the largest protein family with more than 1000 members
(Wess, 1997), regulating cellular function through interactions
Abbreviations: GPCR, G-protein-coupled receptor; B2AR, beta 2-adrenergic
receptor; GLP1R, glucagon-like peptide 1 receptor; NPY5R, neuropeptide Y
receptor 5; ABP, albumin binding protein; HPA, Human Protein Atlas.
⁎ Corresponding author. Tel.: + 46 8 5537 8330; fax: +46 8 5537 8481.
E-mail address: sophia.hober@biotech.kth.se (S. Hober).
hormones, light, odors and taste (Lundström, 2005).
Due to the regulating activities, GPCRs are very interesting
as drug targets and currently 60–70% of the drug development
is focused on GPCRs (Lundström, 2005). However, because of
difficulties in production and purification of these membrane
proteins (Lundström, 2005; Sarkar et al., 2008), only a handful
of solved structures exists in the Protein Data Bank, and thus
information valuable for the drug development process is often
lacking (Berman et al., 2000). In order to further understand
and study GPCRs, genetic fusion to tags has commonly been
used. To this end, small tags like His6-, HA- or Flag have been
expressed together with the target receptor (Jongsma et al.,
2007). Hence, monoclonal antibodies recognizing the tag can

0022-1759/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jim.2011.05.001
 K. Larsson et al. / Journal of Immunological Methods 370 (2011) 14–23
A bodies were purified using the antigen as affinity ligand and
B Gene Collection via geneservice (glucagon peptide 1 receptor,
His6-ABP
Linker (GSSSG)
Fig. 1. Schematic representation of an intact G-protein-coupled receptor and
outline of the antigen used for generation of GPCR-specific antibodies. (A) A
seven transmembrane domain core is characteristic for all GPCRs. The N-
terminus together with the extracellular loops is responsible for ligand binding
and the intracellular loops and C-terminus for intracellular signaling. (B) The
extracellular loops and N-terminus were assembled to a single antigen, separated
by a polar linker (GSSSG) and N-terminally fused to a purification/immunization
tag, His6-ABP.
be utilized to elucidate localization and function of the
receptors. Another approach is to express the receptor fused
to a fluorescent group, such as green fluorescent protein (GFP)
(Castro et al., 2005). In both these methods a genetic alteration
is made by adding a protein tag, a procedure that may affect the
behavior of the receptor regarding localization, quantity and
activity. Therefore, specific antibodies recognizing the native
forms of GPCRs are invaluable tools for the study of these cell
surface receptors and their function (Michel et al., 2009).
Antibodies are normally produced by immunization of an
animal with the pure protein of interest. Due to the inherent
problems in the production and purification of GPCRs, the use
of the intact GPCR-proteins as antigen has been difficult.
Hence, most antibodies recognizing GPCRs have so far been
generated using synthetic peptide fragments of the receptor
protein as antigens (Gupta and Devi, 2006; Mackrill, 2004).
Zhang et al. (2004) further developed this strategy by
synthetically producing cyclic peptides, thought to mimic
the extracellular loops of the CCR5 receptor, for the selection
of single-chain Fv (scFv) fragments. These approaches for
production of antigens have given rise to functional anti-
bodies and scFv-fragments.
To facilitate the production of antigens and allow the
exposure of optional loop regions, we present an alternative
method for antigen design of GPCRs. This antigen design
strategy has the potential to be used as a general method for
the generation of GPCR specific antibodies. As a proof of concept,
three different receptors were chosen, and the N-terminal tail
together with the three extracellular loops were genetically
linked together to form a continuous antigen, where the
different segments were separated by polar and flexible
glycine/serine (GSSSG) containing linkers. The acquired anti-
15
successfully validated for target specificity in a number of
different applications such as flow cytometry, immunofluores-
cence-based confocal microscopy and immunohistochemistry
using human cell lines and tissues.
2. Materials and methods
2.1. Materials
The oligonucleotides for assembly of the GPCR constructs
were synthesized by Thermo Electron Corporation and Operon.
All primers were purchased from Thermo Fisher Scientific. Full-
length GPCR cDNA clones were acquired from the Mammalian
GLP1R, AccNo: BC113493 and beta2-adrenogenic receptor,
B2AR, AccNo: BC063486) and imaGenes (neuropeptide Y
receptor, NPY5R, AccNo: BC042416). Commercial antibodies
used were: rabbit anti-B2AR (Abcam, ab13300); mouse anti-
GLP1R (R&D systems, MAB28141); rabbit anti-GLP1R (Strategic
Diagnostics, GA1940) and rabbit anti-GLP1R (Chemicon,
AB9433). For NPY5R, two additional rabbit antibodies produced
by the Human Protein Atlas (HPA) project were used:
HPA013790 and HPA014056.
2.2. Plasmid construction
The extracellular parts of the GPCRs were cloned in series,
separated by a polar five amino acid linker—GSSSG. The GPCR
constructs were generated using solid-phase gene assembly
(Nguyen et al., 1994; Ståhl et al., 1993), where a longer DNA
sequence was built on a solid support from shorter oligonucle-
otides by parallel ligations. The oligonucleotides (Thermo
Electron Corporation and Operon) were 50 bases long with
10 bases overlap. In brief, 300 μg of streptavidin-coated
DynaBeads (DynaBeads M-280 streptavidin (10 mg/ml),
Dynal Biotech) was used to couple 80 pmol of the first
biotinylated segment. After incubation for 1 h at RT and
subsequent washing, 40 pmol each of the remaining segments
was mixed for 30 min at 65 °C. T4 DNA ligase (Invitrogen) was
then added, and the ligation reaction proceeded for 4 h at RT.
The assembled constructs were amplified using nested PCR
with AmpliTaq Gold DNA polymerase (Applied Biosystems).
Due to the long N-terminal region of GLP1R, it was PCR
amplified separately and ligated to the rest of the construct
using a naturally occurring PstI site. The assembled GPCR
constructs were NotI/AscI digested and ligated into the
expression vector pAff8c (Larsson et al., 2000) in frame with a
hexahistidine (His6) purification tag, as well as an Albumin
binding protein (ABP) derived from streptococcal protein G
used for immunization purposes.
2.3. Protein production and immunization
BL21(DE3) cells, containing the expression vector encod-
ing the antigen, were grown, harvested and lysated essen-
tially as formerly described (Agaton et al., 2003), and the
lysed cell cultures were IMAC-purified under denaturing
conditions using an automated set-up for affinity chroma-
tography (Steen et al., 2006). Moreover, the fusion proteins
were analyzed with SDS-PAGE (BioRad), and the
16 K. Larsson et al. / Journal of Immunological Methods 370 (2011) 14–23
concentrations were determined using a BCA™ Protein Assay
Kit (Pierce). The dual affinity handle protein, His6ABP, was
also expressed and purified from the pAff8c vector as
previously described (Agaton et al., 2004). The three GPCR
antigens were immunized in two New Zealand white rabbits
each (Young In Frontier, South Korea) in accordance with the
national guidelines. The primary immunization was per-
formed using 200 μg of antigen in Freund's complete
adjuvant, followed by three booster injections of 100 μg
each in Freund's incomplete adjuvant with four week in-
tervals. The rabbit serum was collected two weeks after the
final booster injection.
2.4. Antibody purification and characterization
Antisera from immunized animals were purified by a two-
step immunoaffinity protocol (Agaton et al., 2004). Initially,
12 ml serum was filtrated and buffered with PBS. Subse-
quently, antibodies specific for the tag protein were depleted
from the sample using vector protein, His6ABP, coupled to
NHS-activated Sepharose™ 4 Fast Flow matrix (Amersham
Biosciences). The depleted sera samples were additionally
affinity purified on prepacked NHS-activated one ml HiTrap
columns (GE Healthcare), coupled with the respective antigen,
using the ÄKTAxpress chromatography system (GE Healthcare).
After subsequent washing of the columns, antibodies were
eluted with a low pH buffer (0.2 M glycine, 1 mM EGTA, pH 2.5),
followed by a buffer exchange to 1× PBS. Finally, glycerol and
NaN3 were added to a final concentration of 50% and 0.02%,
respectively. Binding to respective antigen and lack of cross-
reactive binding to tag sequence and linker region was confirmed
by Western blot. Briefly, the respective antigen and one of the
other non-related antigens were separated on SDS-PAGE
gradient gels (10–20% Criterion, Bio-Rad Laboratories) and
transferred to PVDF membranes (Criterion Gel™ Blotting
Sandwiches, Bio-Rad Laboratories). Subsequent to blocking of
the membrane in blocking buffer (5% dry milk, 0.5% Tween20, 1×
TBS; 0.1 M Tris–HCl, 0.5 M NaCl), the anti-GPCR antibodies,
diluted 1/300–1/600, were added and the membrane was
incubated for 1 h. Secondary HRP conjugated goat-anti rabbit
antibody (DakoCytomation), diluted 1/3000, was added and
following incubation for 1 h, the membranes were soaked in
substrate solution (Immobilon Western Chemiluminescent HRP
Substrate, Millipore Corporation), and binding events were
visualized in a Chemidoc CCD camera system (Bio-Rad
Laboratories).
2.5. Generation of GPCR-expressing cell lines
The HEK 293T cell line was obtained from LGC Promochem
and the A431 cell line from DSMZ. Cells were cultivated in 5%
CO2 at 37 °C in Dulbecco's modified Eagle's media (DMEM,
Invitrogen) supplemented with 10% fetal bovine serum (FBS)
and 1% Antibiotic–Antimycotic solution. For selection purposes,
the medium of infected cell lines contained blasticidin
(20 μg/ml media). B2AR, GLP1R and NPY5R were overexpressed
in the human epidermoid carcinoma cell line, A-431, using a
lentiviral gene delivery system (Suter et al., 2006). Full-length
cDNA clones of B2AR, GLP1R and NPY5R were purchased from
the Mammalian Gene Collection (Genservice and imaGenes,
GmbH). The gene fragments were amplified with PCR using
Phusion High-Fidelity DNA polymerase (Finnzymes) and
furthermore cloned into the lentivector 2K7bsd (Suter et al.,
2006) using Gateway® cloning (Invitrogen). The 2K7bsd-GPCR
vectors were thereafter transfected to HEK293-T cells together
with packaging and envelope plasmids using a Calcium chloride
transfection procedure, as previously described (Pear et al.,
1993; Vernet et al., 2009). The produced viral particles were in
turn used to infect A-431 cells in the presence of Polybrene
(8 μg/ml).
2.6. Flow cytometry analysis
All generated antibodies as well as several commercial
antibodies were analyzed for binding to the full-length
protein using the overexpressing cell lines and non-trans-
fected controls on a flow cytometer. Briefly, the cells were
trypsinated, washed with PBS and resuspended in PBS
containing 1% bovine serum albumin (PBSB). 200,000 cells
were used for staining with each antibody. Primary anti-
bodies were diluted with the PBS containing BSA to a final
concentration of 5 μg/ml and secondary antibodies to
10 μg/ml. The cells were incubated with 70 μl of primary
antibody for 1 h at RT followed by washing with PBSB. After
incubation with 70 μl of secondary antibody (Alexa Fluor 488
goat anti-rabbit or goat anti-mouse, Molecular probes) and
subsequent wash with PBSB the cells were analyzed on a FACS
Vantage SE (BD Biosciences). Analyzed samples included: no
antibodies on normal A-431 cells, secondary antibodies on
normal A-431 control cells, primary and secondary antibodies
on normal A-431 cells and primary and secondary antibodies
on GPCR overexpressing A-431 cells.
2.7. Immunofluorescence and immunohistochemistry analysis
To confirm that the antibodies bind a protein on the cell
surface, confocal fluorescence microscopy was utilized, essen-
tially as described previously (Barbe et al., 2008). Briefly, glass
bottom plates (Whatman) were coated with fibronectin
(Sigma) and seeded with 10,000 cells per well, grown for 5 h
and fixed with ice-cold 4% paraformaldehyde (Sigma). After
permeabilization with 0.1% Triton X-100 (Sigma) the cells were
incubated with the primary antibody as well as a mouse anti-
tubulin antibody (Abcam) overnight at 4 °C. Moreover, the cells
were incubated for 1.5 h with secondary antibodies, goat anti-
rabbit IgG conjugated with Alexa Fluor 488 and rabbit anti-
mouse IgG Alexa Fluor 555 (Molecular Probes) and counter-
stained with DAPI, a nuclear probe. Image acquisition was
performed with a LSM 510 Meta confocal laser-scanning
microscope (Carl Zeiss GmbH). For the immunofluorescence
staining of human brain tissue, unfixed cryostat sections were
incubated with antibodies NPY5R:1 and NPY5R:2 against the
NPY5R (dilution 1/1000) and processed according to the TSA+
method (Adams, 1992) using a commercial kit and with green
fluorescein isothiocyanate (FITC) as marker, as described in
detail elsewhere (Mulder et al., 2009). Immunohistochemical
staining of human pancreatic islets with the GLP1R-specific
antibodies (GLP1R:1, GLP1R:2, MAB28141, GA1940 and
AB9433) was performed using formalin fixed, paraffin embed-
ded tissue samples. The staining was performed according to
previously described methods (Nilsson et al., 2005).
 K. Larsson et al. / Journal of Immunological Methods 370 (2011) 14–23
2.8. RT-PCR
The expression of NPY5R, B2AR and GLP1R in the transduced
cell lines was investigated on mRNA level. Overexpressing cells
and normal A-431 cells were trypsinated, washed with PBS and
frozen as pellets in −80 °C. Messenger RNA was extracted from
cells using RNeasy Mini Kit (Qiagen) with on-column digestion
of chromosomal DNA. Prior to amplification, the mRNA was
converted to cDNA using Super Script III (Invitrogen). PCR
amplification was conducted using Phusion High-Fidelity DNA
polymerase (Finnzymes) with specific primers. Each GPCR-
specific primer pair was used for amplification of both normal
A-431 cells and the respective overexpressing cell line.
Amplification of all cDNA fragments was also conducted
utilizing β-actin-specific primers, generating a PCR product of
70 bp, for normalization.
2.9. Epitope mapping of GPCR antigens
Peptides covering the extracellular loops and N-terminus
were designed for all GPCR antigens. The total 25 biotinylated
peptides (PEPscreen library, Sigma-Aldrich) were between 7
and 20 amino acids long depending on loop/N-terminal size.
For longer loops overlapping peptides were designed, except
for the N-terminus of GLP1R, which was kept intact and
expressed as a 118 amino acid long peptide. All peptides were
resolved in 80% DMSO and diluted in BRE (Blocking reagent
for ELISA, Roche) to a final concentration of 20 μM. Coupling
of the Luminex beads was essentially done as previously
described (Larsson et al., 2009). Neutravidin (Pierce) diluted
to 0.1 μg/μl in MES was immobilized on 106 carboxylated beads
(COOH Microspheres, Luminex-Corp.) per color-coded ID,
according to the manufacturer's protocol. To couple the
biotinylated peptides about 5 × 10 5 neutravidin-coupled
beads per ID were used. In total 100 μl of 20 μM diluted peptide
was utilized in the coupling reaction and following incubation
for 1 h and subsequent washing steps, the Luminex beads were
stored in azide containing BRE buffer. Prior to analysis of the
antibodies, a bead stock was made containing 10 μl of each
peptide-coupled bead ID. All antibodies were diluted to 50 ng/
ml of which 45 μl was added to 5 μl of the bead stock, incubated
for 1 h and washed in PBST. Following incubation with a
secondary R-Phycoerythrine labeled anti-rabbit IgG (0.5 μg/ml,
Jackson ImmunoResearch) and additional washing, binding
events were measured using a Luminex LX200 instrument with
xPONENT software. For each experiment 50 events per bead ID
were counted and the median fluorescence intensity (MFI) was
used. Each experiment was performed in triplicates.
3. Results
In this study, we have developed a novel strategy for
antigen design to generate specific antibodies against GPCRs.
Three GPCRs were selected for this study: the neuropeptide Y
receptor 5 (NPY5R), the beta 2 adrenergic receptor (B2AR)
and the glucagon-like peptide 1 receptor (GLP1R). By
combining the N-terminal tail with the three extracellular
loops to a single antigen (Fig. 1, Supplementary) an easily
produced and coherent antigen only displaying the surface
accessible parts of the GPCR receptor was created. Immuni-
zation of two rabbits with each of the acquired antigens and
17
subsequent purification of the polyclonal antibodies resulted
in antibodies able to specifically recognize the desired target
protein.
3.1. Antigen production and antibody purification
To enable simple and efficient protein purification, the
antigen-constructs were cloned in fusion to a hexahistidine
(His6) tag. Moreover, due to its immunopotentiating effects,
an albumin binding protein (ABP) was inserted between the
His6-tag and the antigen (Sjölander et al., 1997). The
generated constructs were expressed in Escherichia coli and
purified by IMAC utilizing the His6-tag. Purified proteins were
used for immunization of two rabbits, giving two polyclonal
antisera for each receptor. To purify antibodies targeting the
three receptors included in this study and thereby achieving
specific antibodies, a stringent two-step affinity purification
method was applied, where the antigen was used as affinity
ligand (Nilsson et al., 2005). Antibody yield varied from 0.2 mg
to 1.0 mg with an average of 0.5 mg of pure antibody from
12 ml of antiserum. Binding to the antigen and lack of cross-
reactive binding to the tag sequence and linker regions was
confirmed by Western blot analysis (data not shown). The
antibodies will further on be referred to as NPY5R:1 and
NPY5R:2 for the antibodies generated toward the neuropeptide
Y receptor 5; B2AR:1 and B2AR:2 for the beta 2 adrenergic
receptor antibodies and GLP1R:1 and GLP1R:2 for the glucagon-
like peptide 1 receptor antibodies. All antibodies used, both
those produced in this study and acquired commercially, are
listed in Table 1.
3.2. Cell line generation and flow cytometric analysis
In order to analyze if the GPCR-targeting antibodies were
able to recognize the full-length receptors as expressed on the
surface, three cell lines overexpressing NPY5R, B2AR and
GLP1R, respectively, were constructed. RT-PCR was performed
on the respective cell lines and all three cell lines showed a high
level of receptor specific transcripts (Fig. 2A). Subsequently,
several antibodies were used for analysis of the surface exposed
GPCRs with flow cytometry to assess differences between
differently produced antibodies. The resulting histograms show
that all NPY5R antibodies (NPY5R:1, NPY5R:2, HPA013790 and
HPA014056) bind to the A-431 cells overexpressing NPY5R
(A-431NPY5R) but not to non-transfected A-431 cells (A-431)
(Fig. 2B). For B2AR only one antibody, B2AR:1, generated a shift
in fluorescence, hence indicating binding to the receptor.
B2AR:2 and ab13300 showed no elevated fluorescence signals
when analyzing the overexpressing A-431 cells (A-431B2AR)
(Fig. 2C). All antibodies generated toward GLP1R (GLP1R:1 and
GLP1R:2) as well as the commercial antibodies resulted in
binding to the cells overexpressing GLP1R (A-431GLP1R). There
was no binding of the antibodies to A-431 cells except for the
mouse antibody MAB28141 that displayed a slight shift in
fluorescence (Fig. 2D). In general, the flow cytometry analyses
gave rise to broad peaks, indicating different levels of receptor
expression. Depending on genomic position of the integration
of the gene encoding the GPCR, expression of the transgene
may vary due to chromatin packing.
18 K. Larsson et al. / Journal of Immunological Methods 370 (2011) 14–23
Table 1
Antibodies used in this study are listed together with their corresponding
receptors.
GPCR Antibody
NPY5R NPY5R:1
NPY5R:2
HPA013790
HPA014056
B2AR B2AR:1
B2AR:2
ab13300
GLP1R GLP1R:1
GLP1R:2
AB9433
GA1940
MAB28141
3.3. Immunohistochemical and immunofluorescence analyses
To further validate binding to the full-length protein and
assure specificity, i.e. no unspecific staining in other cell
compartments besides the membrane or membrane-associated
compartments (e.g. Golgi), immunofluorescence analysis with
confocal microscopy was performed. Included in the study were
the GPCR overexpressing A-431 cell lines (A-431NPY5R, A-431B2AR
and A-431GLP1R) as well as non-transfected A-431 cells.
All antibodies directed toward NPY5R (NPY5R:1, NPY5R:2,
HPA013790 and HPA014056) showed staining of the plasma
membrane, in general more pronounced in the cell junctions.
Additional Golgi-like staining was seen as exemplified in Fig. 3A
(NPY5R:2). In the control samples (A-431 cells), only very faint
staining could be observed (Fig. 3B), except for antibody
NPY5R:1 which generated a nuclear staining pattern, not
detectable in A-431NPY5R cells (data not shown). The B2AR:1
antibody exhibited binding to the plasma membrane (Fig. 3C).
Additional staining, likely in the Golgi apparatus, was detected
but less pronounced than for NPY5R antibodies. No significant
staining of B2AR:1 was detected in A-431 cells (Fig. 3D).
Staining with B2AR:2 resulted in a fluorescent signal in the
membrane but compared to A-431 cells, no elevated levels
were detected (data not shown). The commercial ab13300 was
completely negative both in A-431B2AR cells and A-431 cells
(data not shown). The antibodies expected to bind to GLP1R
were also used for confocal microscopy. In these cases no
certain membrane staining could be established (Fig. 3E). On
the other hand MAB28141 stained the plasma membrane.
However, none of these antibodies gave staining in non-
transfected A-431 cells (Fig. 3F), which indicates binding to the
overexpressed receptor. Antibody GA1940 gave staining in the
plasma membrane, but also displayed staining in the nucleus of
A-431 cells, a pattern that was not detected in the A-431GLP1R
cell line (data not shown). No staining could be observed for
antibody AB9433, neither in A-431GLP1R cells nor in A-431 cells
(data not shown).
To further assess the specificity of the produced anti-
bodies, tissue sections from parts of the human brain possibly
expressing NPY5R were analyzed by immunofluorescence
microscopy. No detectable staining was obtained with the
NPY5R:1 antibody, but as can be seen in Fig. 4A, the NPY5R:2
antibody gave rise to a large number of fluorescent processes.
Moreover, when staining human pancreatic tissues, known to
express GLP1R (Thorens, 1992) using GLP1R:1, distinct
staining of the plasma membrane was observed in pancreatic
islets. Antibody GLP1R:2 also showed a membranous staining
pattern, although accompanied with an additional cytoplas-
mic signal of the surrounding exocrine glandular cells. The
other antibodies used (GA1940 and AB9433) both gave rise to
a more unspecific staining pattern (Fig. 4B).
3.4. Epitope mapping
In order to characterize the binding of the different
antibodies to the GPCRs, peptides covering the primary sequence
of the antigens were synthesized. The peptides corresponded to
the N-terminus and extracellular loops of the antigens respec-
tively (see Supplementary data for sequences). The biotinylated
peptides differed in length between 7 and 20 amino acids,
depending on the length of the loops and N-terminus. When
loops exceeded 20 amino acids, the sequences were covered
using overlapping peptides, except for the N-terminus of GLP1R,
which was expressed as a continuous 118 amino acid long
peptide. The peptides were immobilized on streptavidin-coated
beads via an N-terminal biotin and the analyses were performed
with a Luminex instrument. All antibodies showed different
patterns of recognition, regarding which part of the antigen they
interacted with. Even though a majority of the response was
directed to the N-terminus, possibly due to the longer amino
acid sequences, this was not always the case.
Antibody NPY5R:1 bound to all segments of the antigen but
with very low signal intensities, while NPY5R:2 generated a peak
with high intensity for the first peptide of the N-terminus
indicating an epitope (Fig. 5A). B2AR:1 recognized the N-
terminus and the second extracellular loop, while B2AR:2
bound both to the second and third extracellular loop as well
as to the N-terminus (Fig. 5B). The third extracellular loop of
B2AR is short, only 7 amino acids, therefore an extra peptide was
inserted where the 7 amino acids were placed two times in a
row, head-to-tail, in order to extend the distance of the possible
epitope from the bead surface. The antibody B2AR:2 showed
binding to this ‘double’ peptide but no binding to the original
one. Binding of GLP1R:1 was only detected toward the N-
terminus, while GLP1R:2 showed binding to the extracellular
loop 3, as well as the N-terminus (Fig. 5C).
4. Discussion
Here we describe a novel method for design of GPCR antigens
and the subsequent generation of antigen-purified antibodies
with successful outcome. This was accomplished by merging the
N-terminus and extracellular loops into a single antigen, thus
facilitating the production compared to full-length GPCR
production, often requiring time-consuming optimization of
expression parameters (Sarkar et al., 2008). By using this
approach it can be decided beforehand whether an antibody
should bind to the extracellular or intracellular regions of the
receptor. Moreover, the immunization with a longer antigen
including several potential epitopes increases the probability of
success in a broader repertoire of applications compared to
peptide immunizations previously used for the generation of
GPCR specific antibodies. In fact, data from the Human Protein
Atlas program (www.proteinatlas.org), where antibodies are
 K. Larsson et al. / Journal of Immunological Methods 370 (2011) 14–23 19

A
B2AR 1002 bp

NPY5R 781 bp

GLP1R 848 bp
+ -
β-actin 70 bp

31
AR

1R
5R

A4
B2

LP
PY

G
N

B NPY5R:1 NPY5R:2 HPA013790 HPA014056

100

0
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10

C B2AR:1 B2AR:2 ab13300

120
Cell count

0
1 2 3 4 1 2 3 4 1 2 3 4
10 10 10 10 10 10 10 10 10 10 10 10

D GLP1R:1 GLP1R:2 AB9433

100

0
1 2 3 4 1 2 3 4 1 2 3 4
10 10 10 10 10 10 10 10 10 10 10 10
GA1940 MAB28141

100
Secondary antibody + A-431 cells
Primary/secondary antibody + A-431 cells
Primary/secondary antibody + over expressing A-431 cells

0
1 2 3 4 1 2 3 4
10 10 10 10 10 10 10 10

Fluorescence intensity

Fig. 2. (A) Detection of GPCR mRNA in the cell lines was performed by RT-PCR of extracted mRNA from overexpressing (+) as well as non-transfected A-431 cells
(−) using specific primers from NPY5R, B2AR and GLP1R respectively. Primers specific for the housekeeping gene β-actin were used to verify equal amount of
sample loading. (B) Flow cytometric analysis of antibodies specific for NPY5R, (C) B2AR and (D) GLP1R. Each histogram shows data obtained from three samples:
non-transfected A-431 cells incubated solely with secondary antibody (thin lines), non-transfected A-431 cells incubated with both primary and secondary
antibodies (dashed lines) and over expressing cells immunostained with the corresponding primary antibody and secondary antibody (thick lines).
20 K. Larsson et al. / Journal of Immunological Methods 370 (2011) 14–23

A B

C D

E F

Fig. 3. Representative immunofluorescent staining of GPCR-overexpressing as well as non-transfected A-431 cells. Antibody NPY5R:2 staining of (A) NPY5R
overexpressing A-431 cells and (B) A-431 cells, B2AR:1 staining of (C) B2AR overexpressing cells and (D) A-431 cells and finally GLP1R:2 staining of (E) GLP1R
overexpressing cells and (F) A-431 cells. The cells were counterstained with the nuclear probe DAPI (blue). Antibody binding is depicted in green. Bar indicates
10 μm.

produced toward the non-redundant set of human proteins,
suggest that immunization with antigens shorter than 50 amino
acids generate significantly lower antibody titers as well as
lower success rate in subsequent applications (unpublished
data). We chose to assemble the antigen using the extracellular
parts of the GPCRs in order to detect the receptor protein on
intact cells, thus allowing research on endogenous receptor in its
native membrane bound form on living cells, but the same design
strategy could also be envisioned for the intracellular parts.
Validation of antibodies in general, and GPCR-specific
antibodies in particular, is of great importance and seldom an
easy task. The specificity of antibodies targeting GPCRs is a
well-known issue and has recently been discussed by Michel
et al.(2009) Since most GPCRs are endogenously expressed at
very low levels (Sarkar et al., 2008) validation of them is
difficult, and therefore we generated three different cell lines,
overexpressing NPY5R, B2AR and GLP1R respectively. Over-
expression of the specific G-protein-coupled receptors was
 K. Larsson et al. / Journal of Immunological Methods 370 (2011) 14–23 21

Fig. 4. Immunofluorescence staining of human brain tissue and immunohistochemical staining of human pancreas. (A) Immunofluorescence micrograph taken from the
dorsal part of the periaqueductal gray in the human mesencephalon after incubation of an unfixed cryostat section with NPY5R:2 antibody. A large number of
fluorescent processes are seen, but no cell bodies. Bar indicates 1 μm. (B) Immunohisotochemical staining of human formalin fixed paraffin embedded pancreatic tissues
using GLP1R antibodies. GLP1R:1 gives a clear staining of plasma membranes (brown) in the pancreatic islets with no staining observed in exocrine glandular cells. No
specific immunohistochemical staining pattern could be detected after incubation with GA1940. The nuclei of the cells are stained with hematoxylin (blue). Bar indicates
100 μm.

successfully examined by RT-PCR, showing high expression in
the transfected cells and no detectable expression in the non-
transfected cells. This permitted validation of the obtained
antibodies with flow cytometry and immunofluorescence-
based confocal microscopy. According to the flow cytometric
analysis, most of the antibodies showed specific binding, i.e.
increased fluorescence intensity, to their respective GPCR-
overexpressing cell line but not to non-transfected control
cells. However, when using B2AR:2 and ab13300 no increased
fluorescence signals could be detected. This could possibly be
explained by low affinity of the antibodies to their target
receptor or that epitopes are hidden or conformational
different on the native receptor compared to the antigen.
Immunofluorescence staining and analysis by confocal mi-
croscopy of overexpressing cells showed that many of the
used antibodies gave staining of the plasma membrane, hence
indicating binding to the membrane-bound target protein.
Moreover, no or very weak staining could be detected in non-
transfected A-431 cells. The antibodies NPY5R:1 and GA1940
showed nuclear staining in non-transfected A-431 cells that
could not be observed in the corresponding overexpressing
cell line (data not shown), indicating unspecific binding in the
absence of the intended target protein. The antibodies
generated against NPY5R and GLP1R were also used for
immunohistochemical and immunofluorescent staining of
human tissues expected to express the target proteins. In
both cases the antibodies showed staining that also supports
the earlier acquired validation data, indicating correct and
selective binding of the antibodies.
The epitope mapping showed that sub-populations of
antibodies within the purified antibody pools recognized
different regions of the respective antigen. Hence, there is a
large potential in this novel strategy. Through the genetic fusion
of many different epitopes, and by avoiding family specific
regions, specific antibodies can potentially be raised with high
success rate. For large-scale production of antibodies toward
GPCRs using synthetic peptides, B-cell epitope prediction would
be a requirement in order to reduce the number of experiments
needed to achieve a good antibody. However, prediction of B-cell
epitopes has recently been demonstrated to be very difficult
(Blythe and Flower, 2005). The results from the epitope mapping
showed that a majority of the response was directed to the N-
22 K. Larsson et al. / Journal of Immunological Methods 370 (2011) 14–23

Fig. 5. Epitope mapping of GPCR antigens using synthetic peptides and Luminex technology. Intensity plots showing binding of (A) NPY5R:1 and NPY5R:2, (B)
B2AR:1 and B2AR:2 and (C) GLP1R:1 and GLP1R:2 to the N-terminus and extracellular loops of the respective antigen. Intensity is measured as median fluorescent
intensity (MFI). For sequences longer than 20 amino acids, overlapping peptides were used. The N-terminal of the GLP1R was represented as a 118 amino acids
long peptide.

terminus, probably due to the longer amino acid sequences,
especially for GLP1R. However, the B2AR:2 antibody almost
exclusively recognized the second and third extracellular loop.
Therefore our method to include all or selected extracellular parts
of the receptor is a promising approach for GPCR-specific
antibody generation. Since there are some inherited limitations
with polyclonal antibodies our strategy to use these combined
antigens could also be envisioned for production of monoclonal
antibodies. The antigen could then be used for immunization, in
order to have a broader range of B-cells to choose from when
producing the hybridoma cells.
Supplementary materials related to this article can be
found online at doi:10.1016/j.jim.2011.05.001
Acknowledgements
The authors would like to thank Dr. J. Galli, Dr. J. Feldwisch
and K. Johansson (Affibody AB) for fruitful discussions and
technical assistance. C. Stadler is acknowledged for assistance
with the confocal microscopy imaging, Dr. David Suter for
providing the 2K7-plasmid and Dr. Erwan LeMaitre for taking
immunofluorescence micrographs. The Knut and Alice Wal-
lenberg Foundation financially supported this work. The
support of the Swedish Research Council (04X-2888) is
gratefully acknowledged.
References
Adams, J.C., 1992. Biotin amplification of biotin and horseradish peroxidase signals
in histochemical stains. J. Histochem. Cytochem. 40 (10), 1457.
Agaton, C., et al., 2003. Affinity proteomics for systematic protein profiling of
chromosome 21 gene products in human tissues. Mol. Cell. Proteomics 2
(6), 405.
Agaton, C., et al., 2004. Selective enrichment of monospecific polyclonal antibodies
for antibody-based proteomics efforts. J. Chromatogr. A 1043 (1), 33.
Barbe, L., et al., 2008. Toward a confocal subcellular atlas of the human
proteome. Mol. Cell. Proteomics 7 (3), 499.
Berman, H.M., et al., 2000. The Protein Data Bank. Nucleic Acids Res. 28 (1), 235.
 K. Larsson et al. / Journal of Immunological Methods 370 (2011) 14–23
Blythe, M.J., Flower, D.R., 2005. Benchmarking B cell epitope prediction:
underperformance of existing methods. Protein Sci. 14 (1), 246.
Castro, M., et al., 2005. Turn-on switch in parathyroid hormone receptor by a
two-step parathyroid hormone binding mechanism. Proc. Natl. Acad. Sci.
U. S. A. 102 (44), 16084.
Gupta, A., Devi, L.A., 2006. The use of receptor-specific antibodies to study G-
protein-coupled receptors. Mt. Sinai J. Med. 73 (4), 673.
Jongsma, M., et al., 2007. Validation of a rapid, non-radioactive method to
quantify internalisation of G-protein coupled receptors. Naunyn-
Schmiedebergs Arch. Pharmacol. 375 (5), 329.
Larsson, M., et al., 2000. High-throughput protein expression of cDNA
products as a tool in functional genomics. J. Biotechnol. 80 (2), 143.
Larsson, K., et al., 2009. Characterization of PrEST-based antibodies towards
human Cytokeratin-17. J. Immunol. Methods 342 (1–2), 20.
Lundström, K., 2005. Structural genomics of GPCRs. Trends Biotechnol. 23 (2), 103.
Mackrill, J.J., 2004. Generation, use, and validation of receptor-selective
antibodies. Methods Mol. Biol. 259, 47.
Michel, M.C., Wieland, T., Tsujimoto, G., 2009. How reliable are G-protein-
coupled receptor antibodies? Naunyn-Schmiedebergs Arch. Pharmacol.
379 (4), 385.
Mulder, J., et al., 2009. Tissue profiling of the mammalian central nervous system
using human antibody-based proteomics. Mol. Cell. Proteomics 8 (7), 1612.
Nguyen, T.N., Uhlén, M., Ståhl, S., 1994. De novo gene assembly using a
paramagnetic solid support. In: Uhlén, M., Hornes, E., Olsvik, Ö. (Eds.),
Advances in Biomagnetic Separation. Eaton Publishing Co., Natick, p. 73.
Nilsson, P., et al., 2005. Towards a human proteome atlas: high-throughput
generation of mono-specific antibodies for tissue profiling. Proteomics 5
(17), 4327.
23
Pear, W.S., et al., 1993. Production of high-titer helper-free retroviruses by
transient transfection. Proc. Natl. Acad. Sci. U. S. A. 90 (18), 8392.
Sarkar, C.A., et al., 2008. Directed evolution of a G protein-coupled receptor
for expression, stability, and binding selectivity. Proc. Natl. Acad. Sci. U. S. A.
105 (39), 14808.
Sjölander, A., et al., 1997. The serum albumin-binding region of streptococcal
protein G: a bacterial fusion partner with carrier-related properties.
J. Immunol. Methods 201 (1), 115.
Ståhl, S., et al., 1993. Solid-phase gene assembly of constructs derived from
the Plasmodium falciparum malaria blood-stage antigen Ag332. Bio-
techniques 14 (3), 424.
Steen, J., et al., 2006. High-throughput protein purification using an
automated set-up for high-yield affinity chromatography. Protein Expr.
Purif. 46 (2), 173.
Suter, D.M., et al., 2006. Rapid generation of stable transgenic embryonic
stem cell lines using modular lentivectors. Stem Cells 24 (3), 615.
Thorens, B., 1992. Expression cloning of the pancreatic beta cell receptor for
the gluco-incretin hormone glucagon-like peptide 1. Proc. Natl. Acad. Sci.
U. S. A. 89 (18), 8641.
Vernet, E., et al., 2009. Affibody-mediated retention of the epidermal growth
factor receptor in the secretory compartments leads to inhibition of
phosphorylation in the kinase domain. N. Biotechnol.
Wess, J., 1997. G-protein-coupled receptors: molecular mechanisms involved
in receptor activation and selectivity of G-protein recognition. FASEB J.
11 (5), 346.
Zhang, Y., et al., 2004. Selection of active ScFv to G-protein-coupled receptor
CCR5 using surface antigen-mimicking peptides. Biochemistry 43 (39),
12575.