Mistic and TarCF as fusion protein partners for functional expression of the cannabinoid receptor 2 in Escherichia coli

Abstract

G protein coupled receptors (GPCRs) are key players in signal recognition and cellular communication making them important therapeutic targets. Large-scale production of these membrane proteins in their native form is crucial for understanding their mechanism of action and target-based drug design. Here we report the overexpression system for a GPCR, the cannabinoid receptor subtype 2 (CB2), in Escherichia coli C43(DE3) facilitated by two fusion partners: Mistic, an integral membrane protein expression enhancer at the N-terminal, and TarCF, a C-terminal fragment of the bacterial chemosensory transducer Tar at the C-terminal of the CB2 open reading frame region. Multiple histidine tags were added on both ends of the fusion protein to facilitate purification. Using individual and combined fusion partners, we found that CB2 fusion protein expression was maximized only when both partners were used. Variable growth and induction conditions were conducted to determine and optimize protein expression. More importantly, this fusion protein Mistic-CB2-TarCF can localize into the E. coli membrane and exhibit functional binding activities with known CB2 ligands including CP55,940, WIN55,212-2 and SR144,528. These results indicate that this novel expression and purification system provides us with a promising strategy for the preparation of biologically active GPCRs, as well as general application for the preparation of membrane-bound proteins using the two new fusion partners described.

Full text

Mistic and TarCF as fusion protein partners for functional expression
of the cannabinoid receptor 2 in Escherichia coli
Ananda Chowdhury
a,b
, Rentian Feng
a,b
, Qin Tong
a,b
, Yuxun Zhang
a,b
, Xiang-Qun Xie
a,b,c,d,
⇑
a
Department of Pharmaceutical Sciences, School of Pharmacy, and Computational Chemical Genomics Screening Center, University of Pittsburgh, Pittsburgh, PA 15260, USA
b
Drug Discovery Institute, University of Pittsburgh, Pittsburgh, PA 15260, USA
c
Departments of Computational Biology and Structural Biology, University of Pittsburgh, Pittsburgh, PA 15260, USA
d
Pittsburgh Chemical Methods and Library Development (CMLD) Center, University of Pittsburgh, Pittsburgh, PA 15260, USA
article info
Article history:
Received 20 October 2011
and in revised form 14 January 2012
Available online 3 March 2012
Keywords:
Mistic
TarCF
Fusion partner
E. coli expression
Cannabinoid receptor 2
abstract
G protein coupled receptors (GPCRs) are key players in signal recognition and cellular communication
making them important therapeutic targets. Large-scale production of these membrane proteins in their
native form is crucial for understanding their mechanism of action and target-based drug design. Here we
report the overexpression system for a GPCR, the cannabinoid receptor subtype 2 (CB2), in Escherichia coli
C43(DE3) facilitated by two fusion partners: Mistic, an integral membrane protein expression enhancer at
the N-terminal, and TarCF, a C-terminal fragment of the bacterial chemosensory transducer Tar at the C-
terminal of the CB2 open reading frame region. Multiple histidine tags were added on both ends of the
fusion protein to facilitate purification. Using individual and combined fusion partners, we found that
CB2 fusion protein expression was maximized only when both partners were used. Variable growth
and induction conditions were conducted to determine and optimize protein expression. More impor-
tantly, this fusion protein Mistic–CB2–TarCF can localize into the E. coli membrane and exhibit functional
binding activities with known CB2 ligands including CP55,940, WIN55,212-2 and SR144,528. These
results indicate that this novel expression and purification system provides us with a promising strategy
for the preparation of biologically active GPCRs, as well as general application for the preparation of
membrane-bound proteins using the two new fusion partners described.
Ó2012 Elsevier Inc. All rights reserved.
Introduction
The physiological effects of endogenous and synthetic cannabi-
noid ligands are mediated by two cell surface receptors, belonging
to the Rhodopsin family of G protein coupled receptors (GPCRs)
1
[1]. These two receptors, cannabinoid receptor subtype 1 (CB1)
expressing abundantly in the brain and subtype 2 (CB2) expressing
mainly in the immune system, share 68% similarity in their trans-
membrane domains and 44% similarity in their overall receptor se-
quences [2–5]. After stimulation, the CB2 receptor couples to G
a
i
to
negatively regulate cyclic AMP levels by inhibiting adenylase cy-
clase activity [6,7], and to the G
b
c
domain to enhance MAPK and
PI3K activation, ceramide production and downstream gene
expression [8–10]. Clinically, modulation of the CB2 signaling
exhibits great potential for the treatment of inflammatory and
autoimmune diseases, cancer, heart and bone disorders as well
as neurodegenerative disorders [11–15]. In addition, CB2 activa-
tion has also shown to have neuroprotective and analgesic effects
in animals via unclear mechanisms [16,17]. CB1 is highly ex-
pressed in the brain and therapeutic modulations of this receptor
have resulted in adverse psychotropic side effects [18,19]. Selective
modulation of CB2, however, would be able to achieve the desired
therapeutic effect without such psychotropic side effects due to no
or very low expression of CB2 in the central nervous system (CNS).
Therefore, the CB2 receptor is a significant and desirable target for
therapeutic intervention requiring more in-depth information
regarding the receptor structure and function to design highly
selective ligands. However, expression levels of CB2 are very low
in native tissues, and structure determination of CB2 has been im-
peded due to the inability to produce sufficient amounts of the
receptor proteins with high homogeneity and natural ligand bind-
ing activity.
Different hosts have been employed to improve the expression
levels of GPCRs. Baculovirus-infected insect cell lines have been
used to produce GPCRs including the cannabinoid receptor 2 [20],
beta 2-adrenergic receptor [21–23], chemokine receptor [24] and
the A2a adenosine receptor [25,26]; most of which have been
1046-5928/$ - see front matter Ó2012 Elsevier Inc. All rights reserved.
doi:10.1016/j.pep.2012.01.008
⇑
Corresponding author. Fax: +1 412 383 7436.
E-mail address: xix15@pitt.edu (X.-Q. Sean Xie).
1
Abbreviations used: GPCRs, G protein coupled receptors; CB2, cannabinoid
receptor subtype 2; CB1, cannabinoid receptor subtype 1; CNS, central nervous
system; MBP, maltose binding protein; IPTG, isopropyl-b-D-thiogalactoside; TEV,
Tobacco Etch Virus; PIC, protease inhibitor cocktail.
Protein Expression and Purification 83 (2012) 128–134
Contents lists available at SciVerse ScienceDirect
Protein Expression and Purification
journal homepage: www.elsevier.com/locate/yprep

structurally modified to facilitate receptor stability and crystalliza-
tion. Yeast cells also provide eukaryotic environment for post-
translational modification of the exogenous GPCRs [27,28]. How-
ever, compared to mammalian cells, they differ in membrane com-
position and posttranslational modification [29]. While lacking
post-translational modifications, the bacterial system offers several
unbeatable advantages for the expression of exogenous proteins:
fast, homogeneity in protein production, low cost and ability to iso-
topically label the protein of interest for subsequent NMR studies
[30]. Previously, Escherichia coli was used in our lab to express
CB2 receptor fragments by directing the fragment expression to
inclusion bodies using the Trp LE leader sequence [31,32]. The
CB2 receptor fragment produced in E. coli and reconstituted in Brij
58 showed >75% preservation of the alpha helical structure [33].
However, the methodology developed in these studies may not be
applied to the intact receptor without substantial modifications.
To heterologously express eukaryotic membrane proteins, fu-
sion protein technology in E. coli has been successfully applied
for the neurotensin receptor, an integral membrane protein for
which the expression level was enhanced 40-fold when neuroten-
sin was fused to maltose binding protein (MBP) at the at the N ter-
minus and the signal peptide sequence Endotoxin B at the C
terminus [34]. Related methods have also been used for the pro-
duction of the rat neurokinin A receptor [35] and human adenosine
A2a receptors [36]. In addition, expression of the CB2 receptor by
using MBP as an N-terminal fusion partner and Thioredoxin as a
C-terminal fusion partner has also been reported [37–39].
Determining the correct fusion partner(s) to optimize GPCR
expression is not empirical but largely depends on the receptor
in question. Mistic is an unusual B. subtilis membrane protein
[40,41]; TarCF is the C-terminal fragment of bacterial aspartate
chemosensory transducer Tar [42,43]. While Mistic and TarCF have
been used as fusion partners to enhance protein expression and
stabilization, their effects on the expression and stabilization of
GPCRs are obscure and remain unexplored. In the present study,
we have evaluated the roles of several fusion partners including
Mistic, TarCF and TrxA, alone or in combination, to drive the func-
tional expression of the CB2 receptor in E. coli. To facilitate the fu-
sion protein release and purification, Factor Xa/TEV sequences and
multi-His tags were introduced into the expression construct. Cul-
ture conditions were optimized to determine the conditions for
maximum fusion protein yield.
Materials and methods
Expression bacteria strain and reagents
The expression bacteria strain E. coli C43(DE3) was purchased
from Lucigen (Middleton, WI). Strain C43(DE3) contains no intrinsic
plasmids and expresses the T7 polymerase from the lacUV5
promoter upon IPTG induction. In addition, C43(DE3) shows no pro-
teolytic activity towards exogenously overexpressed proteins [44].
3
H-CP55,940 (specific activity: 88.3 Ci/mmol), CP55,940,
WIN55,212-2 and SR144,528 were obtained from RTI International
(Research Triangle Park, NC). Isopropyl-b-
D
-thiogalactoside (IPTG),
benzonase nuclease and lysozyme were purchased from EMD
Chemicals (Gibbstown, NJ). Protease inhibitor cocktail was pur-
chased from Sigma (St. Louis, MO). All restriction and DNA modify-
ing enzymes were purchased from New England Biolabs (Ipswich,
MA).
Construction of recombinant CB2 receptor expression vectors
The constructs used in the present study are shown in Fig. 1. All
expression vectors were based upon the pET-21a vector backbone.
Gene fragment encoding octahistidine tagged Mistic (8His-Mistic)
was derived from the pMIS3.0E vector via polymerase chain reac-
tion using specific primers (For: 5
0
-atatacatatgaaacaccaccacc-3
0
;
Rev: 5
0
-aagcttaccactcaggatcatgtaat-3
0
). The forward and reverse
primers included the restriction sites NdeI and HindIII respectively
for subsequent cloning. The human cannabinoid receptor 2 (CNR2)
gene with the Factor Xa sequence (5
0
-attgagggacgc-3
0
) fused at its
5
0
terminal end (Xa-CB2) was extracted from the pMMHb-TrpLE-
Xa-CB2 vector using HindIII and BamHI sites. The pET-21a–TarCF
construct was used as a template. The 8His-Mistic–Xa-CB2 encod-
ing sequence was cloned upstream of the TarCF gene on the pET-
21a–TarCF template using NdeI and BamHI sites. A Tobacco Etch
Virus (TEV) (sequence 5
0
-gaaaacctatacttccaagga-3
0
) protease recog-
nition site was introduced between TarCF and CB2 encoding se-
quences on the expression plasmid pET-21a for higher efficiency
and specificity of protein cleavage. Similarly, the 8His-Mistic
encoding sequence was subcloned into the pET-21a–CB2–TrxA
template using NdeI and HindIII sites to create the construct (2).
Constructs (3) and (4) were created by removing either the TarCF
sequence (using HindIII and XhoI sites) or the 8His-Mistic se-
quence (using NdeI and AvaI sites) from construct (1), followed
by subsequent Klenow treatments (or a subsequent Klenow treat-
ment) and intramolecular ligation reaction. Double digestion of the
constructs with AvaI and HindIII released the CB2 gene fragment
confirming successful cloning. All construct sequences were veri-
fied by automated DNA sequencing at the University of Pittsburgh
Genomics core facility.
Culture of E. coli C43(DE3) for protein expression
Minicultures were inoculated with single colonies from an
LB-Ampicillin plate containing freshly transformed E. coli
C43(DE3). The bacterial cultures were grown overnight in presence
of Ampicillin (100
l
g/ml) in a shaker (at 250 rpm) at 37 °C. Bacte-
rial maxicultures (1 L) were inoculated with the minicultures and
shaken at 250 rpm, 37 °C until the culture reached an OD
600
of
0.6. Expression of the recombinant CB2 protein was induced with
0.5 mM isopropyl-b-
D
-thiogalactoside (IPTG), followed by continu-
ous shaking for another 4 h at 37 °C. Cells were harvested by cen-
trifugation. After a 50 mM Tris–HCl (pH 8.0) wash, the pellets were
stored at 80 °C for further experiments.
Optimization of culture conditions and IPTG concentration were
performed for maximum expression of Mistic–CB2–TarCF. Briefly,
E. coli C43(DE3) cultures were grown to OD
600
0.6, induced with
0.5 mM or 1 mM IPTG and then maintained at 25 °Cor30°C for
8, 22, 32, 48 and 72 h after IPTG induction.
Preparation of bacterial membrane fractions
The harvested bacterial pellet was washed twice with 0.1 M
Tris–HCI (pH 8.0) buffer and resuspended in the same buffer con-
taining 20% (w/v) sucrose. The OD
600
of the cell suspension was ad-
justed to 10.0. The suspended pellet was incubated at 37 °C for
25 min in the presence of the protease inhibitor cocktail (PIC)
(430
l
g/ml) and lysozyme (0.5
l
l/g) followed by immediate
addition of EDTA to a final concentration 10 mM. After a 0.1 M
Tris–HCI wash containing 20% sucrose, the pellet was then sub-
jected to osmotic lysis by suspension in cold water and sonicated
on ice. This suspension was incubated for 1 h with PIC, benzonase
nuclease and MgCl
2
(10 mM). After a low speed centrifugation
(4500g, 10 min), the supernatant was subjected to a high speed
spin (100,000g, 90 min) at 4 °C. The membrane pellet obtained
was dissolved in Tris–HCI buffer with 20% sucrose and PIC. This
was flash frozen and the aliquots were stored at 80 °C for subse-
quent use.
A. Chowdhury et al. /Protein Expression and Purification 83 (2012) 128–134 129

Detection of CB2 fusion protein expression in E. coli
Transformed E. coli cell pellets or membrane fractions were ana-
lyzed for CB2 expression by Coomassie blue staining and Western
blot. Cell pellets or membrane fractions were lysed in buffer
(10 mM Tris–HCl, 10 mM MgCl
2
, 10% SDS, 430
l
g/ml PIC) and
sonicated briefly. The lysate supernatants were subjected to SDS–
PAGE, followed by Coomassie blue staining. For Western blot
analysis, the lysate supernantant (30
l
g) was heat-denatured, sub-
jected to 12% SDS–PAGE, and transferred to polyvinylidene fluoride
membrane. Following, histidine tagged CB2 receptor were probed
with anti His monoclonal (1:1000, Sigma) and anti CB2 polyclonal
(1:1000, Cayman Chemicals) primary antibodies, the protein bands
were detected using Amersham Enhanced Chemiluminescence-
Western blotting detection reagents (GE Healthcare, Piscataway,
NJ).
Saturation binding assay of the fusion protein
The saturation binding of
3
H-CP55,940 to the membrane pro-
teins was performed as described previously [45]. Briefly, the
membrane fractions (20
l
g) were incubated with increasing con-
centrations of
3
H-CP55,940 (0.01–5 nM) in 96-well plates at
30 °C with slow shaking for 1 h. The incubation buffer was com-
posed of 50 mM Tris–HCl (pH 7.4), 5 mM MgCl
2
, 2.5 mM EGTA
and 0.1% (w/v) fatty acid free BSA. Ligand was diluted in incubation
buffer supplemented with 10% dimethyl sulfoxide and 0.4% methyl
cellulose. Non-specific binding was determined in the presence of
1:1000 unlabeled CP55,940 (5000 nM) in excess. The reaction was
terminated by rapid filtration through Unifilter GF/C filter plates
using a Unifilter Cell Harvester (PerkinElmer). After the plate was
allowed to dry overnight, 30
l
l MicroScint-20 cocktail (PerkinEl-
mer) was added to each well and the radioactivity was counted
by using a PerkinElmer TopCounter. Data from these assays were
analyzed using GraphPad Prism 5.0 Software. The difference be-
tween total and nonspecific binding equals the receptor specific
binding. Non-linear regression analysis revealed the receptor den-
sity (B
max
) and the equilibrium dissociation constant (K
d
) values of
3
H-CP55,940 for the CB2 receptor.
Competitive ligand displacement assay
CB2 receptor ligand displacement assay was performed as
described previously [45]. The known CB2 ligands CP55,940
(unlabeled), WIN55,212-2 and SR144,528 were used in this
displacement assay to test whether the fusion proteins expressed
in E. coli C43(DE3) exhibited receptor–ligand binding properties.
Briefly, non-radioactive (or cold) ligands were diluted (10
2
–
10
3
nM) in binding buffer [50 mM Tris–HCl (pH 7.4), 5 mM MgCl
2
,
2.5 mM EGTA and 0.1% (w/v) fatty acid free BSA], supplemented
with 10% dimethyl sulfoxide and 0.4% methyl cellulose. Each assay
plate well contained a total of 200
l
l of reaction mixture com-
prised of 20
l
g of membrane protein, labeled
3
H-CP55,940 ligand
at a final concentration of 4 nM and the unlabeled ligand at its
varying dilutions as stated above. Plates were incubated at 30 °C
for 1 h with gentle shaking. Reactions were terminated and read
as described in the previous section. All assays were performed
in triplicate (n= 3) and data points represented as mean ± SEM.
Bound radioactivity was analyzed for K
i
values using non-linear
regression analysis by GraphPad Prism 5.0 software.
Results and discussion
Recombinant CB2 receptor produced in E. coli does not have the
ability to translocate to the membrane and is devoid of membrane
environment. This phenomenon has been proposed to be toxic to-
wards the host and lead to misfolded protein aggregation, requir-
ing the isolated protein to undergo refolding [46]. To enhance
membrane protein expression and solubility with correct folding,
as well as membrane localization of the recombinant GPCRs,
researchers have employed several approaches including the iden-
tification of fusion partners linked with GPCRs in E. coli [47,48].
Previous studies have shown that MBP or thioredoxin (Trx) can sta-
bilize and improve the expression and solubility of foreign fusion
proteins in E. coli [49]. Furthermore, Trx fusion proteins can be
folded correctly and express complete biological activity [50]. Mis-
tic, a bacterial membrane-associating protein, has been found to
enhance expression of eukaryotic membrane proteins at the bacte-
rial membrane [40,41,51]. The chemosensory aspartate receptor,
Tar, is a resident membrane protein of the bacterial host which is
expected to facilitate membrane protein expression [52]. Combin-
ing different fusion partners at both ends of a target gene has
emerged as a promising strategy to facilitate expression and im-
prove the solubility of recombinant proteins [39]. However, appli-
cation of the fusion tags Mistic and TarCF for the expression of
GPCRs in E. coli has not been investigated previously. For the first
time, we report in this study the use of different fusion partner
combinations (Mistic, TarCF and TrxA) for the functional
Fig. 1. Schematic diagram of human CB2 fusion protein constructs. All expression plasmid vectors (1–4) were constructed on the pET-21a vector backbone under the control
of the T7 promoter. Mistic, the N-terminal fusion tag, was separated from CB2 by the Factor Xa sequence while TarCF or TrxA, C-terminal fusion tag, were separated from the
CB2 receptor by the TEV sequence. The boxes shown are not drawn to scale. TarCF, C-terminal fragment of bacterial aspartate chemosensory transducer Tar; TrxA,
thioredoxin; TEV, tobacco etch virus sequence; His, Histidine residues.
130 A. Chowdhury et al. /Protein Expression and Purification 83 (2012) 128–134

expression of the recombinant CB2 receptor in E. coli C43(DE3). We
show here, that the fusion protein Mistic–CB2–TarCF is overex-
pressed by E. coli and localized to the bacterial membrane with li-
gand binding properties comparable to those on mammalian cells.
Expression of cannabinoid receptor 2 fusion protein in E. coli
E. coli C43(DE3) cells were transformed respectively with the
fusion constructs shown in Fig. 1. Expression of the recombinant
fusion proteins were detected by Western blot or Commassie Blue
staining. Since all constructs contain a multi-histidine tag, we used
either anti-His or anti-CB2 antibody to detect expression of the CB2
fusion protein. As shown in Fig. 2A, Mistic and TarCF alone failed to
boost the CB2 gene expression. Only when both partners were
linked to CB2 in the proper order did the fusion protein expression
increase dramatically. In addition, fusion protein expression was
also observed with the Mistic–CB2–TrxA construct at a comparable
expression level with that of the Mistic–CB2–TarCF (data not
shown). However, the construct Mistic–CB2–TarCF was used for
further optimization due to its novel combination of fusion part-
ners and prominent expression levels of recombinant fusion
protein.
Next, we investigated whether the expressed CB2 fusion protein
possessed membrane affinity or localized to the E. coli membrane.
Coomassie staining of the membrane enriched fractions revealed a
prominent band at MW 86 kDa for the fusion protein Mistic–
CB2–TarCF while no bands were detected for the fusion proteins
that carried either the Mistic or TarCF tag individually (Fig. 2B).
Importantly, the membrane enriched fractions exhibited the same
expression pattern as the whole E. coli cell lysates, indicating that
all or the majority of CB2 fused protein driven by the two partners
could localize or integrate into the E. coli membrane. Our data sug-
gests that combination of the two tags (Mistic and TarCF) may con-
tribute synergistic effects on the CB2 protein expression compared
to either tag used alone. Our data also show that membrane frac-
tions contain concentrated CB2 fusion protein compared to the
whole cell lysate, indicating that most of the fusion protein is local-
ized within the bacterial membrane. This is in accordance with
previous studies where the effects of Mistic and other bacterial
membrane resident protein to stabilize GPCR expression has been
demonstrated [41,53,54].
In absence of induction with IPTG, there was no expression of
the fusion protein. However, after induction with IPTG, the fusion
protein level increased significantly in a time-dependent manner
(Fig. 3), suggesting that the expression cassette is under the tight
control of the lac operon and T7/lac promoter. Also the pET21a car-
ries the lacI gene together these elements explain why the expres-
sion is tightly controlled.
Control of recombinant protein expression under the tight reg-
ulation is necessary to avoid toxicity of protein expression to the
host and ensure sufficient biomass of viable E. coli that would be
available for membrane protein expression after induction.
To determine the time point of maximum receptor production,
IPTG-induced cells were harvested at different time intervals from
1 to 8 h. Whole cell lysates were analyzed by Western blot using
mouse anti-His (1:1000 dilution) and rabbit anti-CB2 antibodies
(1:500 dilution). As shown in Fig. 3A, CB2 fusion protein expression
levels steadily increased and reached maxima at 3–4 h, followed by
significantly reduced expression. Thus, from this experiment we
can conclude that the expression level of the fusion protein peaked
during culture at 37 °C for 3–4 h after IPTG induction. Since IPTG
induction at lower temperature was previously reported to im-
prove the exogenous protein production and correct folding [54],
we optimized the culture conditions by combining different IPTG
concentrations (0.5 mM and 1 mM), culture temperature and time.
We found that the expression levels of the fusion protein Mistic–
CB2–TarCF are weakly detected during culture period (2–48 h) at
22 °C (data not shown). The fusion protein expression at 30 °C
was not distinct from the regular 37 °C culture condition
(Fig. 3B). However, once the transformed cultured underwent IPTG
induction (1 mM) at 25 °C for 8 h, the fusion protein levels were
significantly increased 2-fold of that of regular conditions
(Fig. 3B). Overall, 0.5 mM IPTG used for inducing protein expres-
sion resulted in lower amounts of fusion protein than 1 mM –
especially for the conditions of culture temperatures at 25 °Cor
30 °C (data not shown).
Receptor saturation binding assay
pET-21a–Mistic–CB2–TarCF transformed E. coli membranes
were subjected to a saturation binding assay to determine receptor
saturation with increasing concentrations of
3
H-CP55,940.
pET-21a–TarCF transformed E. coli membranes were used as the
negative control. For the membrane proteins derived from Mis-
tic–CB2–TarCF transformed E. coli, the maximal receptor density
(B
max
) and dissociation constant (K
d
)of
3
H-CP55,940 for specific
binding sites were 928.8 ± 117.6 fmol/mg protein and 3.04 ± 0.69
nM, respectively (Fig. 4A). Membrane fractions clearly showed
CB2 receptor binding characterization by the abundance of binding
sites recognized by agonist
3
H-CP55,940. For the negative control,
however, no difference was observed between specific and nonspe-
cific binding (Fig. 4B), indicating that the overwhelming majority
of the total binding was contributed by the nonspecific binding.
This confirms the absence of CB2 receptor on pET-21a–TarCF trans-
formed E. coli membranes.
Competitive ligand displacement assays
The conformational state of a receptor protein determines the
functional state of a receptor. High affinity binding between a li-
gand and its receptors is often physiologically important when a
portion of the binding energy can be used to cause a conforma-
tional change in the receptor, resulting in altered downstream
signaling pathways. In the present study, to confirm whether the
expressed fusion proteins from the E. coli exhibit functional bind-
ing activity, we used the well-known CB2 ligands to probe the
interactions of these ligands with their cognate binding sites on
1 2 3
AB
250
150
100
75
50
M 1 2 3
*
**
Fig. 2. Enhanced expression of the CB2 receptor fusion protein. (A) Representative
immunoblot of His-tagged CB2 fusion protein detected in E. coli C43(DE3)
membrane fractions using anti-His. Membrane fractions loaded from left are Lane
1: Mistic–CB2–TarCF; Lane 2: Mistic–CB2; Lane 3: CB2–TarCF. (B) Coomassie
Brilliant Blue staining on the SDS–PAGE of extracted membrane fractions. The
expected MW for the respective fusion proteins are as follows – Lane 1: Mistic–
CB2–TarCF (86 kDa); Lane 2: Mistic–CB2 (71 kDa); Lane 3: CB2–TarCF (55 kDa).
Asterisks show the corresponding fusion protein expression. M: protein marker.
Care was taken to normalize the amount of E. coli C43(DE3) membrane fraction
sample loaded on the gel.
A. Chowdhury et al. /Protein Expression and Purification 83 (2012) 128–134 131

the CB2 enriched membrane fractions (competitive binding assay),
by quantifying the equilibrium dissociation constant (K
i
). By using
10
l
g of membrane fractions of Mistic–CB2–TarCF fusion protein
in the binding assay, the K
i
values for these ligands were well con-
sistent with previous reports using the CB2 from mammalian cells:
CP55,940 (K
i
= 1.43 nM), SR144,528 (K
i
= 2.02 nM) and WIN55,
212-2 (K
i
= 0.13 nM). These results indicate that the ligand binding
domain of the CB2 receptor in the fusion protein is not perturbed
by the physical presence of its neighboring fusion partners Mistic
and TarCF (Fig. 5).
Conclusion
In summary, our data has demonstrated that the Mistic–CB2–
TarCF construct can successfully express the CB2 receptor protein
in E. coli C43(DE3). The obtained fusion proteins can localize at the
bacterial membrane. Importantly, the Mistic–CB2–TarCF fusion pro-
teins show effective binding activity with the known CB2 ligands.
This suggests that the conformational state of the native CB2 recep-
tor, used for specific ligand binding, is retained in the presence of
fusion partners. Also, we found that the fusion partners – Mistic
Control +IPTG (0.5 mM)
(-IPTG) 1 2 3 4 6 8 (h)
Anti-His
Anti-CB2
Anti-His
8 22 32 48 72 8 22 32 48 72 4 (h)
25оC30
оC37
оC
+IPTG (1mM)
A
B
Fig. 3. Optimization of conditions for fusion protein expression in E. coli. (A) Cells transformed with Mistic–CB2–TarCF were grown for the indicated hours after induction
with IPTG (0.5 mM). Expression levels of fusion protein Mistic–CB2–TarCF tagged with poly-histidine were detected by Western blot with anti-CB2 or anti-His antibody.
Control group (0 h) represented no IPTG induction. (B) Optimization of Mistic–CB2–TarCF fusion protein production in E coli. Different combinations of the parameters (IPTG,
culture temperature and time) were tested and one representative setting was shown.
Bmax=928.8±117.6
Kd=3.04±0.69
2 4 6
-200
0
200
400
600
nM Radioligand
Bound (cpm)
2 4 6
-200
0
200
400
600
800
1000
nM Radioligand
Bound (cpm)
A
B
Fig. 4. Saturation binding assay of the membrane fractions. Total (s) and non-specific (j) binding was measured and the deduced specific binding saturation isotherm (N)
was obtained as the difference between total and nonspecific binding. (A) Mistic–CB2–TarCF; (B) pET 21-TarCF (negative control). Assay was performed in triplicate (n= 3).
Data presented as mean ± SEM.
-2 0 2
0
20
40
60
80
100
log[CP55940] nM
% of Bound [
3
H]CP55940(cpm)
-2 0 2
0
20
40
60
80
100
log[SR 144528] nM
% of Bound [
3
H]CP55940(cpm)
-2 0 2
0
20
40
60
80
100
log[WIN 55212-2] nM
% of Bound [
3
H]CP55940(cpm)
ABC
Fig. 5. Competitive displacement of the
3
H-CP55,940 was obtained by using an increased amount of cold ligands. Binding profile of (A) CP55,940 (unlabelled), K
i
= 1.43 nM;
(B) SR144,528, K
i
= 2.02 nM; (C) WIN55,212–2, K
i
= 0.13 nM. Assay was performed in triplicate (n= 3). Data presented as mean ± SEM.
132 A. Chowdhury et al. / Protein Expression and Purification 83 (2012) 128–134

and TarCF – in combination, are more effective for enhancing protein
expression in E. coli, than their use alone. Overall findings from this
present study suggest that the targeting of fusion partners to the
bacterial membrane is critical to the conformational stability of
the expressed CB2 protein. The possible role of the fusion partners
for the overexpression and stabilization the CB2 protein is illustrated
schematically (Fig. 6) for easy comprehension. In this putative mod-
el, the CB2 receptor structure was adapted from the 3D CB2 model
reported previously by Xie et al. [5], while the structure of Mistic
and Tsr (structurally related to Tar) were determined by NMR
(PDB:1YGM) [40] and cryo-electron microscopy [55] studies, respec-
tively. However, confirming the putative model will be subject to
further biophysical studies. Currently, we are using the entire fusion
protein and cleaved receptor in parallel to carry out 2-dimensional
crystal growth and analysis by cryo-electron microscopy. The trials
for 2D crystal generation will be favorably facilitated by the in-
creased molecular weight of the fusion protein complex [56].
Our preliminary studies of detergent-screening in small-scale
(data now shown) indicated the feasibility of rapid affinity purifi-
cation for the fusion protein in the presence of detergents. These
results strongly encourage us to optimize the extraction and puri-
fication conditions for large-scale production of human CB2 recep-
tor to study its functional aspects. If necessary, we may use the
refolding mechanism already demonstrated for the CB1 receptor
[46]. Overall, our studies show new fusion partners for the func-
tional expression of the cannabinoid receptor 2 in the bacterial
membrane. We anticipate this approach will produce enough pro-
tein to conduct further biophysical studies.
Acknowledgments
The authors thank Dr. Peijun Zhang (Department of Structural
Biology, University of Pittsburgh) for the pET-21a-TarCF vector
and planned CryoEM studies; Dr. Senyon Choe (Department of
Structural Biology, Salk Institute) for the pMIS3.0E vector. The
work was supported by NIH RO1 DA025612.
References
[1] T.K. Attwood, J.B. Findlay, Fingerprinting G-protein-coupled receptors, Protein
Eng. 7 (1994) 195–203.
[2] S. Galiegue, S. Mary, J. Marchand, D. Dussossoy, D. Carriere, P. Carayon, M.
Bouaboula, D. Shire, G. Le Fur, P. Casellas, Expression of central and peripheral
cannabinoid receptors in human immune tissues and leukocyte
subpopulations, Eur. J. Biochem. 232 (1995) 54–61.
[3] A.C. Howlett, F. Barth, T.I. Bonner, G. Cabral, P. Casellas, W.A. Devane, C.C.
Felder, M. Herkenham, K. Mackie, B.R. Martin, R. Mechoulam, R.G. Pertwee,
International Union of Pharmacology. XXVII. Classification of cannabinoid
receptors, Pharmacol. Rev. 54 (2002) 161–202.
[4] S. Munro, K.L. Thomas, M. Abu-Shaar, Molecular characterization of a
peripheral receptor for cannabinoids, Nature 365 (1993) 61–65.
[5] X.Q. Xie, J.Z. Chen, E.M. Billings, 3D structural model of the G-protein-coupled
cannabinoid CB2 receptor, Proteins 53 (2003) 307–319.
[6] M. Bayewitch, T. Avidor-Reiss, R. Levy, J. Barg, R. Mechoulam, Z. Vogel, The
peripheral cannabinoid receptor: adenylate cyclase inhibition and G protein
coupling, FEBS Lett. 375 (1995) 143–147.
[7] W. Gonsiorek, C. Lunn, X. Fan, S. Narula, D. Lundell, R.W. Hipkin,
Endocannabinoid 2-arachidonyl glycerol is a full agonist through human
type 2 cannabinoid receptor: antagonism by anandamide, Mol. Pharmacol. 57
(2000) 1045–1050.
[8] M. Bouaboula, C. Poinot-Chazel, J. Marchand, X. Canat, B. Bourrie, M. Rinaldi-
Carmona, B. Calandra, G. Le Fur, P. Casellas, Signaling pathway associated with
stimulation of CB2 peripheral cannabinoid receptor. Involvement of both
mitogen-activated protein kinase and induction of Krox-24 expression, Eur. J.
Biochem. 237 (1996) 704–711.
[9] M. Bouaboula, N. Desnoyer, P. Carayon, T. Combes, P. Casellas, Gi protein
modulation induced by a selective inverse agonist for the peripheral
cannabinoid receptor CB2: implication for intracellular signalization cross-
regulation, Mol. Pharmacol. 55 (1999) 473–480.
[10] M. Bouaboula, D. Dussossoy, P. Casellas, Regulation of peripheral cannabinoid
receptor CB2 phosphorylation by the inverse agonist SR 144528. Implications
for receptor biological responses, J. Biol. Chem. 274 (1999) 20397–20405.
[11] A. Alexander, P.F. Smith, R.J. Rosengren, Cannabinoids in the treatment of
cancer, Cancer Lett. 285 (2009) 6–12.
[12] A.N. Lozano-Ondoua, C. Wright, A. Vardanyan, T. King, T.M. Largent-Milnes, M.
Nelson, J.M. Jimenez-Andrade, P.W. Mantyh, T.W. Vanderah, A cannabinoid 2
receptor agonist attenuates bone cancer-induced pain and bone loss, Life Sci.
86 (2010) 646–653.
[13] A.M. Martin-Moreno, D. Reigada, B.G. Ramirez, R. Mechoulam, N. Innamorato,
A. Cuadrado, M.L. de Ceballos, Cannabidiol and other cannabinoids reduce
microglial activation in vitro and in vivo: relevance to Alzheimers’ disease.
Mol. Pharmacol., 2011.
[14] P. Pacher, R. Mechoulam, Is lipid signaling through cannabinoid 2 receptors
part of a protective system?, Prog Lipid Res. 50 (2011) 193–211.
[15] J.P. Zajicek, V.I. Apostu, Role of cannabinoids in multiple sclerosis, CNS Drugs
25 (2011) 187–201.
[16] G.A. Cabral, E.S. Raborn, L. Griffin, J. Dennis, F. Marciano-Cabral, CB2 receptors
in the brain: role in central immune function, Br. J. Pharmacol. 153 (2008)
240–251.
[17] P. Anand, G. Whiteside, C.J. Fowler, A.G. Hohmann, Targeting CB2 receptors
and the endocannabinoid system for the treatment of pain, Brain Res. Rev. 60
(2009) 255–266.
[18] D.L. Kelly, D.A. Gorelick, R.R. Conley, D.L. Boggs, J. Linthicum, F. Liu, S. Feldman,
M.P. Ball, H.J. Wehring, R.P. McMahon, M.A. Huestis, S.J. Heishman, K.R.
Warren, R.W. Buchanan, Effects of the cannabinoid-1 receptor antagonist
rimonabant on psychiatric symptoms in overweight people with
schizophrenia: a randomized, double-blind, pilot study, J. Clin.
Psychopharmacol. 31 (2011) 86–91.
[19] K. Cahill, M. Ussher, Cannabinoid type 1 receptor antagonists (rimonabant) for
smoking cessation. Cochrane Database Syst. Rev. (2007) CD005353.
[20] K.W. Nowell, D.A. Pettit, W.A. Cabral, H.W. Zimmerman Jr., M.E. Abood, G.A.
Cabral, High-level expression of the human CB2 cannabinoid receptor using a
baculovirus system, Biochem. Pharmacol. 55 (1998) 1893–1905.
[21] S.G. Rasmussen, H.J. Choi, J.J. Fung, E. Pardon, P. Casarosa, P.S. Chae, B.T. Devree,
D.M. Rosenbaum, F.S. Thian, T.S. Kobilka, A. Schnapp, I. Konetzki, R.K.
Sunahara, S.H. Gellman, A. Pautsch, J. Steyaert, W.I. Weis, B.K. Kobilka,
Structure of a nanobody-stabilized active state of the beta(2) adrenoceptor,
Nature 469 (2011) 175–180.
[22] S.G. Rasmussen, H.J. Choi, D.M. Rosenbaum, T.S. Kobilka, F.S. Thian, P.C.
Edwards, M. Burghammer, V.R. Ratnala, R. Sanishvili, R.F. Fischetti, G.F.
Schertler, W.I. Weis, B.K. Kobilka, Crystal structure of the human beta2
adrenergic G-protein-coupled receptor, Nature 450 (2007) 383–387.
[23] V. Sarramegna, F. Talmont, P. Demange, A. Milon, Heterologous expression of
G-protein-coupled receptors: comparison of expression systems fron the
standpoint of large-scale production and purification, Cell. Mol. Life Sci. 60
(2003) 1529–1546.
[24] P.D. Kwong, R. Wyatt, J. Robinson, R.W. Sweet, J. Sodroski, W.A. Hendrickson,
Structure of an HIV gp120 envelope glycoprotein in complex with the CD4
receptor and a neutralizing human antibody, Nature 393 (1998) 648–659.
[25] V. Cherezov, D.M. Rosenbaum, M.A. Hanson, S.G. Rasmussen, F.S. Thian, T.S.
Kobilka, H.J. Choi, P. Kuhn, W.I. Weis, B.K. Kobilka, R.C. Stevens, High-
resolution crystal structure of an engineered human beta2-adrenergic G
protein-coupled receptor, Science 318 (2007) 1258–1265.
[26] V.P. Jaakola, M.T. Griffith, M.A. Hanson, V. Cherezov, E.Y. Chien, J.R. Lane, A.P.
Ijzerman, R.C. Stevens, The 2.6 angstrom crystal structure of a human A2A
adenosine receptor bound to an antagonist. Science 322 (2008)
1211–1217.
Mistic
CB2
TarCF
Linker 1
Linker 2
Extracellular
Intracellular
Factor Xa
TEV
Fig. 6. Putative schematic diagram illustrating the plausible structure and mem-
brane bound state of the fusion protein construct. Mistic (golden) [PDB ID:1YGM] is
joined to the N-terminal of the CB2 receptor (structure from homology model [5])
(rainbow) via linker 1 containing Factor Xa sequence. The CB2 receptor is linked to
TarCF (red) via linker 2 containing the TEV sequence. Mistic and TarCF contain 8 and
6 histidine residues in their N- and C-terminal domains respectively. (For
interpretation of the references to color in this figure legend, the reader is referred
to the web version of this article.)
A. Chowdhury et al. / Protein Expression and Purification 83 (2012) 128–134 133

[27] F. Naider, R. Estephan, J. Englander, V.V. Suresh Babu, E. Arevalo, K. Samples,
J.M. Becker, Sexual conjugation in yeast: a paradigm to study G-protein-
coupled receptor domain structure, Biopolymers 76 (2004) 119–128.
[28] T.K. Kim, R. Zhang, W. Feng, J. Cai, W. Pierce, Z.H. Song, Expression and
characterization of human CB1 cannabinoid receptor in methylotrophic yeast
Pichia pastoris, Protein Expr. Purif. 40 (2005) 60–70.
[29] V. Sarramegna, F. Talmont, P. Demange, A. Milon, Heterologous expression of
G-protein-coupled receptors: comparison of expression systems from the
standpoint of large-scale production and purification, Cell. Mol. Life Sci. 60
(2003) 1529–1546.
[30] R.C. Hockney, Recent developments in heterologous protein production in
Escherichia coli, Trends Biotechnol. 12 (1994) 456–463.
[31] X.Q. Xie, J. Zhao, H. Zheng, Expression, purification, and isotope labeling of
cannabinoid CB2 receptor fragment, CB2(180–233), Protein Expr. Purif. 38
(2004) 61–68.
[32] H. Zheng, J. Zhao, S. Wang, C.M. Lin, T. Chen, D.H. Jones, C. Ma, S. Opella, X.Q.
Xie, Biosynthesis and purification of a hydrophobic peptide from
transmembrane domains of G-protein-coupled CB2 receptor, J. Pept. Res. 65
(2005) 450–458.
[33] Y. Zhang, X.Q. Xie, Biosynthesis, purification, and characterization of a
cannabinoid receptor 2 fragment (CB2(271–326)), Protein Expr. Purif. 59
(2008) 249–257.
[34] R. Grisshammer, R. Duckworth, R. Henderson, Expression of a rat neurotensin
receptor in Escherichia coli, Biochem. J. 295 (Pt 2) (1993) 571–576.
[35] R. Grisshammer, J. Little, D. Aharony, Expression of rat NK-2 (neurokinin A)
receptor in E. coli, Receptors Channels 2 (1994) 295–302.
[36] H.M. Weiss, R. Grisshammer, Purification and characterization of the human
adenosine A(2a) receptor functionally expressed in Escherichia coli, Eur. J.
Biochem. 269 (2002) 82–92.
[37] C. Berger, J.T. Ho, T. Kimura, S. Hess, K. Gawrisch, A. Yeliseev, Preparation of
stable isotope-labeled peripheral cannabinoid receptor CB2 by bacterial
fermentation, Protein Expr. Purif. 70 (2010) 236–247.
[38] D. Krepkiy, K. Gawrisch, A. Yeliseev, Expression and purification of CB2 for
NMR studies in micellar solution, Protein Pept. Lett. 14 (2007) 1031–1037.
[39] A.A. Yeliseev, K.K. Wong, O. Soubias, K. Gawrisch, Expression of human
peripheral cannabinoid receptor for structural studies, Protein Sci. 14 (2005)
2638–2653.
[40] T.P. Roosild, J. Greenwald, M. Vega, S. Castronovo, R. Riek, S. Choe, NMR
structure of Mistic, a membrane-integrating protein for membrane protein
expression, Science 307 (2005) 1317–1321.
[41] G. Kefala, W. Kwiatkowski, L. Esquivies, I. Maslennikov, S. Choe, Application of
Mistic to improving the expression and membrane integration of histidine
kinase receptors from Escherichia coli, J. Struct. Funct. Genom. 8 (2007)
167–172.
[42] A. Krikos, M.P. Conley, A. Boyd, H.C. Berg, M.I. Simon, Chimeric chemosensory
transducers of Escherichia coli, Proc. Natl. Acad. Sci. USA 82 (1985) 1326–1330.
[43] F.M. Antommattei, J.B. Munzner, R.M. Weis, Ligand-specific activation of
Escherichia coli chemoreceptor transmethylation, J. Bacteriol. 186 (2004)
7556–7563.
[44] B. Miroux, J.E. Walker, Over-production of proteins in Escherichia coli: mutant
hosts that allow synthesis of some membrane proteins and globular proteins
at high levels, J. Mol. Biol. 260 (1996) 289–298.
[45] W.R. Leifert, O. Bucco, M.Y. Abeywardena, G.S. Patten, Radioligand binding
assays: application of [(125)I]angiotensin II receptor binding, Methods Mol.
Biol. 552 (2009) 131–141.
[46] K. Michalke, C. Huyghe, J. Lichiere, M.E. Graviere, M. Siponen, G. Sciara, I.
Lepaul, R. Wagner, C. Magg, R. Rudolph, C. Cambillau, A. Desmyter, Mammalian
G protein-coupled receptor expression in Escherichia coli: II. Refolding and
biophysical characterization of mouse cannabinoid receptor 1 and human
parathyroid hormone receptor 1, Anal. Biochem. 401 (2010) 74–80.
[47] X. Zuo, S. Li, J. Hall, M.R. Mattern, H. Tran, J. Shoo, R. Tan, S.R. Weiss, T.R. Butt,
Enhanced expression and purification of membrane proteins by SUMO fusion
in Escherichia coli, J. Struct. Funct. Genomics 6 (2005) 103–111.
[48] A. Korepanova, F.P. Gao, Y. Hua, H. Qin, R.K. Nakamoto, T.A. Cross, Cloning and
expression of multiple integral membrane proteins from Mycobacterium
tuberculosis in Escherichia coli, Protein Sci. 14 (2005) 148–158.
[49] R.B. Kapust, D.S. Waugh, Escherichia coli maltose-binding protein is
uncommonly effective at promoting the solubility of polypeptides to which
it is fused, Protein Sci. 8 (1999) 1668–1674.
[50] E.R. LaVallie, E.A. DiBlasio, S. Kovacic, K.L. Grant, P.F. Schendel, J.M. McCoy, A
thioredoxin gene fusion expression system that circumvents inclusion body
formation in the E. coli cytoplasm. Biotechnology (NY) 11 (1993) 187–193.
[51] K.Y. Blain, W. Kwiatkowski, S. Choe, The functionally active Mistic-fused
histidine kinase receptor, EnvZ. Biochem. 49 (2010) 9089–9095.
[52] Y. Meir, V. Jakovljevic, O. Oleksiuk, V. Sourjik, N.S. Wingreen, Precision and
kinetics of adaptation in bacterial chemotaxis, Biophys. J. 99 (2010) 2766–
2774.
[53] F. Baneyx, Recombinant protein expression in Escherichia coli, Curr. Opin.
Biotechnol. 10 (1999) 411–421.
[54] M. Freigassner, H. Pichler, A. Glieder, Tuning microbial hosts for membrane
protein production, Microb. Cell Fact. 8 (2009) 69.
[55] C.M. Khursigara, X. Wu, P. Zhang, J. Lefman, S. Subramaniam, Role of HAMP
domains in chemotaxis signaling by bacterial chemoreceptors, Proc. Natl.
Acad. Sci. USA 105 (2008) 16555–16560.
[56] D.R. Smyth, M.K. Mrozkiewicz, W.J. McGrath, P. Listwan, B. Kobe, Crystal
structures of fusion proteins with large-affinity tags, Protein Sci. 12 (2003)
1313–1322.
134 A. Chowdhury et al. / Protein Expression and Purification 83 (2012) 128–134