Crystal and solution structures of an odorant-binding
protein from the southern house mosquito complexed
with an oviposition pheromone
Yang Maoa,1, Xianzhong Xub, Wei Xuc, Yuko Ishidac,2, Walter S. Lealc,3, James B. Amesb, and Jon Clardya
aDepartment of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, Boston, MA 02115; and Departments ofbChemistry and
cEntomology, University of California, Davis, CA 95616
Edited by Jerrold Meinwald, Cornell University, Ithaca, NY, and approved September 27, 2010 (received for review August 17, 2010)
Culexmosquitoesintroducethe pathogensresponsible for filariasis,
West Nile virus, St. Louis encephalitis, and other diseases into
humans. Currently, traps baited with oviposition semiochemicals
play an important role in detection efforts and could provide an
environmentally friendly approach to controlling their populations.
The odorant binding proteins (OBPs) in the female’s antenna play
a crucial, if yet imperfectly understood, role in sensing oviposition
cues. Here, we report the X-ray crystallography and NMR 3D struc-
tures of OBP1 for Culex quinquefasciatus (CquiOBP1) bound to an
oviposition pheromone (5R,6S)-6-acetoxy-5-hexadecanolide (MOP).
In both studies, CquiOBP1 had the same overall six-helix structure
seen in other insect OBPs, buta detailed analysis revealed an impor-
tant previously undescribed feature. There are two models for OBP-
mediated signal transduction: (i) direct release of the pheromone
from an internal binding pocket in a pH-dependent fashion and (ii)
detection of a pheromone-induced conformational change in the
OBP·pheromone complex. Although CquiOBP1 binds MOP in a pH-
dependent fashion, it lacks the C terminus required for the pH-
dependent release model. This study shows that CquiOBP binds
MOP in an unprecedented fashion using both a small central cavity
for the lactone head group and a long hydrophobic channel for its tail.
throughout the world. In the United States, they spread West Nile
virus while feeding on birds and humans (1). In Africa, Culex
These insecticide-treated nets provide only limited protection
against blood-feeding by C. quinquefasciatus (2), thus misleading
end users regarding the effectiveness of the nets. Environmentally
friendly strategies for controlling Culex mosquitoes would have
broader impacts in medical entomology, including mitigation of
malaria as well as some of the so-called “neglected” diseases. Ef-
fective management of Culex mosquito populations may be ach-
ieved with oviposition attractants and other kairomones (3, 4),
because larval development is a particularly vulnerable point in
theirlife cycle.These semiochemicalsaredetected by theantennae
of female adults with sensilla housing odorant receptors (ORs) (5,
the development of green chemistry-based strategies for insect
vector management (8). Previously, we have isolated an OBP from
the antennae of C. quinquefasciatus, CquiOBP1, which is highly
oviposition pheromone (8), (5R,6S)-6-acetoxy-5-hexadecanolide
(11) (hereafter referred to as MOP). CquiOBP1 bound MOP with
high affinity at high pH but showed no affinity at low pH (8). OBPs
essential role for odorant reception in C. quinquefasciatus has
been demonstrated by RNA interference experiments in which
reduction of CquiOBP1 expression led to lower sensitivity for the
detection of oviposition attractants, including MOP (12). Using
CquiOBP1 as a molecular target in a reverse chemical ecology ap-
proach (10), we identified C. quinquefasciatus oviposition attrac-
tants (8), which are now commercially available for monitoring
Previous structural and functional studies of insect OBPs were
focused on two questions: (i) whether OBPs can specifically rec-
ognize the corresponding ligands and (ii) how OBPs transfer the
carried ligand and/or chemical stimulus to the ORs. Because the
insect olfactory system is both extremely sensitive and selective,
there have been long debates on whether OBPs are involved in
discriminating among potential molecular signals. Plenty of struc-
an OBP and its bound semiochemical (13–16). For example,
a crystal structure of a PBP from the silk moth Bombyx mori
the bound bombykol, a C16 long-chain alkene alcohol (15). Simi-
larly, the crystal structure of the OBP LUSH from Drosophila
melanogaster demonstrates a high-affinity alcohol-binding site (13,
16). However, other binding and structural studies also show the
remarkable plasticity in the ligand-binding site of OBP (17).
There are also two distinctively different models that have been
proposed as the mechanisms that transfer the molecular signal to
ORs. A series of structural studies of the BmorPBP reveals that
BmorPBP exists in two major conformations in a pH-dependent
manner. At a pH lower than 5.5, the C terminus of BmorPBP
adopts a newly formed α-helix conformation occupying the
bombykol-binding site, whereas at a pH higher than 5.5, the same
C-terminus region forms an elongated stretch outside the binding
site, making it available for the pheromone. These results led the
authors to propose a pH-induced ligand-releasing mechanism,
through which BmorPBP ejects the bound pheromone to its re-
ceptor when encountering the low pH generated by the negatively
mechanism is observed for a PBP from the giant silk moth
Antheraea polyphemus (14, 23). Contrary to the direct ligand re-
leasing, the structural studies of the OBP LUSH suggest a com-
pletely different way of signal transduction. It has been shown that
Author contributions: W.S.L., J.B.A., and J.C. designed research; Y.M., X.X., W.X., and Y.I.
performed research; Y.I. and W.S.L. contributed new reagents/analytic tools; Y.M., X.X.,
W.X., W.S.L., J.B.A., and J.C. analyzed data; and Y.M., W.S.L., J.B.A., and J.C. wrote the
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
Freely available online through the PNAS open access option.
Data deposition: The NMR chemical shifts have been deposited in the BioMagResBank,
www.bmrb.wisc.edu (accession no. 16175). The coordinates of CquiOBP1·MOP have been
deposited in the Protein Data Bank, www.pdb.org [PDB ID code 3OGN (X-ray crystallog-
raphy structure) and PDB ID code 212C (NMR structure)].
1Present address: Department of Biochemistry, Boston University School of Medicine,
Boston, MA 02118.
2Present address: Graduate School of Science, Kobe University, 1-1 Rokkodai-cho, Nada-
ku, Kobe 657-8501, Japan.
3To whom correspondence should be addressed. E-mail: firstname.lastname@example.org.
| November 2, 2010
| vol. 107
| no. 44www.pnas.org/cgi/doi/10.1073/pnas.1012274107
LUSH undergoes a conformational change on binding with its
specific pheromone, switching a surface loop from an inactive
conformation to an active conformation that is capable of trig-
ligand released by the OBP, the ORs are proposed to perceive the
information carried by OBPs indirectly through the conforma-
tional changes elicited by specific pheromones (24).
MOP affinity for CquiOBP1 is pH-dependent, as was observed
for moth PBPs. These findings suggest a different mechanism,
because CquiOBP1 lacks a C-terminus region that is long enough
in the moth PBPs to fold into an extra α-helix at low pH. Addi-
from the malaria mosquito Anopheles gambiae and AaegOBP1
from the yellow fever mosquito Aedes aegypti, were recently solved
by X-ray crystallography (25, 26). Those two OBPs, closely related
in sequence to each other and to CquiOBP1, reveal strikingly
ecology of these three species. To explore CquiOBP1 further as
a molecular target for the development of mosquito oviposition
attractants, we determined by X-ray crystallography and NMR the
3D structure of CquiOBP1 complexed with MOP and compared it
with AgamOBP1 and AaegOBP1 structures. We describe here
structural features of mosquito OBPs, including a unique binding
pocket with a long hydrophobic channel.
Results and Discussion
Crystal Structure of the CquiOBP1·MOP Complex. The crystal struc-
to a resolution of 1.3 Å. Initial phases were readily determined by
molecular replacement using the structure of AgamOBP1. The
structure was refined to a final Rfactorand Rfreeof 13.2% and
17.4%, respectively (Table 1).
The overall structure of the CquiOBP1·MOP complex is similar
to that of other previously solved OBP structures, which consist of
six helices (labeled α1 to α6) surrounding a central hydrophobic
cavity. Three conserved disulfide bonds are established between
cysteine pairs 26/57,53/104, and 95/113 (Fig.1A).Inparticular,the
protein structure is almost identical to that of the other two mos-
quito OBPs, AgamOBP1 and AaegOBP1 (Fig. 1B). LSQMAN
alignment using the α-carbons shows a rmsd from CquiOBP1
of 0.344 Å for AgamOBP1 and 0.454 Å for AaegOBP1. Both
AgamOBP1 and AaegOBP1 share a high degree of sequence
identity withCquiOBP1,90% and 87%,respectively.Incontrastto
the C-terminus extension of CquiOBP1 is much shorter and folds
inside the central cavity, making up part of the central cavity wall.
CquiOBP1 has been found to exist in monomer-dimer equi-
librium in solution (8). The protein also forms a noncrystallo-
graphic dimer in the asymmetrical unit, which buries 1136.6 Å2of
surface area at the interface. The same dimerization interface
was also observed in the crystal structures of AgamOBP1 and
AaegOBP1, suggesting the observed dimer might be a physiolog-
ically relevant state that is conserved among mosquito OBPs.
Structure of the MOP-Binding Site. Similar to the ligand-binding
pocket in other OBP structures, a central cavity is observed inside
the protein and is covered exclusively by hydrophobic residues.
However, in contrast to all the previously published ligand-bound
OBP structures, a large part of MOP is not identified inside the
central cavity. Instead, MOP has its long lipid “tail” bound in
a hydrophobic tunnel formedbetween helices4 and 5 and only has
its lactone/acetyl ester “head” sticking into the central cavity (Fig.
2A). The hydrophobic tunnel is lined by Leu73, Leu76, His77, and
Leu80 from helix 4; Met91, Gly92, and Leu96 from helix 5; and
Trp114 from helix6.Because helices 4 and 5 also form thedimeric
interface of the crystal structure, the two MOP-binding tunnels of
each CquiOBP1 dimeric unit meet and connect at the dimer in-
terface. Notably, the same ligand-binding pocket was serendipi-
tously identified in the previous structures of AgamOBP1 and
AaegOBP1 by a bound PEG molecule from the crystallization
helices 1, 3, and 4; runs through the central cavities and the con-
nected hydrophobic tunnels between helices 4 and 5; and comes
out through the second OBP molecule of the dimer (Fig. 2B). All
the residues that line up the MOP-binding tunnel are well con-
served in AgamOBP1 and AaegOBP1. The apparent tunnel cre-
ated by the dimer could well be a solid-state artifact resulting from
the high OBP concentrations used for crystallization. However, in
other insect OBP structures, this tunnel is either blocked or dis-
rupted by mutations and/or relative movement of helices 4 and 5.
Further examination of the complexed crystal structure in-
dicated that most of the lipid chain of MOP was buried in the
tunnel, with the rest of the molecule only occupying about half of
the central cavity—the traditional binding site of OBPs. The rel-
atively less well-defined electron density around the lactone ring
suggests that this part of MOP has several conformations in the
cavity. Still, MOP makes extensive hydrophobic interactions and
van der Waals interactions inside the cavity. The acetyl ester
branch is buried in a hydrophobic patch formed by Tyr10, Leu80,
X-ray crystallographic data collection and refinement
Cell parameters, Å (°)a = 35.92
b = 107.31
c = 38.53
β = 100.38
13.5, 25.8, 24.3
B, Å2(protein, solvent, ligand)
Ramachandran plot, % (favorable,
additional, generous, disallowed)
Parentheses indicate values for the high resolution shell (1.35 to 1.30 Å).
labeled α1 to α6. Conserved disulfide bonds and MOP are shown in ball-and-
stick representations. (B) Overlay of the structures of CquiOBP1 (cyan),
AgamOBP1 (pink), and AaegOBP1 (orange). (C) Chemical structure of MOP.
CquiOBP1·MOP structure. (A) Ribbon diagrams. Individual helices are
Mao et al. PNAS
| November 2, 2010
| vol. 107
| no. 44
Ala88, Met91, His121, and Phe123. Similarly, the lactone ring
makes interactions with Leu15, Leu19, Leu80, His111, Tyr122,
and Phe123. All these residues are also conserved in AgamOBP1,
but in AaegOBP1, Leu15 and Leu19 are replaced with Phe15
and Met19, which are apparently bulkier but equally hydrophobic.
The majority of those interactions are located on one side of the
lactone ring, leaving free space on the other side, which is likely
the cause of the relative flexibility of the lactone ring in the
Intriguingly, residues in the C terminus (e.g., His121, Tyr122,
Phe123) of CquiOBP1 make extensive contact with both the acetyl
ester group and the lactone ring of MOP, acting like a holder that
restrains the MOP in the binding site (Fig. 3). It has been pre-
viously observed in the structures of AgamOBP1 and AaegOBP1
Tyr54 and the δ-nitrogen of His23, making a potential pH-sensing
triad that locks the C terminus onto helix 1 and helix 3 (25). We
the CquiOBP1·MOP complex. It is likely that a drop in pH could
interactions between the C-terminal loop and the rest of the pro-
tein, and therefore displace the C terminus from the central cavity.
That would result in a loss of support of MOP from these residues
and, eventually, release of MOP by the protein. The observation
that MOP occupies part of the long tunnel in the dimeric structure
prompted us to investigate if the same CquiOBP1·MOP complex
exists in solution.
NMR Study of the CquiOBP1·MOP Complex. The
NMR spectrum of15N-labeled CquiOBP1 bound to unlabeled
MOP (CquiOBP1·MOP) at pH 7.0 exhibited close to the expected
number of protein backbone amide resonances (118 of 125) (27).
The large chemical shift dispersion and uniform peak intensities
indicate that the protein complex is structurally homogeneous and
stably folded under NMR conditions. Analysis of NMR relaxation
data reveals that CquiOBP1·MOP is monomeric in solution. A
summary of15N NMR relaxation and heteronuclear NOE data at
pH 7.0 is presented in Fig. 4. The average15N R1and R2values
from residues in structured regions are 0.95 (±0.05) s−1and 14.8
(±0.5) s−1, respectively. Elevated R1values and decreased heter-
onuclear NOE values (<0.65) are apparent for the first eight res-
idues from the N terminus, consistent with significant backbone
flexibility in this region (28). Assuming isotropic tumbling of
CquiOBP1·MOP, the overall rotational correlation time was
obtained from R1/R2ratios of all residues within 1 SD of the av-
erage value(29).Thus,theaverage rotationalcorrelationtime was
calculated to be 9.4 ± 0.5 ns at 298 K, indicating that, contrary to
the crystal structure, the CquiOBP1·MOP complex is monomeric
in solution at pH 7.0 under NMR conditions.
We attempted to perform a detailed spectral analysis of bound
MOP, but the chemical shifts of the aliphatic protons in the
pheromone were severely overlapped with each other and with
aliphatic protons from the protein, thus preventing accurate
NMR assignments for the bound MOP. The structure of the
protein itself in the CquiOBP1·MOP complex was obtained with
more than 95% of the protein NMR resonances assigned (27),
and the assignments have been deposited to the BioMagResBank
(BMRB) repository (accession no. 16175). NMR-derived protein
structures of CquiOBP1·MOP complex were calculated on the
basis of NOE data (from the protein), slowly exchanging amide
protons (NH), chemical shift analysis, and3JNHαspin-spin cou-
pling constants (Methods). The analysis of chemical shift index
(30),3JNHα(31), and hydrogen-deuterium exchange rates of NH
groups (32) determined the secondary structure. Table 2 sum-
marizes the structural statistics calculated for 15 lowest energy
conformers. The ensemble of the 15 lowest energy NMR struc-
tures has a rmsd of 0.68 Å for main chain atoms and 1.17 Å for all
heavy atoms. The energy-minimized average NMR structure of
CquiOBP1·MOP is illustrated in Fig. 5, and the overall fold is
rmsd of the main chain atoms is 1.15 Å when comparing the NMR
and X-ray structures in regions of regular secondary structure.
A long stretch of amino acid residues in helix α5 (Ala112 to
Leu120) exhibits exchange-broadened NMR resonances, sug-
gesting that this region may undergo some type of conformational
exchange. Many of these exchange-broadened residues (Ala112,
Met113, Gly116, and Lys117) are found on the internal surface of
helix α5 and make very close contact with the bound MOP in the
crystal structure. Thus, the apparent flexibility of helix α5 may be
the result of its binding interaction with MOP. We propose that
C-terminal residue Val (V) 125 locks the C-terminal onto MOP, holding the
pheromone molecule in the central cavity. MOP is shown as a stick model in
gray and is overlaid with the same 2Fo− Fcelectron density map as in Fig. 2.
CquiOBP1 is colored in cyan and represented in the ribbon diagram. The side
chains of H23, Y54, and residues in the C-terminal are shown in stick models.
A surface representation of the C-terminal of CquiOBP1 is also shown. All
oxygen atoms in stick models are shown in red, and nitrogen atoms are
shown in blue. Hydrogen bonds are shown as orange dotted lines.
Network of hydrogen bonds between His (H) 23, Tyr (Y) 54, and the
bound dimer. The lipid chain of MOP is bound in a hydrophobic tunnel lo-
cated outside the central cavity. The tunnel from each monomer connects at
the dimer interface. MOP molecules are represented in stick models and
overlaid with a 2Fo − Fc electron density map contoured at 1σ phased
without MOP molecules modeled in. (B) Diagram of the structure of
AgamOBP1 and a bound PEG molecule from the crystallization condition.
PEG is shown in stick models.
Structure of CquiOBP1 binding pocket. (A) Cut-through view of the
| www.pnas.org/cgi/doi/10.1073/pnas.1012274107Mao et al.
conformational fluctuations in helix α5 may function as a gate to
help create an opening to allow entrance of MOP inside the
Our results show that MOP binds to CquiOBP1 by simulta-
neously engaging both the traditional central cavity and a hydro-
in solution, but binding is retained. This unprecedented binding,
which involves more than the binding pocket, has implications on
how mosquito OBPs may contribute to the selectivity of the
We found that the interactions between the MOP and
CquiOBP1 are exclusively hydrophobic interactions and van der
Waals interactions, in direct contrast to the hydrogen bonds ob-
served in the BmorPBP·bombykol complex and LUSH·alcohol
them is making hydrogen bond with the protein. In fact, the
electron density around those functional groups is less defined
the cavity. This result agrees with our previous binding studies (8),
which show that CquiOBP1 binds not only to the natural phero-
mone stereoisomer (5R,6S)-MOP but to its antipode (5S,6R)-
MOP. Because there is no strong recognition of the functional
groups and enough room in the cavity, the antipode can have its
functional groups bind to the cavity in another direction and still
have its lipid chain bound in the hydrophobic tunnel.
The remarkable similarity between the structure of CquiOBP1
and that of AgamOBP1 and AaegOBP1 suggests those two
mosquito OBPs might bind to and recognize their ligands in
a similar fashion. Both AgamOBP1 and AaegOBP1 structures
were solved with a molecule of PEG bound in the position where
MOP is bound in CquiOBP1, suggesting that this unique ligand-
binding site might be universal among mosquito OBPs. Further-
more, all the residues that make contact with MOP in CquiOBP1
are conserved in the other two mosquito OBPs, except for two
conservative mutations in the central cavity of AaegOBP1. We
therefore hypothesize that those two OBPs might also be able to
bind to long-chain fatty acid-derived compounds like MOP. In-
deed, binding assays showed that the three mosquito OBPs,
with similar affinity (Fig. 6A). We further hypothesize that
CquiOBP1 does not recognize the specific functional group of
MOP but rather recognizes the length of lipid chain that fits its
15N NOEs (A), spin-lattice relaxation rate constants (B), and spin-spin re-
laxation rate constants (C) are plotted as a function of residue number. All
data were determined at 600 MHz1H frequency and 298 K.
15N NMR relaxation data for CquiOBP1·MOP at pH 7.0. Steady-state
structures of CquiOBP1·MOP
Structural statistics for the ensemble of 15 NMR
NOE restraints (total)
Medium (1 ≤ |i − j| ≤ 4)
Long (|i − j| > 4)
Dihedral angle restraints (ϕ and ψ)
Hydrogen bond restraints in β-sheet regions
rmsd from ideal geometry
Bond length, Å
Bond angles, °
Most favored region
rmsd of atom position from
β-sheet and α-helical regions
(main chain atoms)
β-sheet and α-helical regions
0.0092 ± 0.00009
2.11 ± 0.001
0.68 ± 0.14 Å
1.17 ± 0.1 Å
structure (magenta) for CquiOBP1·MOP. The rmsd of the main chain atoms is
1.19 Å when comparing the NMR and X-ray structures in the regions of
regular secondary structure.
Overlay of the X-ray crystal structure (cyan) and average NMR
Mao et al.PNAS
| November 2, 2010
| vol. 107
| no. 44
binds aldehydes and geranylacetone but not γ-octalactone (Fig.
6B). Interestingly, octanal showed apparent higher affinity than
nonanal and decanal, thus suggesting that a shorter hydrophobic
chain fits better in the hydrophobic tunnel. Taken together, these
findingssuggestthat CquiOBP1acts more likea “broadband filter”
that can pick up ligands selected on the length of the hydrophobic
chain rather than the functional group. These structural insights
could prove useful in designing the next generation of oviposition
attractants for mosquito control.
In CquiOBP1 structure, we identified the same hydrogen bond
triad in the C terminus of the protein as previously observed in
triad is a pH-sensing “lock” that clamps the “hinge,” the C ter-
minus, onto the bound MOP. The disruption of this hydrogen
bond network at low pH would destabilize the C-terminal loop
and “unlock” the bound ligands. We did not detect any significant
ligand-induced conformational change in the structure of the
AaegOBP1, suggesting that the pheromone-induced conforma-
tional change mechanism of Drosophila LUSH probably does not
apply to CquiOBP1. It is therefore likely that different OBPs
exhibit different ligand-releasing/signal transduction mechanisms
in communicating with ORs.
Materials and Methods
Expression and Purification of CquiOBP1, AaegOBP1, and AgamOBP1. Non-
expression that is known to generate properly folded functional OBPs
(8). Uniformly15N-labeled and13C,15N-labeled CquiOBP1 was expressed in
Escherichia coli and purified by ion exchange and gel filtration chromatog-
raphy as described previously (23). Typically, 5 mg of purified protein was
obtained from a 1-L culture. The identity and integrity of the final protein
Crystallization of the CquiOBP1·MOP Complex. To prepare the CquiOBP1·MOP
complex for crystallization study, concentrated and delipidated CquiOBP1
protein (10 mg/mL) was mixed with MOP at a 5:1 molar ratio and incubated
overnight at 4 °C. Crystals of complexes were obtained by a hanging drop
vapor diffusion method at 25 °C by mixing 2 μL of the sample with 2 μL of
well solution [20% (wt/vol) PEG 4,000, 100 mM Hepes (pH 8.2), 200 mM
MgCl2]. Crystals were transferred into a solution consisting of 15% (vol/vol)
ethylene glycol, 20% (wt/vol) PEG 4,000, 100 mM Hepes (pH 8.2), and 200
mM MgCl2and were flash-frozen in liquid nitrogen.
X-Ray Data Collection, Processing, and Structure Determination. A complete
dataset for the CquiOBP1·MOP crystal was collected at 100 K at the Ad-
vanced Photon Source, Argonne National Laboratory, using the micro-
diffractometer facility on beam line ID-24. Data were processed and scaled
with HKL2000 (33). Molecular replacement was performed with PHASER (34)
of the CCP4 suite (35), using one monomer of the structure of AgamOBP1
(PDB ID code 2ERB) (26) without solvent molecules and ligand as the search
model. A clear solution was found in the cross-rotation function and sub-
sequent translation function. The molecular replacement solution was then
submitted to ARP/wARP (36, 37) for automated model building. The protein
model resulting from this automated procedure was completed and further
refined by interactive rounds of manual fitting in COOT (38) and REFMAC
(37). Inspection of 2Fo− Fcand Fo− Fcelectron density maps confirmed the
presence of MOP molecules, which were modeled after the protein structure
was satisfyingly refined. The CquiOBP1/MOP structure was further refined
with the addition of the restraints for ideal geometry of MOP. Data collec-
tion and refinement statistics are given in Table 1.
NMR Spectroscopy. Samples for NMR analysis were prepared by dissolving
15N- or15N/13C-labeled CquiOBP1 protein (0.5 mM) in 0.3 mL of a 95% H2O/
5% D2O solution containing 10 mM phosphate at pH 7.4. One equivalent of
MOP was added to saturate the protein with MOP. All NMR experiments
were performed at 25 °C on a Bruker Avance 600-MHz spectrometer
equipped with a four-channel interface and triple-resonance cryogenic
parameters: The numbers of complex points and acquisition times were 256
points and 180 ms for15N (F1) and 512 points and 64 ms for1H (F2). As-
signment of backbone and side-chain resonances were obtained by ana-
lyzing the following spectra: HNCACB; HN(CO)CACB,HNCO,CBCA(CO)NH;
HBHA(CO)NH; C(CO)NH-TOCSY; H(CCO)NH-TOCSY,HCCH-TOCSY; and H(CCH)-
COSY. The NMR data were processed using NMRPipe (F. Delaglio, National
Institutes of Health) and analyzed using Sparky (39).
15N-1H HSQC spectrum was recorded with the following
15N NMR Relaxation Measurements.15N R1,15N R2, and15N NOE experiments
were performed on CquiOBP1/MOP at 25 °C using standard pulse sequences
described previously (40). Longitudinal magnetization decay was recorded
using seven different delay times: 0.01, 0.05, 0.15, 0.2, 0.3, 0.4, and 0.8 s.
Transverse magnetization decay was recorded with eight different delays:
0.0, 0.016, 0.032, 0.048, 0.064, 0.08, 0.096, and 0.112 s. To check sample
stability, transverse magnetization decay at 0.032 s was verified to be un-
changed before and after each set of measurements. A recycle delay of 1.5 s
was used in measurements of both15N R1and15N R2experiments. Steady-
state15N NOE values (28) were obtained by recording two sets of spectra in
the presence and absence of a 3-s proton saturation period. The NOE
experiments were repeated three times to calculate the average and SD of
the NOE values. The overall rotational correlation time for backbone amide
motion was determined using the protocol described previously (41).
NMR Structure Calculation. Backbone and side chain NMR resonances were
assigned as described previously (42). Analysis of NOESY data determined
nearly 2,000 interproton distance relationships throughout the protein (43).
The NMR-derived distances and dihedral angles then served as constraints
(Table 2) for calculating the 3D structure of the protein using distance ge-
ometry and restrained molecular dynamics. Structure calculations were per-
formed using the YASAP protocol within X-PLOR (44, 45) as previously
described (46). A total of 1,856 interproton distance constraints were
obtained as described (42) by analysis of13C-edited and15N-edited NOESY-
HSQC spectra (100-ms mixing time) of
addition to the NOE-derived distance constraints, the following constraints
were included in the structure calculation: 196 dihedral angle constraints (ϕ
and ψ) and 122 distance constraints for 61 hydrogen bonds verified by iden-
tifying slowly exchanging amide protons in hydrogen-deuterium exchange
experiments (28). Fifty independent structures were calculated, and the 15
structures of lowest energy were selected. The average total and experi-
average rmsds from an idealized geometry for bonds and angles are 0.0081 Å
and 1.98°. None of the distance and angle constraints were violated by more
than 0.40 Å and 4°, respectively.
13C,15N-labeled CquiOBP1/MOP. In
Binding Assays. Binding was measured by separately incubating 5 μg of
CquiOBP1 (8), AgamOBP1 (26), or AegOBP1 (25) with 1 μL, 3.2 mM MOP in
a 50-μL solution. Likewise, octanal, nonanal, decanal, geranylacetone, and
γ-octalactone were separately incubated with CquiOBP1. The unbound and
bound proteins were separated using an ultracentrifugal device; the ligand
was extracted from the bound protein with hexane after lowering the pH
OBP1s from three mosquito species. Each protein was incubated separately
with the same amount of the Culex attractant. Although CquiOBP1 showed
slightly higher affinity, the OBPs from the malaria mosquito, AgamOBP1, and
the yellow fever mosquito, AaegOBP1, bound with high affinity at high
pH. None of the OBPs showed significant binding at low pH. (B) Binding of
other test ligands to CquiOBP1 at pH 7. Octanal, nonanal, decanal, and ger-
Binding of various ligands to mosquito OBPs. (A) Binding of MOP to
| www.pnas.org/cgi/doi/10.1073/pnas.1012274107 Mao et al.
and quantified by gas chromatography according to a previously reported
“cold-binding assay” (20). Additionally, binding was measured by a com-
petitive binding assay using N-phenyl-1-naphthylamine as a fluorescent re-
porter (47). Fluorescence spectra were recorded on a Shimadzu RF-5301
The coordinates of CquiOBP1·MOP have been deposited in the Protein
Data Bank (3OGN, X-ray crystallography structure; 212C, NMR structure).
ACKNOWLEDGMENTS. We are grateful to laboratory members for insightful
discussion, Dr. Julien Pelletier for copyediting a draft version of the
manuscript, and Dr. Josep Rayo for preparing a sample of15N-CquiOBP1.
This work was partially supported by National Science Foundation Grant
0918177 (to W.S.L.); National Institutes of Health Grants CA024487 (to J.C.),
EY012347 (to J.B.A.), and NS059404 (to J.C.); US Department of Agriculture-
Agriculture and Food Research Institute Grant 2010-65105-20582 (to W.S.L.);
and unrestricted funds (gift) from Bedoukian Research, Inc.
1. Syed Z, Leal WS (2009) Acute olfactory response of Culex mosquitoes to a human- and
bird-derived attractant. Proc Natl Acad Sci USA 106:18803–18808.
2. Chandre F, et al. (1998) Pyrethroid resistance in Culex quinquefasciatus from west
Africa. Med Vet Entomol 12:359–366.
3. Barbosa RMR, Furtado A, Regis L, Leal WS (2010) Evaluation of an oviposition-
stimulating kairomone for the yellow fever mosquito, Aedes aegypti, in Recife, Brazil.
J Vector Ecol 35:204–207.
4. Barbosa RMR, Regis L, Vasconcelos R, Leal WS (2010) Culex mosquitoes (Diptera:
Culicidae) egg laying in traps loaded with Bacillus thuringiensis variety israelensis and
baited with skatole. J Med Entomol 47:345–348.
5. Hughes DT, Pelletier J, Luetje CW, Leal WS (2010) Odorant receptor from the southern
house mosquito narrowly tuned to the oviposition attractant skatole. J Chem Ecol 36:
6. Pelletier J, Hughes DT, Luetje CW, Leal WS (2010) An odorant receptor from the
southern house mosquito Culex pipiens quinquefasciatus sensitive to oviposition
attractants. PLoS ONE 5:e10090.
7. Ishida Y, Cornel AJ, Leal WS (2002) Identification and cloning of a female antenna-
specific odorant-binding protein in the mosquito Culex quinquefasciatus. J Chem
8. Leal WS, et al. (2008) Reverse and conventional chemical ecology approaches for the
development of oviposition attractants for Culex mosquitoes. PLoS ONE 3:e3045.
9. Pelletier J, Leal WS (2009) Genome analysis and expression patterns of odorant-
binding proteins from the Southern House mosquito Culex pipiens quinquefasciatus.
PLoS ONE 4:e6237.
10. Leal WS (2005) Pheromone reception. Top Curr Chem 240:1–36.
11. Laurence BR, Pickett JA (1982) Erythro-6-acetoxy-5-hexadecanolide, the major
compound of a mosquito attractant pheromone. J Chem Soc Chem Commun 59–60.
12. Pelletier J, Guidolin A, Syed Z, Cornel AJ, Leal WS (2010) Knockdown of a mosquito
odorant-binding protein involved in the sensitive detection of oviposition attractants.
J Chem Ecol 36:245–248.
13. Kruse SW, Zhao R, Smith DP, Jones DNM (2003) Structure of a specific alcohol-binding
site defined by the odorant binding protein LUSH from Drosophila melanogaster.
Nat Struct Biol 10:694–700.
14. Mohanty S, Zubkov S, Gronenborn AM (2004) The solution NMR structure of
Antheraea polyphemus PBP provides new insight into pheromone recognition by
pheromone-binding proteins. J Mol Biol 337:443–451.
15. Sandler BH, Nikonova L, Leal WS, Clardy J (2000) Sexual attraction in the silkworm
moth: Structure of the pheromone-binding-protein-bombykol complex. Chem Biol 7:
16. Thode AB, Kruse SW, Nix JC, Jones DNM (2008) The role of multiple hydrogen-
bonding groups in specific alcohol binding sites in proteins: Insights from structural
studies of LUSH. J Mol Biol 376:1360–1376.
17. Lautenschlager C, Leal WS, Clardy J (2007) Bombyx mori pheromone-binding protein
binding nonpheromone ligands: Implications for pheromone recognition. Structure
18. Horst R, et al. (2001) NMR structure reveals intramolecular regulation mechanism for
pheromone binding and release. Proc Natl Acad Sci USA 98:14374–14379.
19. Lautenschlager C, Leal WS, Clardy J (2005) Coil-to-helix transition and ligand release
of Bombyx mori pheromone-binding protein. Biochem Biophys Res Commun 335:
20. Leal WS, et al. (2005) Kinetics and molecular properties of pheromone binding and
release. Proc Natl Acad Sci USA 102:5386–5391.
21. Lee D, et al. (2002) NMR structure of the unliganded Bombyx mori pheromone-
binding protein at physiological pH. FEBS Lett 531:314–318.
22. Wojtasek H, Leal WS (1999) Conformational change in the pheromone-binding
protein from Bombyx mori induced by pH and by interaction with membranes. J Biol
23. Damberger FF, Ishida Y, Leal WS, Wüthrich K (2007) Structural basis of ligand binding
and release in insect pheromone-binding proteins: NMR structure of Antheraea
polyphemus PBP1 at pH 4.5. J Mol Biol 373:811–819.
24. Laughlin JD, Ha TS, Jones DNM, Smith DP (2008) Activation of pheromone-sensitive
neurons is mediated by conformational activation of pheromone-binding protein.
aegypti suggests a binding pocket covered by a pH-sensitive “Lid.” PLoS ONE 4:e8006.
26. Wogulis M, Morgan T, Ishida Y, Leal WS, Wilson DK (2006) The crystal structure of an
odorant binding protein from Anopheles gambiae: Evidence for a common ligand
release mechanism. Biochem Biophys Res Commun 339:157–164.
27. Xu X, et al. (2009) (1)H, (15)N, and (13)C chemical shift assignments of the mosquito
pheromone. Biomol NMR Assign 3:195–197.
28. Ames JB, Tanaka T, Stryer L, Ikura M (1994) Secondary structure of myristoylated
recoverin determined by three-dimensional heteronuclear NMR: Implications for the
calcium-myristoyl switch. Biochemistry 33:10743–10753.
29. Marion D, et al. (1989) Overcoming the overlap problem in the assignment of1H NMR
spectra oflargerproteinsbyuseofthree-dimensional heteronuclear1H-15NHartmann-
Hahn-multiple quantum coherence and nuclear Overhauser-multiple quantum
coherence spectroscopy: Application to interleukin 1 beta. Biochemistry 28:6150–6156.
30. Wishart DS, Sykes BD, Richards FM (1991) Relationship between nuclear magnetic
resonance chemical shift and protein secondary structure. J Mol Biol 222:311–333.
31. AnglisterJ, et al.(1994)1H,13C,15Nnuclear magnetic resonancebackbone assignments
and secondary structure of human calcineurin B. Biochemistry 33:3540–3547.
32. Wuthrich K (1986) NMR of Proteins and Nucleic Acids (John Wiley and Sons, Inc., New
33. Otwinowski Z, Minor W (1997) Processing of x-ray diffraction data collected in
oscillation mode. Methods Enzymol 276:307–326.
34. McCoy AJ, et al. (2007) Phaser crystallographic software. J Appl Cryst 40:658–674.
35. Collaborative Computational Project, Number 4 (1994) The CCP4 suite: Programs for
protein crystallography. Acta Crystallogr D Biol Crystallogr 50:760–763.
36. Langer G, Cohen SX, Lamzin VS, Perrakis A (2008) Automated macromolecular model
building for X-ray crystallography using ARP/wARP version 7. Nat Protoc 3:1171–1179.
37. Murshudov GN, Vagin AA, Dodson EJ (1997) Refinement of macromolecular structures
by the maximum-likelihood method. Acta Crystallogr D Biol Crystallogr 53:240–255.
38. Emsley P, Lohkamp B, Scott WG, Cowtan K (2010) Features and development of Coot.
Acta Crystallogr D Biol Crystallogr 66:486–501.
39. Goddard TD, Kneller DG (2008) SPARKY 3 (University of California, San Francisco).
40. Farrow NA, et al. (1994) Backbone dynamics of a free and phosphopeptide-
complexed Src homology 2 domain studied by 15N NMR relaxation. Biochemistry 33:
41. Freedberg DI, et al. (2002) Rapid structural fluctuations of the free HIV protease flaps
in solution: Relationship to crystal structures and comparison with predictions of
dynamics calculations. Protein Sci 11:221–232.
42. Tanaka T, Ames JB, Kainosho M, Stryer L, Ikura M (1998) Differential isotype labeling
strategy for determining the structure of myristoylated recoverin by NMR
spectroscopy. J Biomol NMR 11:135–152.
43. Clore GM, Gronenborn AM (1997) NMR structures of proteins and protein complexes
beyond 20,000 M(r). Nat Struct Biol 4(Suppl):849–853.
44. Badger J, Kumar RA, Yip P, Szalma S (1999) New features and enhancements in the X-
PLOR computer program. Proteins 35:25–33.
45. Brünger AT (1992) X-PLOR, Version 3.1: A System for X-Ray Crystallography and NMR.
(Yale Univ Press, New Haven, CT).
46. Bagby S, Harvey TS, Eagle SG, Inouye S, Ikura M (1994) NMR-derived three-dimensional
solution structure of protein S complexed with calcium. Structure 2:107–122.
47. Ban L, Zhang L, Yan Y, Pelosi P (2002) Binding properties of a locust’s chemosensory
protein. Biochem Biophys Res Commun 293:50–54.
Mao et al.PNAS
| November 2, 2010
| vol. 107
| no. 44