7/30/2019 nature10236.pdf

1/8

ARTICLEdoi:10.1038/nature10236

Structure of the human histamine H1

receptor complex with doxepinTatsuro Shimamura1,2,3*, Mitsunori Shiroishi1,2,4*, SimoneWeyand1,5,6, HirokazuTsujimoto1,2, Graeme Winter6, Vsevolod Katritch7,Ruben Abagyan7, Vadim Cherezov3, Wei Liu3, Gye Won Han3, Takuya Kobayashi1,2, Raymond C. Stevens3 & So Iwata1,2,5,6,8

The biogenicamine histamine is an importantpharmacological mediatorinvolved in pathophysiologicalprocesses suchasallergies and inflammations. HistamineH1 receptor (H1R) antagonists arevery effectivedrugs alleviating thesymptoms ofallergicreactions. Here we show thecrystal structureof theH1R complexwithdoxepin,a first-generationH1R antagonist.Doxepin sits deep in the ligand-binding pocket and directly interacts with Trp 4286.48, a highly conserved key residuein G-protein-coupled-receptor activation. This well-conserved pocket with mostly hydrophobic nature contributes tothelowselectivity of thefirst-generation compounds. Thepocketis associated with an anion-binding regionoccupied by

a phosphate ion. Docking of various second-generation H1R antagonists reveals that the unique carboxyl group presentin this class of compounds interacts with Lys 1915.39 and/or Lys179ECL2, both of which form part of the anion-bindingregion. This regionis notconserved in other aminergicreceptors, demonstrating how minordifferences in receptors leadto pronounced selectivity differences with small molecules. Ourstudyshedslighton themolecular basis of H1R antagonistspecificity against H1R.

Histamine is a biogenic amine and an important mediator in variousphysiological and pathophysiological conditions such as arousal state,allergy and inflammation13. Histamine exerts its effects through theactivation of four distinct histamine receptors (H1, H2, H3 and H4)that belong to the G-protein-coupled-receptor (GPCR) superfamily.The H1R, originally cloned from bovine H1R

4, is now known to beexpressed in various human tissues including airway, intestinal and

vascular smooth muscle and brain2

. In type I hypersensitivity allergicreactions, H1R is activated by histamine released from mast cells,which are stimulated by various antigens5. Many studies have beenperformed to develop H1R antagonists, known generally as anti-histamines. Many of these compounds inhibit the action of histamineon H1R to alleviate the symptoms of allergic reactions, making H1Rone of the most validated drug targets judging from the number ofdrugs approved6. H1R shows constitutive activity, and H1R antago-nists generally act as inverse agonists for H1R

7,8. Development of H1Rantagonists has progressed through two generations. First-generationdrugs such as pyrilamine and doxepin (Supplementary Fig. 1) areeffective H1R antagonists. These compounds are, however, knownto show considerable side effects such as sedation, dry mouth andarrhythmia, because of penetration across the bloodbrain barrier

and low receptor selectivity. These H1R antagonists can bind not onlyto H1R but also to other aminergic GPCRs, monoamine transportersand cardiac ion channels. Second-generation drugs such as cetirizineand olopatadine (Supplementary Fig.1) are lesssedatingand in generalhave fewer side effects. The improved pharmacology of the second-generation zwitterionic drugs can be attributed to a new carboxylicmoiety, in combination withthe protonated amine, which significantlyreduces brain permeability, although residual central nervous systemeffects are still reported9. The introduction of the carboxyl moiety also

improves the H1R selectivity of these compounds, but certain second-generation H1R antagonistssuch as terfenadinestill show cardio-toxicity becauseof their interaction with cardiac potassiumchannels10,11.

A first-generation H1R antagonist, doxepin, can cause many typesof side effects due to its antagonistic effects on histamine H2 (ref. 12),serotonin 5-HT2, a1-adrenergic, and muscarinic acetylcholine recep-tors13 in addition to the inhibition of the reuptake of serotonin and

noradrenaline14

. Although even rawhomology models of GPCRs mayfacilitate discovery of novel ligands15,16, reliable receptor structures areessential to improve the reliability of the predictions and understandthe structural basis of subtype specificity. Recently determined GPCRstructures have enabled structure-based approaches to modellingligand interactions in the binding pocket1723 and are already yieldingnovel chemotypes predicted by virtual screening of large chemicallibraries24,25. Here we report the 3.1 A resolution structure of theH1RT4-lysozyme fusion protein (H1RT4L) complex with doxepin.Thecrystal structurereveals the atomicdetails of doxepin binding andits inverse agonistic activity. The H1R crystal structure and models ofsecond-generation H1R antagonists will be highly beneficial for guid-ing rational design of ligands that do not penetrate the bloodbrainbarrier while maintaining H1 selectivity.

Overall architecture of H1RIn the H1R construct, T4L

26 was inserted into the third cytoplasmicloop (intracellular loop 3 (ICL3)) (Gln222Gly 404) and 19 residueswere truncated from the amino-terminal region (Met 1Lys 19) (seeMethods). H1RT4L showed similar binding affinities for H1R antago-nists and for histamine as the wild-type H1R expressed in yeast cells(Supplementary Table 1) and in COS-7 cells27. The structure of theH1RT4L crystals obtained in the lipidic cubic phase (see Methods)

*These authors contributed equally to this work.

1HumanReceptor CrystallographyProject,ERATO, JapanScience andTechnology Agency, Yoshidakonoe-cho,Sakyo-ku, Kyoto606-8501, Japan.2Departmentof CellBiology,GraduateSchoolof Medicine,

KyotoUniversity, Yoshidakonoe-cho, Sakyo-Ku,Kyoto 606-8501, Japan.3Departmentof MolecularBiology,The ScrippsResearchInstitute,10550NorthTorreyPines Road, LaJolla,California92037, USA.4Graduate School of Pharmaceutical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan.5Division of Molecular Biosciences, Membrane Protein Crystallography Group,

Imperial College, London SW7 2AZ, UK. 6Diamond Light Source, Harwell Science and Innovation Campus, Chilton, Didcot, Oxfordshire OX11 0DE, UK. 7Skaggs School of Pharmacy and Pharmaceutical

Sciences and San Diego Supercomputer Center, University of California, San Diego, La Jolla, California 92093, USA. 8Systems and Structural Biology Center, RIKEN, 1-7-22 Suehiro-cho Tsurumi-ku,

Yokohama 230-0045 Japan.

7 J U L Y 2 0 1 1 | V O L 4 7 5 | N A T U R E | 6 5

Macmillan Publishers Limited. All rights reserved2011

http://www.nature.com/doifinder/10.1038/nature10236http://www.nature.com/doifinder/10.1038/nature10236

7/30/2019 nature10236.pdf

2/8

was determined in complex with the H1R antagonist doxepin at 3.1Aresolution (Supplementary Table 2).

H1R is structurally most similar to the aminergic receptors (Fig. 1):b2-adrenergic (b2-AR)

18, b1-adrenergic (b1-AR)19 and dopamine D3

(D3R)23 receptors, while having larger deviations from the morephylogenetically distant rhodopsin17,21, A2A adenosine receptor(A2AAR)

20 and chemokine receptor CXCR4 (ref. 22) (Supplemen-tary Table 3). H1R also shares common motifs with other GPCRs

including D(E)RY in helix III, CWxP in helix VI and NPxxY in helixVII, as well as a disulphide bond connecting extracellular loop 2(ECL2) with the extracellular end of helix III (Cys 1003.25 toCys 180; superscripts indicate residue numbers as per theBallesterosWeinstein scheme28) but lacks the palmitoylation site atthe end of helix VIII found in many other GPCRs29.

Previous GPCR structures revealed that not only the residues in thetransmembrane segments but also those in the loops are critical forligand specificity1723. ECL2 connecting helicesIV and V is attached tohelix III through a disulphide bond between Cys 180 in ECL2 andCys1003.25 in helix III. Seven residues (Phe 168Val 174) before thedisulphide are not included in the structure, as they did not haveinterpretable densities. A section of ECL2, between the disulphidebridge and the extracellular end of helix V, is particularly important

because it is located at the entrance to the ligand-binding pocket. Thissection of ECL2 contains 7 amino acids in H1R, as compared to 5 inb2-AR, 4 in D3R, and 8 in A2AAR. The extra length of this ECL2section is apparently accommodated by the increased distancebetween the extracellular ends of helices III and V by,1.5A and,3.1A when compared to b2-AR and D3R, respectively (Fig. 1b, c).This creates more space within the ligand-binding pocket, which cannow accommodate the larger second-generation H1R antagonists, asdiscussed later.

Some unique features are also observed in the transmembranesegments. A conserved Pro 1614.59-induced kink in helix IV forms atight i13 helical turn, instead of i14 as in b2-AR and D3R (Fig. 2a).This tighter turn allows accommodation of a bulky Trp side chain atposition 4.56, which seems to be essential for ligand specificity of

aminergic GPCRs because this position is occupied by Ser in b2-AR

andD3R, andmutations of this Trpin guinea-pig H1RtoAla,MetandPhe reduce affinity against the antagonist pyrilamine30.

The ionic lock, a salt bridge between Arg3.50 in the conservedD(E)R3.50Y motif and Asp/Glu6.30, which is suggested to stabilizethe inactive conformation, was observed in rhodopsin structures17,21

and D3R23, but was broken in all the other GPCRs1820,22. In H1R,Arg1253.50 of the D(E)R3.50Y motif does not form a salt bridge eitherwith Glu4106.30 or with Asp 1243.49. Instead, the side chain ofArg1253.50 adopts in a new conformer relative to previous structures,forming a hydrogen bond to Gln4166.36 in helix VI (Fig. 2b). Different

structures of the ionic lock regions of the receptors could be causedby modifications of ICL3. Otherwise, they might be related to thedifferent levels of constitutive activities of the receptors.

Doxepin isomers and conformersThe doxepin used in this study contains a mixture of E and Z isomers,and each isomer can take two distinct rotational conformers of thedibenzo[b,e]oxepin ring, resulting in four distinct conformers (con-formers 14; Supplementary Fig. 2). Two conformers, one E isomer(conformer1) andone Z isomer(conformer 4),fit the electron densitybetter than the other two (Supplementary Fig. 3). This result is alsoconsistent with the Rfree and the averaged B-factor values for eachconformer (Supplementary Table 4). A 1:1 mixture of the E and Zisomers was used in the refinement. The two conformers are indis-

tinguishable at this resolution and have nearly identical interactions

ECL1

N

C

ECL2ECL3

ICL1

ICL2ICL3

IIIVII

VIII

III

VIV

IV

CWxP

D(E)RY

NPxxY

PO43

Doxepin

D107

W428

SS

SS

a b

c

H1R2-AR

H1R

D3R

ECL2

Doxepin

Doxepin

III

V

ECL1

ECL2

ECL3

ECL1

IV

I

VIVII

VIII

II

ECL3

III

V

IV

I

VI VII

VIII

II

N

N

C

C

Figure 1 | Structureof H1R complex with doxepin. a, Ribbon representationof the H1R structure. Doxepin is shown as yellow spheres whereas thephosphate ion is shown as spheres withthe carbonand oxygenatoms colouredorange and red, respectively. Disulphide bonds are shown as yellow sticks, andTrp 428 and Asp 107 are shown as pink sticks. Three conserved motifsD(E)R3.50Y, CWxP6.50 and NP7.50xxY are highlighted in blue.b, Superimposition of the H1R (green) and b2-AR (cyan) structures. c, Same as

b but with the D3R structure coloured violet.

a

b

H1R

2-AR

D3R

P1614.59

Doxepin

W1584.56

Carazolol

Eticlopride

Q4166.36

E4106.30

R1253.50

D1243.49

III

IV

S1654.57

S1654.57

VI

Figure 2 | Comparison of thestructuresof H1R, b2-ARand D3R. a, Proline-induced kink in helix IV (H1R, green; b2-AR, cyan; D3R, magenta). The sidechain of Trp1584.56 and Pro1614.59 of H1R and the equivalent residues ofb2-AR (Ser 165

4.57 and Pro 1684.60) and ofD3R(Ser 1654.57 and Pro 1674.59)arealso shown. b, Variations in the D(E)RY motif structures of H1R, b2-AR andD3R are coloured in green, cyan and magenta, respectively. Side chains ofAsp1243.49, Arg1253.50, Glu4106.30 of H1R and the equivalents in b2-AR andD3R are represented as stick models. For H1R, Gln416

6.36, which forms a

hydrogen bond with Arg 1253.50, is also shown. Possible hydrogen bonds areindicated by dotted lines.

RESEARCH ARTICLE

6 6 | N A T U R E | V O L 4 7 5 | 7 J U L Y 2 0 1 1

Macmillan Publishers Limited. All rights reserved2011

7/30/2019 nature10236.pdf

3/8

with the binding pocket; consequently, in the following sections weare referring to the E isomer, unless noted otherwise.

Ligand-binding pocketDoxepin binds in a pocket mainly definedby the side chains of helicesIII, V and VI (Fig. 3a, b). Asp 1073.32, a strictly conserved residue inaminergic receptors (Supplementary Table 5), forms an anchor saltbridge with the amine moiety of the ligand. This interaction has been

reported to be essential for the binding of H1R antagonists as well asagonists in mutationalstudies3133. This amine moiety is connected viaa flexible carbon chain to the tricyclic dibenzo[b,e]oxepin ring in ahydrophobic pocket comprised of the side chains of helices III, V andVI. The tricyclic ring of doxepin sits much deeper (by,5 A) in thebinding pocket than the ligands in the other non-rhodopsin GPCRstructures (Fig. 3c). The ligand is surrounded mainly by highly con-served residues among aminergic receptors, including Ile 1153.40,Phe 4246.44, Trp4286.48 and Phe 4326.52, whereas the non-conservedresidues Trp 1584.56 and Asn 1985.46 in the pocket make only minorhydrophobic interactions with doxepin (Fig. 3a,b). The importance ofa large side chain at position 6.52 hasbeen suggested forthe binding ofpyrilamine30,33. Thr1123.37 can form a hydrogen bond to the oxygenatom of the E isomer (but not the Z isomer) of doxepin, as shown in

Fig. 3a, b. A suboptimal geometry and the bifurcated nature of thishydrogen bond indicate that it does not contribute significantly tobinding affinity, which is supported by lack of this hydrogen-bondinteraction in the H1Rolopatadine complex, as described later. This

well-conserved pocket and its mostly hydrophobic nature shouldcontribute to low selectivity of doxepin and other first-generationH1R-antagonists

13,32. Moreover, because of its deep binding position,doxepin does not interact with ECL2, whose highly variable primaryand tertiary structures are known to contribute to binding specificityof GPCR ligands34.

A novel feature of the H1Rdoxepin complex is the existence of ananion-binding site at the entrance to the ligand-binding pocket

(Fig. 3d). We modelled a phosphate ion, which is present at a highconcentration in the crystallization buffer (300 mM ammoniumphosphate), into the observed strong density in the site. This modelis supported by the fact that a phosphate ion affects the binding ofsomeligands and the stabilityof H1R (Supplementary Tables 1 and 6).The phosphate ion is coordinated by Lys 179ECL2, Lys 1915.39,Tyr4316.51 and His4507.35; all of which, except for Tyr 4316.51, areunique to H1R (Supplementary Table 5). This encasement of theligand in the pocket combined with an ionic interaction betweenthe phosphate ion and the tertiary amine of doxepin (NO distance4.8A) indicates that a phosphate ion may serve as a positive modu-lator of ligand binding. This hypothesis has been validated by com-paring thermostability (Supplementary Table 6) and ligand affinity(Supplementary Table 1) in buffers with and without phosphate.

Thermostability of the receptor is increased in the presence of phos-phate for all ligands except for cetirizine, which can prevent the phos-phate binding, according to the modelling study discussed later. Thephosphate effect is observed at a concentration as low as 1.5 mM,

NH+

O

Y4587.43D1073.32

Y4316.51

F4356.55Y1083.33

S1113.36W4286.48

F4326.52

T1945.43

W1584.56

N1985.46T1123.37

F4246.44

I1153.40

a b

III

II VII

VI

V

IV T1945.43N1985.46

W1584.56

T1123.37Y108

3.33

I1153.40

Doxepin

F4356.55F4326.52

Y4316.51W4286.48

F4246.44

Y4587.43

D1073.32

PO43

c

V

VI

W4286.48

Doxepin

Carazolol

Eticlopride

ZM241385

d

H4507.35

K179ECL2

K1915.39

Y4316.51

PO43

ECL2

III

VI

VII

ECL3

V

Doxepin

A1955.44

Figure 3 | Binding interactions of doxepin. a, Doxepin is shown as stickswith yellow carbon atoms, whereas the contact residues within 4 A are shownwith grey carbonatoms. Nitrogen andoxygen atoms arecolouredblueand red,respectively. Hydrogen bonds/salt bridges are indicated as blue dotted lines.

b, Doxepin binding interactions. Hydrophobic interactions are shown in black

dotted lines. c, Ligand-binding positions in non-rhodopsin GPCRs. Carbonatoms of doxepin (H1R) are shown in yellow, carazolol (b2-AR) in cyan,eticlopride (D3R) in magenta and ZM241385 (A2AAR) in grey. d, Structure ofthe anion-binding region with a phosphate ion.

ARTICLE RESEARCH

7 J U L Y 2 0 1 1 | V O L 4 7 5 | N A T U R E | 6 7

Macmillan Publishers Limited. All rights reserved2011

7/30/2019 nature10236.pdf

4/8

indicating its physiological relevance. The affinity of histamineand pyrilamine to the receptors also increased in the presence ofphosphate.

H1 selectivity of H1R antagonistsSupplementary Fig. 1 lists the first and second generations of H1Rantagonists. It has been shown that second-generation H1R antagonistsare much more specific to H1R and show much lower affinity to the

otheraminergicreceptors32,35

. H1R-antagonist specificityhasbeenprevi-ously analysed using H1R-homology models based on the bacteriorho-dopsin or bovine rhodopsin crystal structure in combination with theH1R antagonist pharmacophore model and mutational studies

30,36,37.These studies have successfully determined some residues importantfor selectivity, including Lys1915.39; however, contributions of theECL residues have not been examined because these loops could notbe modelled accurately based on the bacteriorhodopsin or bovine rho-dopsin structure. Our H1R structure with the extracellular loops shouldsignificantly improve understanding of H1R-antagonist selectivity.Using flexible ligand-receptor docking38,39 in the ICM molecularmodelling package40 (alsoseeMethods),wehavestudiedH1Rselectivityfor representative second-generation zwitterionic H1R-antagonists:olopatadine, acrivastine, R-cetirizine (levocetirizine) and fexofenadine

(Fig. 4). Olopatadine (Fig. 4a) is a close doxepin analogue with amethylcarboxyl substitution in one of its benzene rings. Its bindingmode closely resembles doxepin, while the carboxyl group extends outof the pocket towards the extracellular space and interacts withLys1915.39 and Tyr1083.33 without displacing the phosphate ion.These additional interactions can explain a reduced effect of the muta-tion of the conserved Asp1073.32 to Ala on olopatadine binding(14-fold for olopatadine as compared to 280-fold for doxepin)32,41.

The orientation of the carboxyl moiety in the ECL region dictates thattheoxygen atom of thedibenzo[b,e]oxepin ring is in a position where itcannotforma hydrogen bond with Thr1123.37. Althoughthe marketeddrug is only the Z isomer, both olopatadine Z and E isomers showsimilar H1R affinities

41.Acrivastin (Fig. 4b) has a different chemical scaffold with a

carboxyl group in its pyridine ring. Its longer carbon chain positionsthe carboxyl group higher in the ECL region, where it can form saltbridges to both Lys 1915.39andLys 179ECL2 amine moieties.R-cetirizine(Fig. 4c) has its carboxylic moiety attached directly to a piperazineamino group. The conformational modelling indicates that thecarboxyl moiety can reach towards the ECL region, forming saltbridges to Lys 1915.39 and Lys 179ECL2. Finally, fexofenadine (R iso-mer; Fig. 4d) has the most extended carboxyl-containing substituent,

PO43

K179ECL2

K1915.39

Y4587.43

D1073.32

S1113.36

Y1083.33

W1584.56

T1945.43

A1955.44

N1985.46

T1123.37 I1153.40

VII

IIIIV

V

VID178ECL2 Y4316.51

H4507.35

a b

c d

K179ECL2

K1915.39

K179ECL2

K1915.39

K179ECL2

K1915.39

D1073.32

Y4587.43

Y4587.43

D1073.32

Y4587.43

D1073.32

Y1083.33

Figure 4 | Interactions of second-generation selective H1R antagonists withthe H1R ligand-binding pocket. ad, Conformation of each complex waspredicted by global optimization of the ligand in the all-atom flexible H1Rmodel3840 based on the H1Rdoxepin complex structure. Carbon atoms for

Z-olopatadine co-bound with the phosphate ion (a), acrivastine

(b), R-cetirizine (levocetirizine) (c) and fexofenadine (d) are coloured violet.Nitrogen and oxygen atoms are coloured blue and red, respectively. Ligandcontact residues of H1R are shown with grey carbon atoms; parts of helices III,IV and ECL2 are not shown for clarity. Hydrogen bonds are shown in cyan.

RESEARCH ARTICLE

6 8 | N A T U R E | V O L 4 7 5 | 7 J U L Y 2 0 1 1

Macmillan Publishers Limited. All rights reserved2011

7/30/2019 nature10236.pdf

5/8

whichreaches outside of the binding cavityand forms a salt bridge toLys 1915.39.

Modelling of second-generation H1R-antagonist binding to H1Rindicates that no significant protein backbone rearrangements arerequired to accommodate these diverse ligands. The enhanced H1Rselectivity of these compounds32,35 can be explained by the specificinteraction of the carboxyl group with Lys residues in the ECL region,unique to H1R. The result also shows a good agreement with earlier

modelling and site-directed mutagenesis studies. Lys 1915.39

is knownto be important for increasing affinity for some of these ligands 30,42,43,whereas the involvement of Lys179ECL2 was suggested in themodelling study of 8R-lisuride into the ligand-binding pocket44.Our modelling results also indicate that olopatadine is the onlysecond-generation compound studied here for which the carboxylmoiety does not interfere with phosphate binding. The results are alsosupported by the fact that the presence of the phosphate ion increasedthe thermal stability of the H1Rdoxepin or H1Rolopatadine com-plexes, whereas it did not affect the stability of the H 1Rcetirizinecomplex (Supplementary Table 6).

Mechanism of H1R inactivationH1R antagonists act as highly effective inverse agonists of H1R, which

reduce basal activity of the receptor and therefore are expected tointerfere with the key molecular switches involved in the GPCRactivation mechanism. One of the switches is represented by Trp6.48

of the conserved CWxP6.50 motif, which helps to stabilize rhodopsinin its inactive dark state through a direct interaction with retinal. Therecently published structure of the active-state A2AAR

45 also showedthat Trp6.48 participates in the activation-related conformationalchanges, where a small ligand-induced shift of Trp6.48 was observedin concert with the large movement of the intracellular part of helixVI. In other receptors, the roleof Trp6.48 is less obvious; forinstance, itlacks direct ligand interactions with either inverse agonists or fullagonists ofb2-AR

46. It is interesting to note that in the H1R structure,like in inactive rhodopsin, the H1R-antagonist doxepin does makeextensive hydrophobic interactions with the Trp 4286.48 rings, which

is unique among the known non-rhodopsin GPCR structures andcould stabilize the hydrophobic packing around helix VI (Fig. 3c).Another important ligand-induced switch described in b2-AR isactivation-related contraction of the extracellular ligand-bindingpocket37. Because the natural agonist histamine is much smaller thanbulky H1R antagonists, some contraction of the binding pocket islikely to accompany ligand-induced H1R activation. Bulky com-pounds, capable of blocking both activation-related contraction ofthe pocket and the Trp 4286.48 switch would be very efficient inlocking H1R in an inactive conformation, which is likely to explainthe as much as 78% reduction of H1R basal activity by some H1Rantagonists8.

METHODS SUMMARY

H1RT4L was expressed in yeast Pichia pastoris. Ligand-binding assays wereperformedas described in Methods. P. pastoris membranes were solubilizedusing1% (w/v) n-dodecyl-B-D-maltopyranoside and 0.2% (w/v) cholesteryl hemisucci-nate, and purified by immobilized metal ion affinity chromatography (IMAC).After IMAC, the C-terminal GFP was cleaved by Tobacco Etch virus (TEV)protease. Then the sample mixture was passed through IMAC to remove thecleaved His-tagged GFP and TEV protease. Receptor crystallization was per-formed by the lipidic cubic phase (LCP) method. The protein-LCP mixture con-tained 40% (w/w) receptor solution, 54% (w/w) monoolein, and 6% (w/w)cholesterol. Crystals were grown in 4050 nl protein-laden LCP boluses overlaidby 0.8ml of precipitant solution (2630% (v/v) PEG400, 300 mM ammoniumphosphate, 10 mM MgCl2, 100mM Na-citrate pH 4.5 and 1 mM doxepin) at20 uC. Crystals were harvested directly from LCPmatrix andflashfrozen inliquidnitrogen. X-ray diffraction data were collected at 100 K with a beam size of103 10mm on the microfocus beamline I24 at the Diamond Light Source(UK). Data collection, processing, structure solution and refinement are

described in Methods.

Full Methods and any associated references are available in the online version ofthe paper at www.nature.com/nature .

Received 15 March; accepted 1 June 2011.

Published online 22 June 2011.

1. Schwartz, J. C., Arrang, J. M., Garbarg, M., Pollard, H. & Ruat, M. Histaminergictransmission in the mammalianbrain. Physiol. Rev. 71, 151(1991).

2. Hill, S. J. Distribution,properties, and functional characteristicsof three classes ofhistamine receptor. Pharmacol. Rev. 42, 4583 (1990).

3. Hill,S.J. etal. International Union of Pharmacology. XIII.Classification of histaminereceptors. Pharmacol. Rev. 49, 253278 (1997).4. Yamashita,M. et al. Expression cloning of a cDNA encoding the bovine histamine

H1 receptor. Proc. Natl Acad. Sci. USA 88, 1151511519 (1991).5. Simons, F. E. Advances in H1-antihistamines. N. Engl. J. Med. 351, 22032217

(2004).6. Overington,J. P.,Al-Lazikani,B. & Hopkins,A. L. Howmanydrugtargets are there?

Nature Rev. Drug Discov. 5, 993996 (2006).7. Bakker, R. A., Wieland, K., Timmerman, H. & Leurs, R. Constitutive activity of the

histamine H1 receptor reveals inverse agonism of histamine H1 receptorantagonists. Eur. J. Pharmacol. 387, R5R7 (2000).

8. Bakker,R. A., Schoonus, S.B.,Smit, M.J., Timmerman, H.& Leurs, R.HistamineH1-receptor activation of nuclear factor-kB: roles for Gbc- and Gaq/11-subunits inconstitutive and agonist-mediated signaling. Mol. Pharmacol. 60, 11331142(2001).

9. Tashiro, M. et al. Dose dependency of brain histamine H1 receptor occupancyfollowing oraladministrationof cetirizine hydrochloride measured using PETwith[11C]doxepin. Hum. Psychopharmacol. 24, 540548 (2009).

10. Woosley, R. L., Chen, Y., Freiman,J. P. & Gillis, R. A. Mechanism of the cardiotoxic

actions of terfenadine. J. Am. Med. Assoc. 269, 15321536 (1993).11. Yap,Y. G.& Camm,A. J.Potentialcardiac toxicityof H1-antihistamines. Clin.Allergy

Immunol. 17, 389419 (2002).12. Okamura, N. et al. Functional neuroimaging of cognition impaired by a classical

antihistamine, d-chlorpheniramine. Br. J. Pharmacol. 129, 115123 (2000).13. Cusack, B., Nelson, A. & Richelson, E. Binding of antidepressants to human brain

receptors: focus on newer generation compounds. Psychopharmacology (Berl.)114, 559565 (1994).

14. Sarker,S. etal. Thehigh-affinitybinding sitefor tricyclicantidepressants residesinthe outer vestibule of the serotonin transporter. Mol. Pharmacol. 78, 10261035(2010).

15. Klabunde, T. & Hessler, G. Drug design strategies for targeting G-protein-coupledreceptors. ChemBioChem 3, 928944 (2002).

16. de Graaf, C., Rognan, D. & Customizing, G. Protein-coupled receptor models forstructure-based virtual screening. Curr. Pharm. Des. 15, 40264048 (2009).

17. Palczewski, K. et al. Crystal structure of rhodopsin: a G protein-coupled receptor.Science 289, 739745 (2000).

18. Cherezov, V. et al. High-resolution crystal structure of an engineered human b2-adrenergic G protein-coupled receptor. Science 318, 12581265 (2007).

19. Warne, T. et al. Structure of a b1-adrenergic G-protein-coupled receptor. Nature454, 486491 (2008).

20. Jaakola, V. P. et al. The 2.6 angstrom crystal structure of a human A2A adenosinereceptor bound to an antagonist. Science 322, 12111217 (2008).

21. Murakami, M. & Kouyama, T. Crystal structure of squid rhodopsin. Nature 453,363367 (2008).

22. Wu, B. et al. Structures of the CXCR4 chemokine GPCR with small-molecule andcyclic peptide antagonists. Science 330, 10661071 (2010).

23. Chien,E. Y. etal. Structureof thehuman dopamineD3 receptor in complex withaD2/D3 selective antagonist. Science 330, 10911095 (2010).

24. Kolb, P. et al. Structure-based discovery of b2-adrenergic receptor ligands. Proc.Natl Acad. Sci. USA 106, 68436848 (2009).

25. Katritch, V. etal. Structure-based discovery of novel chemotypes for adenosine A2Areceptor antagonists. J. Med. Chem. 53, 17991809 (2010).

26. Rosenbaum, D. M. et al. GPCR engineering yields high-resolution structuralinsights intob2-adrenergic receptor function. Science 318, 12661273 (2007).

27. Ratnala, V. R. et al. Large-scale overproduction, functional purification and ligandaffinities of the His-tagged human histamine H1 receptor. Eur. J. Biochem. 271,

26362646 (2004).28. Ballesteros, J. A. & Weinstein, H. Integrated methods for the construction of three-

dimensional models and computational probingof structurefunction relationsinG protein-coupled receptors. Methods Neurosci. 25, 366428 (1995).

29. Qanbar, R. & Bouvier, M. Role of palmitoylation/depalmitoylation reactions inG-protein-coupled receptor function. Pharmacol. Ther. 97, 133 (2003).

30. Wieland,K. etal. Mutationalanalysis ofthe antagonist-bindingsiteof thehistamineH1 receptor. J. Biol. Chem. 274, 2999430000 (1999).

31. Ohta, K. et al. Site-directed mutagenesis of the histamine H1 receptor: roles ofasparticacid107, asparagine198and threonine194. Biochem. Biophys.Res. Commun.203, 10961101 (1994).

32. Nonaka, H. etal. Uniquebindingpocketfor KW-4679in thehistamineH1 receptor.Eur. J. Pharmacol. 345, 111117 (1998).

33. Bruysters, M. et al. Mutational analysis of the histamine H1-receptor bindingpocket of histaprodifens. Eur. J. Pharmacol. 487, 5563 (2004).

34. Peeters, M. C., van Westen, G. J., Li, Q. & Ijzerman, A. P. Importance of theextracellular loops in G protein-coupled receptors for ligand recognition andreceptor activation. Trends Pharmacol. Sci. 32, 3542 (2011).

35. Gillard,M. etal. H1 antagonists: receptoraffinity versusselectivity. Inflamm. Res. 52(Suppl. 1), S49S50 (2003).

ARTICLE RESEARCH

7 J U L Y 2 0 1 1 | V O L 4 7 5 | N A T U R E | 6 9

Macmillan Publishers Limited. All rights reserved2011

http://www.nature.com/naturehttp://www.nature.com/nature

7/30/2019 nature10236.pdf

6/8

36. Kiss,R.,Kovari,Z. & Keseru,G. M.Homologymodellingandbinding sitemappingofthe human histamine H1 receptor. Eur. J. Med. Chem. 39, 959967 (2004).

37. Jongejan, A. & Leurs, R. Delineation of receptorligand interactions at the humanhistamineH1 receptorby a combined approachof site-directed mutagenesisandcomputational techniquesorhow to bind the H1 receptor.Arch. Pharm.(Weinheim) 338, 248259 (2005).

38. Totrov, M. & Abagyan, R. Flexible proteinligand docking by global energyoptimization in internal coordinates. Proteins (Suppl. 1) 29, 215220 (1997).

39. Katritch, V. et al. Analysis of full and partial agonists binding to b2-adrenergicreceptor suggests a role of transmembrane helix V in agonist-specificconformational changes. J. Mol. Recognit. 22, 307318 (2009).

40. Katritch, V., Kufareva, I. & Abagyan, R. Structure based prediction of subtype-selectivity for adenosine receptor antagonists. Neuropharmacology60, 108115(2011).

41. Matsumoto, Y., Funahashi, J., Mori, K., Hayashi, K. & Yano, H. The noncompetitiveantagonism of histamine H1 receptors expressed in Chinese hamster ovary cellsby olopatadine hydrochloride: its potency and molecular mechanism.Pharmacology81, 266274 (2008).

42. Gillard, M., Van Der Perren, C., Moguilevsky,N., Massingham, R. & Chatelain, P.Binding characteristics of cetirizine and levocetirizine to human H1 histaminereceptors: contributionof Lys191andThr194. Mol.Pharmacol. 61, 391399 (2002).

43. Leurs,R., Smit, M. J.,Meeder, R.,TerLaak, A.M. & Timmerman,H. Lysine200 locatedin the fifth transmembrane domain of the histamine H1 receptor interacts withhistaminebut not withall H1 agonists. Biochem. Biophys. Res. Commun. 214,110117 (1995).

44. Bakker, R. A. et al. 8R-lisuride is a potent stereospecific histamine H1-receptorpartial agonist. Mol. Pharmacol. 65, 538549 (2004).

45. Xu, F. etal. Structureof an agonist-bound human A2A adenosine receptor. Science332, 322327 (2011).

46. Rasmussen, S. G. et al. Structure of a nanobody-stabilized active state of theb2

adrenoceptor. Nature 469, 175180 (2011).

Supplementary Information is linked to the online version of the paper atwww.nature.com/nature.

Acknowledgements This work was supported by the ERATO Human ReceptorCrystallography Project from the Japan Science and Technology Agency and by theTargeted Proteins Research Program of MEXT (S.I.), Japan; the NIH Common Fundgrant P50GM073197 for technology development (R.C.S.)and NIH PSI:Biologygrant

U54 GM094618 (R.C.S., V.C., V.K. and R.A.); R.A. was also partly funded by NIH R01GM071872. The work was also partly funded by the Biotechnology and BiologicalSciences Research Council(BBSRC)BB/G023425/1 (S.I.),Grant-in-Aidfor challengingExploratory Research (T.S.), the Mochida Memorial Foundation for Medical andPharmaceutical Research (T.S. and T.K.), Takeda Scientific Foundation (M.S.) and theSumitomo Foundation (T.K.). A part of the work was performed in the MembraneProtein Laboratory funded by the Wellcome Trust (grant 062164/ Z/00/Z) at theDiamond Light Source Limited and at The Scripps Research Institute. We thankD. Axford,R. OwenandG. Evans forhelpwith datacollectionat I24of theDiamond LightSource Limited,H. Wufor helpwiththe preparation of Supplementary Fig. 1 andQ. Xufor help on validation of data processing and A. Walker for assistance withmanuscript

preparation. The authors acknowledge Y. Zheng (The Ohio State University) andM. Caffrey, TrinityCollege (Dublin,Ireland), forthe loan of the inmeso robot(built withsupport from the NIH (GM075915), the National Science Foundation (IIS0308078),and Science Foundation Ireland (02-IN1-B266)). S.I. is thankful for the help ofL. E. Johnson, a co-founder of the Diamond-MPL Project and R. Tanaka, the technicalcoordinator of the ERATO Human Receptor Crystallography Project.

Author Contributions T.S. purified and crystallized the receptor in LCP, optimizedcrystallization conditions, grew crystalsfor data collection, solved and refined thestructure, and prepared the manuscript. M.S. designed, characterized and screenedthe constructs, purified the receptor, and prepared the manuscript. S.W. and S.I.collected the data and processed diffraction data with G.W.H. H.T. expressed thereceptor, prepared the membrane, and performed the ligand-binding assay. V.K. andR.A. performed flexible ligand-receptor docking, and prepared the manuscript. V.C.assisted with the crystallization in LCP and prepared the manuscript. W.L. performedthe thermal stability assay and assisted with the crystallization in LCP. G.W.H. refinedthe structure and assisted with preparing the manuscript. T.K. designed the receptorproduction strategy and assisted with preparing the manuscript. R.C.S. and S.I. wereresponsible for the overall project strategy and management and wrote the

manuscript.

Author Information Thecoordinates and thestructure factors havebeen deposited inthe Protein Data Bank under the accession code 3RZE. Reprints and permissionsinformation is available at www.nature.com/reprints. The authors declare nocompeting financial interests. Readersare welcome to comment on theonline versionof this article at www.nature.com/nature. Correspondence and requests for materialsshould be addressed to R.C.S. (stevens@scripps.edu) , S.I. (so_iwata@me.com) or T.K.(t-coba@mfour.med.kyoto-u.ac.jp) .

RESEARCH ARTICLE

7 0 | N A T U R E | V O L 4 7 5 | 7 J U L Y 2 0 1 1

Macmillan Publishers Limited. All rights reserved2011

http://www.nature.com/naturehttp://www.nature.com/reprintshttp://www.nature.com/naturemailto:stevens@scripps.edumailto:so_iwata@me.commailto:t-coba@mfour.med.kyoto-u.ac.jpmailto:t-coba@mfour.med.kyoto-u.ac.jpmailto:so_iwata@me.commailto:stevens@scripps.eduhttp://www.nature.com/naturehttp://www.nature.com/reprintshttp://www.nature.com/nature

7/30/2019 nature10236.pdf

7/8

METHODSConstruction of the H1R expression vectors for Pichia pastoris. The codingsequence of the full-length human histamine H1 receptor (H1R-fl), in whichN-linked glycosylation sites (Asn5 and Asn 18) were mutated to glutamines,was synthesized with optimization of codon usage for P. pastoris (TAKARABio), and cloned into the pPIC9K expression vector (Invitrogen). The H 1RT4L construct with an N-terminal 19-residue deletion and insertion ofcysteine-less (C54T, C97A) T4 lysozyme into the third intracellular loop wasgenerated by the yeast homologous recombination technique in Saccharomyces

cerevisiae with the SmaI linearized plasmid pDDGFP2 (ref. 47) and three PCRproducts with ,30 bp overlapping sequences. The three fragments were indi-

vidually generated by standard PCR techniques with the indicated primers. Thegenerated plasmid integrating H1RT4L followed by TEV cleavage sequence(ENLYFQG), yeast enhanced GFP and octa-histidine tag (H1RT4LGFP) wasisolated from S. cerevisiae. Coding regions of the H1RT4LGFP fusions wereamplified by PCR using a forward primer containing a BamHI site (59-CTAGAACTAGTGGATCCACCATG-39) and a reverse primer containing anEcoRI site (59-GCTTGATATCGAATTCCTGCAGTTAATG-39). The PCR pro-ducts were digested with BamHI and EcoRI, and subcloned into the pPIC9K

vector.Expression and membrane preparation. The PmeI linearized pPIC9K express-ion vector integrating H1R-flGFP or H1RT4LGFP was then transformed intothe P. pastoris SMD1163 strain by electroporation (2,000 V, 25 mF, and 600V)using a Gene PulserI (Bio-Rad).Clone selection was performed on the YPD-agarplate containing 0.1 mg ml21 geneticine. A single colony of P. pastoris transfor-mant was inoculated into BMGY medium (1% (w/v) yeast extract, 2% (w/v)peptone, 1.34% (w/v) yeast nitrogen base without amino acids, 0.00004% (w/v)biotin, 1% (w/v) glycerol, 0.1M phosphate bufferat pH 6.0) at 30 uC with shakingat 250 r.p.m. until an OD600 of 26 was reached. The cells were harvested bycentrifugation. To induce expression, the cell pellet was resuspended to an OD600of 1.0 in BMMY medium (1% (w/v) yeast extract, 2% (w/v) peptone, 1.34% (w/v)yeast nitrogen base without amino acids, 0.00004% (w/v) biotin, 0.5% (v/v)methanol, 0.1 M phosphate buffer at pH 6.0) containing 2.5% (v/v) DMSO at30 uC. Cells were harvested within 20 to 24 h after induction, and stored at280 uC. Yeast cells were disruptedwith 0.5mm glass beads in a buffercontaining50 mM HEPES, pH 7.5, 120 mM NaCl, 5%(v/v) glycerol, 2 mM EDTA andEDTA-free protein inhibitor cocktail (Roche). Undisrupted cells and cell debriswere separated by centrifugation at 3,000g, and yeast membranes were collectedby ultracentrifugation at 100,000gfor 30min at 4 uC. Washing of the membraneswas performed by repeating dounce homogenation and centrifugation in a high

salt buffer containing 10 mM HEPES, pH 7.5, 1 M NaCl, 10 mM MgCl2, 20mMKCl and EDTA-free protease inhibitor cocktail. Prepared membranes were resus-pended in a buffer containing 50 mM HEPES pH 7.5, 120 mM NaCl, 20% (v/v)glycerol and EDTA-free protease inhibitor cocktail, and snap-frozen in liquidnitrogen and stored at 280 uC until use. Membrane proteins were quantifiedusing the bicinchoninic acid method (Pierce).Purification of H1RT4L. Membrane suspension containing H1RT4LGFPwas thawed and incubated on ice for 30 min in the presence of 5 mM doxepin,10mgml21 iodoacetamide, and EDTA-free protease inhibitor cocktail (Roche).Themembrane suspension was pouredinto the buffercontaining 20 mM HEPESpH 7.5, 500 mM NaCl, 1% (w/v) n-dodecyl-B-D-maltopyranoside (DDM,Anatrace), 0.2% (w/v) cholesteryl hemisuccinate (CHS, Sigma), 20% (v/v) gly-cerol and 23mg ml21 membrane, and stirred gently at 4 uC for 12h. Theunsolubilized material was separated by centrifugation at 100,000g for 30 min.The supernatant was incubated with TALON IMAC resin (Clontech) overnight.The resin was washed with twenty column volumes of 20 mM HEPES pH 7.5,

500 mM NaCl, 10% glycerol, 0.025% (w/v) DDM, 0.005% (w/v) CHS, 100mMdoxepin and20 mM imidazole. Theprotein was eluted with 4 column volumes of20 mM HEPES pH 7.5, 500mM NaCl, 10% glycerol, 0.025% (w/v) DDM, 0.005%(w/v) CHS, 500mM doxepin and 200 mM imidazole. The eluted fractions wereconcentrated to 2.5 ml with a 100 kDa molecular weight cut-off AmiconUltra(Millipore). Imidazole was removed using PD-10 column (GE healthcare).Then the protein was loaded onto the Ni-Sepharose high performance resin(GE healthcare) (1.5 ml resin for,10 mg of protein). The resin was washed with20 column volumes of 20 mM HEPES pH 7.5, 500mM NaCl, 10% glycerol,0.025% (w/v) DDM, 0.005% (w/v) CHS, 500 mM doxepin and 20 mM imidazole.The sample was elutedwith3 column volumes of20 mMHEPES pH7.5,500mMNaCl, 10% glycerol, 0.025% (w/v) DDM, 0.005% (w/v) CHS, 1 mM doxepin and500 mM imidazole. Imidazole was removed using PD-10 column (GE health-care). The protein was processed overnight with His-tagged TEV protease(expressed and purified in house). TEV protease and the cleaved His-taggedGFP were removed by passing the sample through the Ni-Sepharose high per-

formance resin.The receptor was concentrated to 3040mg ml21 with a 100kDa

molecular weightcut-off Vivaspin concentrator(Vivascience). Protein purity andmonodispersity were tested by SDSPAGE and by size-exclusion chromato-graphy using Superdex 200 (GE healthcare).

Lipidic cubicphase crystallization. Lipidiccubic phase(LCP)crystallization trialswere performed using an in meso crystallization robot as previously described48.Ninety-six-well glass sandwich plates were filled with 4050 nl protein-laden LCPboluses overlaidby 0.8ml ofprecipitant solution ineach well and sealed with a glasscoverslip. The protein-LCP mixture contained 40% (w/w) receptor solution, 54%(w/w) monoolein, and 6% (w/w) cholesterol. Crystallization set-ups were per-

formed at room temperature (2022uC). Plates were incubated and imaged at

20 uC using an automated incubator/imager (RockImager 1000, Formulatrix).Crystals were obtained in 2630% (v/v) PEG400, 300 mM ammonium phosphate,10mM MgCl2, 100mM Na-citrate pH 4.5 and 1 mM doxepin (Sigma)(Supplementary Fig. 4). Crystals were harvested directly from LCP matrix usingMiTeGenmicromounts andwere flash-frozenin liquidnitrogen withoutadditionalcryoprotectant.

Data collection and refinement. X-ray diffraction data were collected at 100 Kwith a wavelength of 0.97780 A and with a beamsize of 103 10mm on the micro-focus beamline I24 at the Diamond Light Source (UK) with a Pilatus 6M detector.Each loop was subjected to a grid scanning49 in order to locate the crystals, whichare invisiblein the LCPonce theyare mounted.The exactlocations anddimensionsof the chosen crystals were determined by further grid scanning with a smallersearch area. Data collection was carried out by collecting several overlappingwedges of data from adjacent positions within a single crystal. The data wereprocessedinitially withxia250 usingMosflm51andScala52withthe mergingstatistics

used to determine an optimumsubsetof measurementsto merge.The final data setconsisted of data from five of the eight positions recorded, giving a total of 75degrees of data. These data were then remerged with Scala to give the final dataset summarized in Supplementary Table 2. The space group was determined to beI422 with one molecule in the asymmetric unit. Diffraction data were slightlyanisotropic, extending to 2.9 A in the c* direction and to 3.1 A in the a* and b*directions. The structure factors up to 3.1 A resolution were anisotropically scaledby PHASER53 and thenused for the subsequent molecular replacement and refine-ment. The structure was determined by molecular replacement with the programPHASER53 using two independent search models (polyalanine of the 7 trans-membrane a-helices, and T4L) from the b2-AR (PDB code 2RH1) structure. Wechose b2-AR as a model structure because it has the highest homology of trans-membrane helices with H1R (41.7%) among the human GPCR structures. For theinitial map calculation after molecular replacement, however, we used a b2-ARmodel withoutside chains,loops, ligand,lipids andany solvents; therefore, thefinal

H1

R structure is not biased to theb2

-AR structure. This is supported by lowRwork

and Rfree values (Supplementary Table 2). All refinements were performed withREFMAC554 and autoBUSTER55 followed by manual examination and rebuildingof the refined coordinates in the program Coot56. The non-lysozyme portion con-tainshigherB factors (116A2) owingto fewercontacts as comparedto T4lysozyme(36A2). Calculation of the surface area buried by crystal contacts also explainsthis.For the non-lysozyme portion, only 8% (1,225 A2) of 15,689 A2 solvent-accessiblesurface area is buried by crystal contacts. In contrast, for the T4 lysozyme portion,32% (2,733 A2) of the solvent accessible area (8,648A2) is buried by crystal inter-actions.Supplementary Fig. 5 alsoshows that there are strong interactions betweenT4 lysozyme domains, but relatively fewer between non-lysozyme domainsthroughout the crystal packing. Althoughthe average B factor of the non-lysozymedomain is high as compared to T4 lysozyme, electron densities were clear forunambiguous model building (Supplementary Figs 3 and 5). The eight H1RN-terminal residues (Thr20Leu 27), two C-terminal residues (Arg 486Ser487),and seven residues (Phe168Val 174) in the second extracellular loop (ECL2) are

not included in the structure, as they did not have interpretable densities.Strong and spherical electron densities (about 4 sigma) were found in the

anion-binding region in the Fo2Fc omit map. We excluded the presence of awater molecule in this region owing to strong residual positive Fo2 Fc densitieswhen we modelled it as a water molecule. The coordination geometry in thehighly electropositive environment surrounded by His 4507.35, Lys179ECL2 andLys1915.39 implied that either a phosphate or sulphate ion could be modelled.Becauseammoniumphosphatewas added to ourcrystallization buffer, we modelledit as a phosphate ion. The average B factorsof the phosphate ionand the interactingatoms are 177 A2 and 154 A2, respectively.

Ligand-binding assays. For the saturation binding experiment, yeast membranesuspensions containing H1R-flGFP (20mg) or H1RT4LGFP (5mg) were incu-batedwith increasingconcentrations of [3H]pyrilamine (from 0.15to 40nM)in atotal assay volumeof 200ml for1 h at 25uC. To investigatethe effectof phosphateon the ligand binding, assays were performed in PBS buffer pH 7.4 (138 mMNaCl, 8.1mM Na2HPO4, 27 mM KCl, 1.8mM KH2PO4) or in the HEPES buffer

containing 20 mM HEPES pH 7.5 and 150 mM NaCl. Nonspecific binding was

ARTICLE RESEARCH

Macmillan Publishers Limited. All rights reserved2011

7/30/2019 nature10236.pdf

8/8

determined in the presence of 1,000-times excess unlabelled pyrilamine.Membranes were trapped on Whatman GF/B filters pre-soaked in 0.3% poly-ethylenimine, and unbound radioligands were washed with 9 ml of the PBS orHEPES buffers. The retained radioactivity was measured on an LCS-5100 liquidscintillation counter (ALOKA) in a Clearzol I scintillation liquid (Nakarai,Japan). Data were analysed by nonlinear curve-fitting with a rectangular hyper-bola function using the Prism 4.0 software (GraphPad) to determine dissociationconstant (Kd).

For competitionbinding assays, yeast membrane suspensions containing H1R-

flGFPor H1RT4LGFP wereincubatedwith 4 nMor 20nM [3

H]pyrilamine inthePBS bufferor theHEPES bufferin thepresence of 10nM to 100mM histaminehydrochloride or 0.001 nM to 1mM doxepin, or 0.01nM to 10mM cetirizine,pyrilamine, olopatadine and fexofenadine. Data were analysed by nonlinear curvefitting with a sigmoidalfunction usingthe Prism 4.0to determinethe half maximalinhibitory concentrations (IC50). All data shown were calculated based on morethan three independent experiments. Inhibition constant Ki was calculatedbased on the equation Ki5 IC50/(11L/Kd), where L is the concentration of[3H]pyrilamine with the dissociation constant Kd.Thermal stability assay. N-[4-(7-diethylamino-4-methyl-3-coumarinyl)phenyl]maleimide (CPM) dye was purchased from Invitrogen and dissolved inDMSO (Sigma) at 4 mg ml21 as the stock solution for future use. The stocksolution was kept at 280 uC and was diluted 1:40 in dye dilution solution(10 mM buffer, 500 mM NaCl, 10% glycerol, 0.025% DDM and 0.005% CHS)before use. The thermal denaturation assay was performed with a total volume of200ml sample in a quartz fluorometer cuvette (Starna Cells). H1R (4mg) was

diluted in the appropriate buffer solution to a final volume of 200 ml. Five micro-litresof the diluted dyewas added to the protein solution andit wasincubatedfor30 min at 4 uC. The mixed solution was transferred to the cuvette and the datawere collected by a Cary Eclipse spectrofluorometer (Varian) with a temperatureramping rate at 1 uCmin21. The excitation wavelength was 387nm and theemission wavelength was 463 nm. All assays were performed over a temperaturerange starting from 20 uC to 80 uC. The stability data were processed withGraphPad Prism program (Graphpad Sofware). To determine the melting tem-perature (Tm), a Bolzmann sigmoidal equation was used to fit to the data.Flexible ligand-receptor docking. Docking of ligands was performed using theall-atom flexible receptor docking algorithm in the ICM-Pro molecular modellingpackage57 as described previously39,58. The initial H1R model was generated inICMby building hydrogen atoms forthe crystalstructure of H1R. Internal coord-inate (torsion) movements were allowed in the side chains of the binding pocket,defined as residues within 8 A distance of doxepin in the H1Rdoxepin complex.Other side chains and the backbone of the protein were kept as in the crystal

structure. An initial conformation for each of the ligands was generated byCartesian optimization of the ligand model in MMFF force field. Docking wasperformedby placing theligand ina randomposition within5 A from the binding

pocketand globaloptimization of the complex conformational energy. The globalenergy of the complex was calculated as a sum of van der Waals, electrostatic,hydrogen-bonding and torsion stress terms. Stochastic global energy optimiza-tion of the complex was performed using the ICM Monte Carlo procedure withminimization59. To facilitate side-chain rotamer switches in flexible H1R models,

the first 106 steps of the Monte Carlo procedure used soft van der Waals poten-tials and high Monte Carlo temperature, followed by another 10 6 steps withexact van der Waals method and gradually decreasing temperature. A harmonicdistance restraint has been applied between the amino group of the ligand and

thecarboxyl ofthe Asp107 side chainin theinitial 106

stepsto facilitate formationof the known salt-bridge interaction between these two groups. At least tenindependent runs of the docking procedure were performed for each H1R ligand.The docking results were considered consistent when at least 80% of the indi-

vidual runs resulted in conformations clustered within a root mean squareddeviation of,0.5A to the overall best energy pose of the ligand.

47. Newstead, S., Kim, H., von Heijne, G., Iwata, S. & Drew, D. High-throughputfluorescent-based optimization of eukaryotic membrane protein overexpressionand purification in Saccharomyces cerevisiae. Proc. Natl Acad. Sci. USA 104,1393613941 (2007).

48. Cherezov, V.,Peddi,A., Muthusubramaniam,L., Zheng, Y. F. & Caffrey, M. A roboticsystem for crystallizing membrane and soluble proteins in lipidic mesophases.Acta Crystallogr. D 60, 17951807 (2004).

49. Aishima, J. et al. High-speed crystal detection and characterization using a fast-readout detector. Acta Crystallogr. D 66, 10321035 (2010).

50. Winter, G.xia2: an expert system for macromolecular crystallography data

reduction. J. Appl. Cryst. 43, 186190 (2010).51. Leslie, A. G. W. Recent changes to the MOSFLM package for processing film andimageplate data. Joint CCP4 ESF-EACMB Newslet. Protein Crystallogr. No. 26(1992).

52. Evans, P. Scaling and assessment of data quality.Acta Crystallogr. D 62, 7282(2006).

53. McCoy, A. J. et al. Phaser crystallographic software. J. Appl. Cryst. 40, 658674(2007).

54. Skubak, P., Murshudov, G. N. & Pannu, N. S. Direct incorporation of experimentalphaseinformationin modelrefinement.ActaCrystallogr.D 60, 21962201(2004).

55. Murshudov, G. N., Vagin, A. A. & Dodson, E. J. Refinement of macromolecularstructures by the maximum-likelihood method. Acta Crystallogr. D 53, 240255(1997).

56. Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and Development ofCoot.Acta Crystallogr. D 66, 486501 (2010).

57. Abagyan,R., Orry, A.,Raush,E. & Totrov, M. ICMManual v.3.0 (MolSoftLLC, 2011).58. Totrov,M. & Abagyan,R. Derivation of sensitive discrimination potential for virtual

ligand screening. Proceedings of the third annual international conference on

computational molecular biology312317 (1999).59. Abagyan, R. & Totrov,M. Biased probability Monte Carlo conformationalsearchesand electrostatic calculations for peptides and proteins.J. Mol. Biol. 235,9831002 (1994).

RESEARCH ARTICLE