Studying Haloanisoles Interaction with Olfactory Receptors - ACS

Subscriber access provided by TULANE UNIVERSITY

Article

Studying haloanisoles interaction with olfactory receptors Carla Sílvia Sívia Silva Teixeira, Antonio Cesar Silva Ferreira, and Nuno M. F. Sousa A Cerqueira ACS Chem. Neurosci., Just Accepted Manuscript • DOI: 10.1021/acschemneuro.5b00335 • Publication Date (Web): 19 Apr 2016 Downloaded from http://pubs.acs.org on April 24, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Chemical Neuroscience is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Neuroscience

Studying haloanisoles interaction with olfactory receptors Carla S. Silva Teixeiraa, António C. Silva Ferreirab,c , Nuno M. F. S. A. Cerqueira*a

a

UCIBIO@Requimte/Departamento de Química e Bioquímica, Faculdade de Ciências,

Universidade do Porto, Rua do Campo Alegre s/n, 4169-007 Porto, Portugal b

Escola Superior de Biotecnologia, Universidade Católica Portuguesa, Rua Arquiteto Lobão

Vital, Apartado 2511, 4202-401 Porto, Portugal c

Institute for Wine Biotechnology, Department of Viticulture and Oenology, University of

Stellenbosch, Private Bag XI, Matieland 7602, South Africa

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 34

Abstract In this paper, computational means were used to explain and predict the interaction of several odorant molecules, including three haloanisoles, 2,4,6-trichloroanisole (TCA), 2,4,6-tribromoanisole (TBA) and 2,4,6-Trichlorophenol (TCP), with three olfactory receptors (ORs): OR1A1, OR1A2, and OR3A1. As the X-ray structure of these ORs is not known, the three-dimensional structure of each OR was modeled by homology modeling. The structures of these ORs were stabilized by molecular dynamic simulations and the complexes of the odorant molecules with each ORs were generated by molecular docking. The theoretical results have shown that each OR has distinct but well-defined binding regions for each type of odorant molecules (aldehydes and alcohols). In OR3A1, the aldehydes bind in the bottom region of the binding pocket nearby Ser257 and Thr249. In the paralogs OR1A1 and OR1A2, the aldehydes tend to interact in the top region of the binding pocket and close to a positively charged lysine. On the other hand, the alcohols interact in the bottom region of the active site and close to a negatively charged aspartate. These results indicate that when aldehydes and alcohols odorants compete in these two ORs, the aldehydes can block the access of the alcohols odorants to their specific binding site. This observation goes in line with the experimental data that reveals that when the odorant is an aldehyde, a lower quantity of ligand is needed to cause 50% of the maximum response (lower EC50), when compared with the alcohols. The theoretical results have also allowed to explain the differences in the activity of (S)-(-)-citronellol in the wild-type and mutated OR1A1. The theoretical results show that Asn109 has a preponderant role in this matter, since when it is mutated, it leads to a conformational rearrangement of the binding pocket that prevents the interaction of (S)-(-)-citronellol with Asp111 that was shown to be important for the OR activation. The good agreement between the theoretical and experimental results also lead us to study the potential interaction of the haloanisoles, TCA, TBA and TCP with these ORs. The results have shown that these compounds can compete with other known agonists/antagonists for the access to the binding regions of ORs. These results may partially explain the capability of these compounds to give a musty odor to food and beverages at very low concentrations.

Keywords: haloanisoles, homology, olfactory receptors, molecular dynamics, molecular docking.


Page 3 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction Haloanisoles are small, intriguing molecules that are generally considered off-flavor substances with the capacity to give a moldy or musty aroma to food and beverages. These compounds are characterized by low molecular weight and have at least one halogen atom (fluorine, chlorine, bromine or iodine) as part of their chemical composition (Table 1). To date they have already been detected and associated with unwanted off-flavors in drinking water 1, beer

2

, Japanese sake

3

coffee 4, wine 5, eggs and broiler meat 6.

In the last two decades, the contamination of food or beverages by haloanisoles has been the focus of several studies that continue to be highly motivated by the economic implications of the problem, especially in the food industry. A key aspect of this issue is that one of the most popular haloanisoles, the 2,4,6-trichloroanisole (TCA), belongs to a small group of substances which humans are capable of detect at extremely low concentration levels along with trans-2-nonenal, 3mercapto-2-pentanone, furan-2-methanethiol, 1-p-menthene-8-thiol, 2-isobutyl-3-methoxypyrazine or methylthiol among others 7. Those potent odorants play a critical role in the perceived quality of the food system on which they are found. Therefore, understanding the respective mechanisms of odor detection namely the interaction with olfactory receptors as agonist or antagonist may contribute to explain the reason of such low sensory thresholds. One of the most studied sources of haloanisole contamination is the natural cork that confers an unwanted musty aroma to the wine. The problem has been deeply studied in the last 40 decades, and several authors identified the haloanisoles, 2,4,6-trichloroanisole (TCA), 2,4,6-tribromoanisole (TBA) and 2,4,6-Trichlorophenol (TCP), as the main cause for cork taint. TCA exists in corks by the action of bacteria and yeasts 8, as well as by the use of hypochlorite in the bleaching process 9. Due to substantial improvements in the processing of natural cork, such as the avoidance of hypochlorite as a bleaching agent, and rigorous quality management on microbial contamination, today the typical TCA-based corky off-flavor is of reduced importance in the cork industry 10. TBA is another compound identified in wines with a significant musty character. It is produced by orthomethylation of its direct precursor 2,4,6-tribromophenol (TCP) that generally comes from sources in the winery environment. TCP is originated from naturally-occurring phenol and chlorine from sanitizers and cleaning products, and town water 9. It is reported that TCP is not the main contributor to cork taint; however, it is one of the main precursors of TCA and therefore has a major role in producing cork taint in wine.

Table 1- Molecular structure of the haloanisoles TCA, TBA and TCP.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Cl

OCH3 Cl

OCH 3 Br

Br

Page 4 of 34

OH Cl

Cl

Cl

Br

Cl

TCA

TBA

TCP

Although the haloanisoles that are responsible for the unwanted flavor in food and beverages are well identified and characterized for a long time, the biological process by which they produce the unwanted off-flavors even at low concentrations is poorly understood. It is known that the sensory threshold of an odorant molecule can vary according to the medium in which it is present and/or on the individual that conducts the sensory evaluation. In the case of the haloanisoles under study, it was reported that TCA and TBA, which have a sensory threshold in water ranging from 0.008 to 0.03 ng/L, have a threshold in wine ranging from 2 to 6 ng/L. It was also shown that in wine, and depending on the sensitivity of the individual, the presence of TCA or TBA will cause a loss of fruit intensity and aroma at lower concentrations (1 to 3 ng/L) and an actual taint perception at higher concentrations (more than 3 ng/L) 11.

In mammals, ORs are G protein-coupled receptors (GPCRs), that are characterized by seven hydrophobic membrane–spanning domains. classified as class A rhodopsin like

13

12

. Due to their domain organization, the ORs are

and have an average length of about 320 ± 25 amino acid

residues. The signal cascade induced by ORs after G protein activation leads to adenylate cyclase activation and cAMP production. cAMP then binds and opens Cyclic nucleotide-gated (CNG) ion channel. This opening allows for an influx of both Na+ and Ca2+ ions into the cell, thus depolarizing it. The Ca2+, in turn, activates chloride channels, causing efflux of Cl−, which results in a further depolarization of the cell 14.

As any odorant, it would be expected that haloanisoles present in food or drink should activate or block some specific ORs. In theory, this interaction would initiate a neuronal response and generate an olfactory code triggering the perception of a smell that is different from the one that is characteristic from the uncontaminated food or drink. But the facts show that the TCA molecule has some peculiar feature that results in an unexpected psychophysical phenomenon. While some studies report its remarkable capability to evoke musty odors at concentrations much lower than the smallest effective concentration of odorants known to induce responses in ORs (~1 µM)

15, 16

,

other studies performed with amphibian receptors (Cynops pyrrhogaster) shows that TCA, even at attomolar concentrations, attenuates olfactory transduction by suppressing cyclic nucleotide-gated channels, without evoking odorant responses15. This phenomenon, known as olfactory masking, was previously studied by the same group who found a positive correlation between it and the


Page 5 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


odorant suppression of the transduction current through the cyclic nucleotide–gated (CNG) and Ca2+-activated Cl− (Cl(Ca)) channels 17. To date, it is not available any study concerning the interaction, either excitatory or inhibitory, between TCA, TBA and TCP haloanisole molecules and the human ORs. In line with this facts, the aim of this paper is to use a computational protocol to explain and predict a possible interaction between TCA, TBA and TCP and three dehorphanized human ORs: OR1A1, OR1A2, and OR3A1. To date, no three-dimensional structure of the OR1A1, OR1A2, and OR3A1 is available and therefore, no information is available to understand how haloanisoles, such as TCA, TBA, and TCP interact with these ORs. Taking this into account, in this article, we propose, to model the threedimensional structure of the three ORs and study their potential interaction with TCA, TBA, and TCP. The three ORs (OR1A1, OR1A2, and OR3A1) were chosen based on the fact that they all have known agonists and/or antagonists with experimentally determined EC/IC50. Additionally, OR1A1 and OR1A2 are known to be activated by odorants present in foods or beverages with record of contaminations by halionalisoles. For example, the terpenes geraniol, citronellol, and citronellal, agonists of OR1A1 and OR1A2, are odorant molecules present in beer Linear aldehydes, nonanal, and decenal are found in coffee beans

18

and wine

19

.

20

. The OR3A1 was chosen

because it has a common agonist with OR1A1 and OR1A2 (helional) but is narrowly tuned to aldehydes with specific structural characteristics. The GPCRs unit of each receptor was built by homology modeling and the overall structure of the proteins emerged on lipid bilayer and a water box that was subsequently equilibrated in a molecular dynamic simulation. In addition, one mutated OR, mOR1A1, was also modeled and studied in order to understand the different activities that the wild-type and mutated OR exhibit against (S)-(-)-citronellol. The interaction of the agonists and antagonists, including the TCA, TBA and TCP with each OR was then studied by molecular docking. The theoretical results obtained in this work were then compared with the available data present in the literature, namely functional and theoretical studies

21-24

. This will allow to explain these results and at the same time validate

the computational protocol and the modeled ORs.

Results and Discussion In this work, it is our aim to study the interaction of the haloanisoles TCA, TBA, and TCP with the ORs using computational means. As the three-dimensional structure of any OR is not known, we proposed to build by homology modeling three olfactory receptors that are denoted in this article by OR1A1, OR1A2, and OR3A1. We have chosen these three olfactory OR because there are



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 34

several experimental data regarding the activity of several compounds with these receptors in the literature. This information will be used to validate the created olfactory receptors models and see if they can be used to explain the available experimental data. In addition, a mutated OR1A1 model, denoted in this article as mOR1A, was built in order to evaluate the differences that are observed on the activity of certain ligands between the wild-type and mutated olfactory receptors. The wildtype models were also used to study the interaction of the haloanisoles molecules TCA, TBA, and TCP. Three computational methodologies are used in this paper: homology modeling, molecular dynamic simulations, and molecular docking. In order to explain more clearly what was done in each one of them and discuss the results that were obtained, the discussion is divided into three main stages. In the first stage, the computational protocol used to build the three-dimensional structure of the wild-type OR1A1, OR1A2, and OR3A1 and the mutant m1OR1A1 will be discussed. In a second stage, the interaction between agonists and antagonist with each OR was modeled and correlated with the available experimental data.

This will allow validating the

computational protocol and the ORs models that were generated. Finally, in the last stage, we studied the interaction of the haloanisoles compounds TCA, TBA, and TCP with the OR and discussed their role in the perception of the undesired off-flavours present in wine. Homology Modeling The three-dimensional structure of the wild-type OR, OR1A1, OR2A1 and OR3A1, and mutant OR1A1 (mOR1A1), were constructed based on homology modeling techniques. The homology modeling method relies on the fact that evolutionarily related proteins share similar amino-acid sequences and structural arrangements. To this end, it is used a template that must share high sequence identity to these proteins. As currently there is no X-ray structures of any OR, the crystal structures of the rhodopsin-like GPCRs are generally a valuable option to serve as a template. From the six possible GPCRs, the adenosine receptor A2A was the one that presents higher percentage of sequence identity in relation to all the studied ORs (Supporting information Table 1) and so, it was used as a template to model the three-dimensional structure of OR1A1, OR1A2, and OR3A. The sequence identity of this structure with the OR1A1, OR1A2 and OR3A1 are 22.16 %, 21.3 % and 20.71 %, respectively (Table 1 and Table 1 from Supporting information). A total of 10 homology models were generated to OR1A1, OR1A2, and OR3A1, using the modSim server

25

and their overall stereochemical quality was subsequently analyzed in detail. To this end,

statistical analysis of each structure was done including the DOPE, Ramanchandran plots, rotamer evaluation, geometry evaluation and Cβ deviation. The statistical analysis of the best 5 scored solutions are shown in Table 2. A visual inspection of all of these structures reveals that there are no significant structural differences between the top five scored structures for each OR. The bestscored solutions (highlighted in bold in Table 2) were selected for further analysis.


Page 7 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


A similar protocol was used to build the three-dimensional structure of mOR1A1. The main difference between the mutated and wild-type protein is the residue at position 109. In the wildtype OR1A1, this residue is an Asn while in the mutated receptor this residue is a Lys.

Table 2 - Statistical data regarding the five best-scored homology models of OR1A1, OR1A2, OR3A1, m1OR1A1. The models that were selected for further calculations are highlighted in bold.

Sequence identity (%)

OR’s

Wild-Type

Mutated

OR1A1

22.16

OR1A2

21.3

OR3A1

20.71

mOR1A1

22.16

DOPE

-30966 -31007 -31025 -31369 -31384 -29903 -30274 -30554 -30601 -30637 -29488 -29654 -29843 -29999 -30515 -30673 -30716 -30798 -30973 -31809

Ramachandran plot favored Allowed Ramachand regions regions ran outliers (98%) (>99.8% ( 0.1*

> 0.1*

octanol

38.2 ± 13

60.4 ± 15

(S)-(-)-citronellol

92.6 ± 10.2

a

(a) No effect up to 300 µM. (*) Effects starting from 0.03 µM, however a fit-function-derived EC50 was not calculated due to the lack of saturation, and ORindependent effects at concentrations higher than 10–30 µM.

The theoretical results obtained in this work show that the aldehyde and alcohol molecules bind in distinct regions of the binding pocket of OR1A1 and OR1A2. In all the analyzed cases the aldehydes bind nearby positively charged residues while the alcohols tend to bind nearby negatively charged residues. In the OR1A1, the aldehydes molecules bind nearby the entrance of the binding pocket and interact very closely with Lys197. The distances between Lys197 and each ligand are presented in Table 2 of Supporting information. This residue behaves like an anchor to all of the aldehydes derivatives. The remaining part of the structure of the aldehydes has two distinct spatial rearrangements. One in the direction of Thr245 and the other in the direction of Ser112. This behavior is endorsed by the presence of Phe206 that allows the formation of a small cavity nearby Thr245. The aldehydes with the best activities are the ones that present both binding positions, and the best-scored solutions are the ones that point towards the cavity where Thr245 is located. The ones that only have one binding position are the ones that point towards Ser112 and are experimentally inactive. The theoretical results suggest that this happens because these compounds have not enough length to achieved the cavity where Thr245 is located at, as it is the


Page 19 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


case of hexanal, or the compound is not flexible enough, as it is the case of 2,4-decadienal. This indicates that the interaction of Thr245 with the odorant molecules may be an important feature for the activation of the OR1A1. Some aldehydes also show two binding positions. The compounds with better activity are the ones that have functional groups in the tail of their structure or/and have bulkier groups in their structures. Any of these conditions allows a better interaction of the compounds with the cavity where Thr245 is located. The first condition allows the aldehydes to interact directly with Thr245 by hydrogen bonds, as it is the case of helional. The presence of bulkier groups in their structure (methyl groups) allows them to become better accommodate inside the same cavity and almost fulfill it (as it is the case of (S)-(-)-citronellal and hydroxy-citronellal). The compounds with an alcohol moiety bind in the deepest region of the OR1A1 binding pocket. In all of the analyzed complexes, the hydroxyl group of these compounds is stabilized by a short hydrogen bond with the negatively charged Asp111, which behaves like an anchor to all of them. The distances between Asp111 residue and each ligand are presented in Table 2 of Supporting information. Similarly, to what is observed with the aldehydes, there are several possible spatial arrangements of the ligand tails in the binding pocket. One in the direction of Phe237 and two in the direction of Asn109. The binding pose of some representative agonists of OR1A1 is represented in Figure 6.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 34

Figure 6: Binding pose of some agonists in the OR1A1 binding pocket. The distances (Å) between the functional groups of aldehyde/alcohol odorant molecules and the hydrogen bond donor or acceptor from the nearest residue from the binding site of the OR1A1 are described in Table 2 from Supporting information. The theoretical results do not provide by themselves any significant explanation that could explain the differences that are observed in the activity of the alcohols that were measured experimentally. From Table 4 it is clear that the presence of bulkier groups in the alcohol scaffolds increases their activity, except when chiral groups are present as it is the case of (S)-(-)-citronellol. The theoretical results have also shown that the compounds with chiral centers and bulkier atoms may hinder a


Page 21 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


favorable binding of the compounds in the bottom region of the active site due to the present of Phe206 that turns the tunnel nearby Glu111 very narrow.

In the OR1A2, the aldehydes and the alcohol molecules also bind in different locations of the binding pocket, as it was observed with OR1A1, and so similar conclusions can be drawn. The aldehydes bind in the top region of the active site and all of them are anchored around Lys186 by hydrogen bonds. The distances between Lys186 and each ligand are presented in Table 2 of Supporting information. Two different rearrangements are then observed for the majority of the compounds. The aldehydes that contain eight carbons in their structure become packed in the cavity where His85 is located at (cavity 1), or aligned along the binding pocket and closer to the wide cavity generated by Phe206 (cavity 2). The aldehydes with larger scaffolds cannot bind in cavity 1 due to size limitation and all of them point towards the wider cavity 2. The compounds with these characteristics are the less active, suggesting that the interaction with cavity 1 is required for the activation of the OR1A2. Similar to what was observed in the OR1A1, compounds composed by 8 carbons and with functional groups in their tails have improved activities. This is due to the fact that they can interact with cavity 1 and establish hydrogen bonds with His85. In OR1A2, hexanal is an inactive compound similarly to what is found in OR1A1. This happens due to the small size of the molecule that precludes any interaction with any of the cavities of the binding pocket, in particular with cavity 1 which may be required for the activation of the OR. The compounds with an alcohol moiety bind in the deepest region of the OR1A2 binding pocket. In all of the analyzed complexes, the hydroxyl group of these compounds is stabilized by a short hydrogen bond with the negatively charged Asp111 that behaves like an anchor to all of them, as it was found in OR1A1. The distances between Asp111 and each ligand are presented in Table 2 of Supporting information. In this ORs two different binding positions are observed. One is pointing towards the same cavity that is occupied by the larger aldehydes (where Phe206 is located at) and another in the opposite direction (where Phe73 is located at). The binding poses of some representative agonists of OR1A2 are represented in Figure 7.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 34

Figure 7: Binding pose of some agonists in the OR1A2 binding pocket. The distances (Å) between

the functional groups of aldehyde/alcohol odorant molecules and the hydrogen bond donor or


Page 23 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


aceptor from the nearest residue from the binding site of the OR1A2 are described in Table 2 from Supporting information.

Similar to what was observed in the OR1A1, the theoretical results did not provide any significant changes in the binding poses of the alcohols. From Table 4 we can infer that alcohols with chiral centers and bulkier atoms in their structure are inactive in OR1A2. Interestingly the same compound, (S)-(-)-citronellol, in OR1A1 is active. The theoretical results have shown that this happens because the access of the alcohol molecules to the bottom region of the binding pocket of the OR1A2 is more restricted than in OR1A1. This occurs due to the presence of Tyr250 and Phe237 that narrows that region of the binding pocket. Consequently, the alcohols that have bulkier groups in their structure have more difficulty in interacting with Asp111 and therefore become less active. This fact is observed with (S)-(-)-citronellol. Geraniol has also bulkier groups in its structure but does not have chiral centers, which turns its overall volume smaller than chiral compound (S)- (-)-citronellol. A few years ago Schmiedeberg and colleagues performed rhodopsin-based homology modeling in an attempt to define the docking sites for the agonists (S)-(-)-citronellal, and (S)-(-)-citronellol within both OR1A1 and OR1A2

23

. Although some of the residues that were identified in this work are in

agreement with their data (residue Phe73, Asn/Lys109, Ser112, and Phe206), some residues that we found to be directly involved in the aldehyde and alcohol moiety interaction are not on their list. These differences in the binding pocket architecture and ligand accommodation are probably due to the template used for the homology modeling. The sequence identity of OR1A1, and OR1A2, to bovine rhodopsin (the template used in their study), is lower (14.1%, and 16.9%, respectively) than the template that was used in this study from Human A2A Adenosine receptor (22.16% and 21.3%, respectively).

C. mutated OR1A1 In 2007, Schmiedeberg and colleagues found that the odorant (S)-(-)-citronellol is ineffective in OR1A2 but effective in OR1A1

23

. In an attempt to find an answer for this difference in odorants

activity they mutated Asn109 on OR1A1 by a Lys. They observed that (S)-(-)-citronellol does not activate the mutant OR1A1 (mOR1A1) as it did in the wild-type OR. This indicates that this residue is, therefore, important for the activity of OR1A1. In an attempt to understand the effect of this mutation and in the activity of OR1A1, we have also modeled this mutated OR.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 34

The theoretical results have shown that in the mOR1A1, Lys109 establish a salt bridge with Asp111 (2.72 Å) and endorses a new configuration of the binding pocket in that region that blocks the access of S-(-) citronellol to Asp111. In the wild-type receptor this residue behaves like an anchor to all the alcohol molecules, and therefore the absence of this interaction can be the source of the inactivity of (S)-( -)-citronellol in mOR1A1. Interestingly the same compound is also inactive in OR1A2, which also has a Lysine (Lys109) and an aspartate (Asp111) interacting by a salt bridge (2.49 Å).

D. Wild-type ORs with TCA, TBA and TCP The theoretical results have shown that TCA, TBA, and TCP can interact with the three ORs as they bind in the same region of the binding pocket where the agonists and antagonist studied in the previous sections bind. In the OR3A1, TCA, TBA, and TCP interact in the same region of the binding pocket where the agonists and antagonists listed on Table 3 interact. The theoretical results showed a single binding position for TCA, TBA and TCP. All of these compounds establish short hydrogen bonds with Ser257 and Thr249 and are precisely aligned with Phe254 with which they interact by πinteractions. All of these interactions were shown to be important for the activity of the agonists and antagonists of OR3A1, which indicates that these compounds can be ligands of this receptor.

In OR1A1, TCA has two binding positions. One position is located in the cavity where Lys197 is located (the binding region of aldehydes agonists) and another nearby Asp111 and Ser112 (the binding region of alcohols agonists). In OR1A2, a similar pattern is observed. TCA binds in one of the aldehyde binding sites, i.e. the one close to Lys186, Tyr250, and Phe206 or, close to the alcohols binding position nearby Asp111 and Lys109. No solution was observed in the cavity where His86 and Phe27 are located (second binding pose of the aldehydes in OR1A2). This happens because of the size and volume of TCA that do not fit inside this cavity.

TBA and TCP show only one binding position either in OR1A1 and OR1A2. In OR1A1, TBA binds in the region where the aldehydes bind, whereas TCP bind in the bottom region of the binding pocket where the alcohols interact. The preference of TCP for the bottom region of the active site is related to the presence of the hydroxyl group in its structure that makes it behave like an alcohol, and therefore interacts preferentially with negatively charged groups (Asp111). TBA tends to interact close to positively charge amino acids, such as lysine residues (Lys197), similarly to what is observed with TCA.


Page 25 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In OR1A2, TBA and TCP bind into the same region where the alcohols interact, namely with Lys109 and Asp111 through hydrogen bonds. In this case, none of the TBA molecules interact with Lys186. Such behavior is endorsed by the shape and volume of the compound that precludes any interaction in the cavity where His85 and Phe27 are located. The distances between the haloanisoles and the ORs residues are presented at Tables 2 and 3 of supporting information. The binding pose of TCA, TCP, and TBA in the OR1A1, OR1A2 and OR3A1 binding pockets is represented in Figure 8.

Figure 8: Binding pose of TCA, TCP and TBA in the OR1A1, OR1A2 and OR3A1 binding pockets. Experimental studies have shown that the olfactory threshold of the TCP molecule (350 ng/L) is much higher than for TCA (2 ng/L) and TBA (4 ng/L). This means that to be perceived higher concentrations of TCP are required when compared to TCA and TBA. This observation goes in line with functional studies that indicate that alcohols are less efficient in OR activation.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 34

The theoretical results show that this might be related with the binding pose of these compounds. In OR3A1, the binding poses between these compounds are the same but in OR1A1 and OR1A2 they are different. In these cases, TCP bind always in the bottom region of the binding pocket site (alcohols position), whereas TCA binds either in the top or bottom regions of the binding pocket (alcohols and aldehydes position). This indicates that the binding pose of TCP can be impaired if any of the TCA molecules are bonded to the ORs. This happens because the TCA competes with the same binding pocket of TCP (bottom region of the binding pocket) or blocks the access to it when it binds to the top region of the binding pocket. Similar conclusions can be drawn to the TBA molecules.

Concluding remarks In this paper, we modeled the three-dimensional structure of three ORs, OR1A1, OR1A2 and OR3A1, by homology modeling and studied their interaction with several alcohols and aldehydes that possess agonist and/or antagonist activity and for which EC/IC50 were known. To this end, the structures of the modeled ORs were stabilized by a long molecular dynamic simulation and the complexes were generated by molecular docking. The results have shown that in OR3A1, the binding pose of the agonists and antagonist aldehydes are similar between each other and all of them interact by hydrogen bonds with Ser257 and Thr249. The agonist compounds that have bulkier groups attached to their main scaffold allow them to become better accommodate in that region of the binding pocket and turn therefore the interaction with it stronger. The compounds that lack this feature are inactive or antagonists. In the OR1A1 and OR1A2, the theoretical results have shown that the aldehydes and alcohols agonists interact with well-defined regions of the binding pocket and are located in distinct regions. The aldehydes tend to interact with the top region of the binding pocket and in close contact with positively charged amino acids, whereas the alcohols interact in the bottom region of the active site nearby negatively charged amino acids. The presence of bulkier groups tends to improve the activity of the compounds allowing them to occupy more efficiently those regions of the binding pocket and at the same time enhance the direction of the polar groups in the direction of the charged residues. Functional studies reveal that alcohols are 10 times less efficient in OR activation than the aldehydes. The theoretical results indicate that this might be related to the distinct binding positions for each type of the compound in the binding pocket of each OR that are required to activate it. As the aldehyde odorants bind in the top region of the binding pocket, these compounds do not have any physical blockage that impairs their interaction with that region of the OR. In the case of the alcohol odorants, the same is not true. As the binding position of these compounds is located in


Page 27 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the bottom region of the binding pocket, their interaction with the OR can be impaired, if the access to that region is blocked. The theoretical results indicate that this can happen when alcohol and aldehyde odorants interact with the same OR. In this case, as the aldehydes bind in the top region of the binding pocket, they will block the access of any molecule to the bottom region of the binding pocket and therefore impair an effective interaction of the alcohols with the OR. This also goes in line with the experimental observations23 that indicates that a lower quantity of aldehyde is needed to cause 50% of the maximum response (and therefore their EC50 is thus lower ), when compared with the alcohols. These results are also in agreement with a previous study were its was reported that although all agonists appear to bind inside the binding pocket of the ORs, they do so in very different ways according to their functional groups 26.

In an attempt to explain why (S)-(-)-citronellol is active in the wild-type OR1A1 and inactive in mOR1A1 where Asn109 is mutated by a Lys we have also modeled these mutated OR and studied its interaction with (S)-(-)-citronellol. The theoretical results have shown that this mutation leads to conformational rearrangement of the binding pocket that precludes any interaction of (S)-(-)citronellol with ASP111 that was found to be crucial for the activation of the wild-type OR. Our next objective was to try to understand how the haloanisoles TCA, TBA and TCP interact with these receptors. We have chosen to study the interaction of these compounds with OR1A1, OR1A2, and OR3A1 because these receptors interact with odorants that are commonly found in wines and beers such as geraniol, citronellol or citronellal. Therefore, and considering the fact that TCA and TBA are known to influence wine and beer aroma, either giving it an off-flavor or reducing its original aroma, it is expected that TCA, TBA, and TCP might somehow affect the interaction between these naturally occurring aroma compounds and the ORs that they activate or inactivate. The theoretical results have shown that TCA, TBA and TCP bind in the same region that are occupied by alcohols and/or aldehydes odorants. This indicates that they can compete with these compounds for their binding site, hindering their access to the binding pocket and therefore canceling their contribution for the global aroma. This key feature is overwhelmed by their small size, its symmetrical shape, and its capability to establish several hydrogen bonds with the binding pocket of the ORs that potentiates the activity of these haloanisoles. Considering this information and the fact that TCA, TBA, and TCP are associated with the same musty aroma, independently of the beverage or food where they are present, it is possible that these haloanisoles will compete with several aromatic compounds for its binding site and activate a specific ORs that will imprint a fingerprint pattern. This codifies the perception of the musty aroma in the olfactory code that is not present in the uncontaminated food or beverage. This hypothesis does not contradict the fact that some haloanisoles may permeate into the membrane and



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 34

suppress cyclic nucleotide-gated channels leading to the attenuation of olfactory transduction and consequently reducing the original flavor of the contaminated food or drink

27

.This dual effect may

be dependent on the haloanisole concentration. Experimental studies done elsewhere have shown that the olfactory threshold of the TCP molecule (350 ng/L) is much higher than for TCA (2 ng/L) and TBA (4 ng/L)

28

. The results obtained from

this work suggest that it may be related to the capability of TCA to bind indiscriminately to several binding pockets of the receptors (alcohols and aldehydes binding sites). TCP only interact with the bottom region of the binding pocket, so if TCA is already bound to the upper region of the receptor it will not have access to its binding site. This fact together with its lower volatility (when compared to TCA) may explain why TCP has the higher olfactory threshold. TBA has bromine instead of the chlorine atoms that are present in TCA, which means that it is bulkier and less prone to interact by hydrogen bonds with the residues found in the binding pocket of ORs than TCA. These two facts can help to explain why the threshold of TCP is the double of TCA. All the results obtained in this article were based on computational means that were based on different computational techniques, such as homology modeling, molecular docking, and molecular dynamics. The protocol presented here provided consistent results that allowed explaining many experimental data that so far could not be explained at atomic level. The good correlation that was obtained between experimental and theoretical results also allowed validating the computational protocol. This means that such computational protocol can be used with certain confidence in similar applications to predict the olfactory receptor-ligand interaction when there is no crystallographic structure of the receptor nor functional data available.

Methods

Homology Modeling The primary sequences of the three ORs were obtained from the Universal Protein Resource (UNIPROT) database (available in http://www.uniprot. org). (UNIPROT ID: Q9P1Q5, Q9Y585, and P47881 for OR1A1, OR1A2, and OR3A1, respectively). The homology modeling procedures of the four ORs were carried out using the ModSim server (http://gpcr-modsim.org/) 25.

The template used to build all the studied ORs was in an active form. This conformation was chosen because the ligands used to validate this study behave as agonists of the studied receptors and according to the literature when a full agonist binds to a receptor it will stabilize its active state 29

.


Page 29 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Each OR sequence was aligned with six active-like or agonist-bound GPCRs (Supporting information Figures 2 to 5): Bovine Rhodopsin MetaII State, (3pqr); Human Beta2 Adrenergic receptor (3p0g); Turkey Beta1 Adrenergic receptor (Carmoterol), (2y02); Turkey Beta1 Adrenergic receptor (Dobutamine),(2y00);Human A2A Adenosine receptor, (3qak); and Human Beta2 Adrenergic receptor (covalent agonist), (3pds). Adenosine receptor A2A was the GPCR with the higher sequence identity percent for all the studied ORs (Supporting information Table 1). In total ten homology models were generated for each receptor and their overall stereo-chemical quality was evaluated by a set of statistical analysis, namely the discrete optimized protein energy (DOPE), Ramanchandran plots, rotamer evaluation, geometry evaluation and

Cβ deviation

(resulting overall distortion of the Cβ position from ideality) 30. The selection of the best models was based on the score obtained from the statistical analysis and from visual inspection. The mOR1A1 was obtained using the same protocol. Molecular Dynamics. The structures of the four different ORs that were obtained from the previous homology modeling protocol were then submitted to a subset of minimizations and a long molecular dynamic simulation. The minimizations and a Molecular Dynamics (MD) simulations of the four ORs models were performed with the GROMACS 4.6.1 suite, using the OPLSAA-ff force field. All hydrogen atoms were added taking into account all residues in their physiological protonation state. The protein was embedded in a pre-equilibrated POPC (1- palmitoyl-2-oleoyl phosphatidylcholine) membrane model so that the TM bundle is parallel to the vertical axis of the membrane. Several counter-ions were added to the system to neutralize the high positive charge of each protein: 7Cl- for OR1A1, 8Cl- for mOR1A1; 7Cl- for OR1A2; and 10Cl- for OR3A1. The system was then soaked with bulk water (SPC model) and inserted into a hexagonal prism-shaped box that is energy-minimized and carefully equilibrated in the framework of PBC (periodic boundary conditions). The distance between the box faces and the closer protein atoms were kept at a minimum distance of 12 Å. The average size of each system was 79 000 atoms for OR1A1, 70 000 for OR1A2, 95 000 for OR3A1 and 79 000 for mOR1A1. By default, all the simulation modules use a random seed of 1993 and a gen-seed of 173529 at 300K.

All systems were minimized in two stages: first, the proteins were kept fixed and only the position of the water molecules was minimized. Afterward the full system was minimized. Subsequently, a MD simulation of 100 ps at constant volume and temperature, and considering periodic boundaries conditions was run, followed by 40 ns of MD simulation with the NPT ensemble, in which Langevin dynamics was used (collision frequency of 1.0 ps-1) to control the temperature at 310.15 K.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 34

Bound lengths involving hydrogen atoms were constrained using the SHAKE algorithm, and the equations of motion were integrated with a 2 fs time step using the Verlet leapfrog algorithm. Electrostatic interactions were calculated using smooth Particle-Mesh Ewald (PME) method with a cut-off of 12 Å, and the non-bound interactions were truncated with a 12 Å cut-off. Molecular Docking Calculations The binding poses of the compounds inside the ORs, (OR1A1, OR1A2, OR3A1 and mutated OR1A1) were studied using the molecular docking software AutoDock

31, 32

and the vsLab plug-in

33

. The structure of the receptor was taken from the molecular dynamic simulation. The ligands

were built with GaussView, protonated at physiological pH and optimized with gaussian09 (HF/631G(d)). In the docking process, the genetic algorithm (GA) was used. The number of generations, energy evaluations, population size and number of solutions were set to 27 000, 2 500 000, 150 and 50, respectively. The types of atomic charges were taken as Kollman-all-atom for the receptor and Gasteiger for the compounds. The docking of the ligands was performed on a specific area of the protein, that was centered on the coordinates X=49Å, 44Å, 49Å; Y=52Å, 44Å, 52Å; and Z=45Å, 44Å , 55Å and with the dimensions : width=15.0Å, 13.125Å, 15.37Å ; height=17.625, 15.37Å, 17.25Å and depth=28.125Å, 28.5Å, 29.25Å for the receptors OR1A1, OR1A2 and OR3A1 respectively. The points spacing was 0.375 Å for all the three receptors. The final solutions were retrieved from the molecular docking process according to the criteria of the scoring energy. Tunnel shape and SASA calculations The images of the tunnels were made measuring the free volume of the binding site of each modelled OR using the software VolArea

34

. The solvent accessible surface area (SASA)

corresponds to the surface area of a molecule that is accessible to a solvent. In this article, we use it to measure the potential area of the binding site that can interact with a potential substrate. In all calculations, a probe of 1.4 Å was used.

Supporting information

The PDB files of the homology modeled olfactory receptors: OR1A1, OR1A2, OR3A1 and mOR1A1. Additional information regarding the homology model protocol, molecular dynamic simulations and molecular docking results are provided.

Author information


Page 31 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Corresponding author E-mail: [email protected] Author Contributions All the authors contributed equally for this work.

Funding This research was funded by the project IF/01310/2013 and PTDC/AGR-ALI/121062/2010 and partially supported by ESB/UCP plurianual funds through the POS-Conhecimento Program that includes FEDER funds through the program COMPETE (Programa Operacional Factores de Competitividade) by national funds through FCT (Fundação para a Ciência e a Tecnologia).

References [1] Anselme, C., N'guyen, K., Bruchet, A., and Mallevialle, J. (1985) Can polyethylene pipes impart odors in drinking water?, Environmental Technology 6, 477-488. [2] McGarrity, M. J., McRoberts, C., and Fitzpatrick, M. (2003) Identification, cause, and prevention of musty off-flavors in beer, Technical quarterly-Master Brewers Association of the Americas 40, 44-47. [3] Miki, A., Isogai, A., Utsunomiya, H., and Iwata, H. (2005) Identification of 2,4,6-trichloroanisole (TCA) causing a musty/muddy off-flavor in sake and its production in rice koji and moromi mash, J Biosci Bioeng 100, 178-183. [4] Spadone, J. C., Takeoka, G., and Liardon, R. (1990) Analytical investigation of Rio off-flavor in green coffee, Journal of agricultural and food chemistry 38, 226-233. [5] Buser, H., Zanier, C., Tanner, H. (1982) Identification of 2,4,6-Trichloroanisole as a Potent Compound Causing Cork Taint in Wine, Journal of agricultural and food chemistry 30, 359362. [6] Engel, C., de Groot, A. P., and Weurman, C. (1966) Tetrachloroanisol: a source of musty taste in eggs and broilers, Science 154, 270-271. [7] Belitz, H. D., Grosch, W., and Schieberle, P. (2009) Food chemistry, 4th rev. and extended ed., Springer, Berlin. [8] Alvarez-Rodriguez, M. L., Lopez-Ocana, L., Lopez-Coronado, J. M., Rodriguez, E., Martinez, M. J., Larriba, G., and Coque, J. J. (2002) Cork taint of wines: role of the filamentous fungi isolated from cork in the formation of 2,4,6-trichloroanisole by o methylation of 2,4,6trichlorophenol, Applied and environmental microbiology 68, 5860-5869. [9] Simpson, R. F., and Sefton, M. A. (2007) Origin and fate of 2, 4, 6‐trichloroanisole in cork bark and wine corks, Australian journal of grape and wine research 13, 106-116. [10] Slabizki, P., Fischer, C., Legrum, C., and Schmarr, H. G. (2015) Characterization of Atypical Off-Flavor Compounds in Natural Cork Stoppers by Multidimensional Gas Chromatographic Techniques, Journal of agricultural and food chemistry 63, 7840-7848. [11] Tracy, R., Skaalen, B. (2008) 2,4,6-TBA next '2,4,6-TCA' in U.S. wine industry, Pratical Winery & Vineyard Journal. [12] Buck, L., and Axel, R. (1991) A novel multigene family may encode odorant receptors: a molecular basis for odor recognition, Cell 65, 175-187. [13] Jacoby, E., Bouhelal, R., Gerspacher, M., and Seuwen, K. (2006) The 7 TM G-protein-coupled receptor target family, ChemMedChem 1, 761-782.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 34

[14] Silva Teixeira, C. S., Cerqueira, N. M. F. S. A., and Silva Ferreira, A. C. (2016) Unravelling the Olfactory Sense: From the Gene to Odor Perception, Chem Senses 41, 105-121. [15] Bhandawat, V., Reisert, J., and Yau, K.-W. (2005) Elementary response of olfactory receptor neurons to odorants, Science 308, 1931-1934. [16] Menini, A., Picco, C., and Firestein, S. (1995) Quantal-like current fluctuations induced by odorants in olfactory receptor cells, Nature 373, 435-437. [17] Takeuchi, H., Ishida, H., Hikichi, S., and Kurahashi, T. (2009) Mechanism of olfactory masking in the sensory cilia, J Gen Physiol 133, 583-601. [18] Takoi, K., Itoga, Y., Koie, K., Kosugi, T., Shimase, M., Katayama, Y., Nakayama, Y., and Watari, J. (2010) The Contribution of Geraniol Metabolism to the Citrus Flavour of Beer: Synergy of Geraniol and β‐Citronellol Under Coexistence with Excess Linalool, Journal of the Institute of Brewing 116, 251-260. [19] Mateo, J. J., and Jimenez, M. (2000) Monoterpenes in grape juice and wines, Journal of chromatography A 881, 557-567. [20] Mondello, L., Costa, R., Tranchida, P. Q., Dugo, P., Lo Presti, M., Festa, S., Fazio, A., and Dugo, G. (2005) Reliable characterization of coffee bean aroma profiles by automated headspace solid phase microextraction-gas chromatography-mass spectrometry with the support of a dual-filter mass spectra library, J Sep Sci 28, 1101-1109. [21] Wetzel, C. H., Oles, M., Wellerdieck, C., Kuczkowiak, M., Gisselmann, G., and Hatt, H. (1999) Specificity and sensitivity of a human olfactory receptor functionally expressed in human embryonic kidney 293 cells and Xenopus Laevis oocytes, J Neurosci 19, 7426-7433. [22] Jacquier, V., Pick, H., and Vogel, H. (2006) Characterization of an extended receptive ligand repertoire of the human olfactory receptor OR17-40 comprising structurally related compounds, J Neurochem 97, 537-544. [23] Schmiedeberg, K., Shirokova, E., Weber, H.-P., Schilling, B., Meyerhof, W., and Krautwurst, D. (2007) Structural determinants of odorant recognition by the human olfactory receptors OR1A1 and OR1A2, Journal of structural biology 159, 400-412. [24] Anselmi, C., Buonocore, A., Centini, M., Facino, R. M., and Hatt, H. (2011) The human olfactory receptor 17-40: requisites for fitting into the binding pocket, Computational biology and chemistry 35, 159-168. [25] Rodríguez, D., Bello, X., and Gutiérrez‐de‐Terán, H. (2012) Molecular Modelling of G Protein‐Coupled Receptors Through the Web, Molecular Informatics 31, 334-341. [26] Launay, G., Teletchea, S., Wade, F., Pajot-Augy, E., Gibrat, J. F., and Sanz, G. (2012) Automatic modeling of mammalian olfactory receptors and docking of odorants, Protein Eng Des Sel 25, 377-386. [27] Takeuchi, H., Kato, H., and Kurahashi, T. (2013) 2,4,6-trichloroanisole is a potent suppressor of olfactory signal transduction, Proceedings of the National Academy of Sciences of the United States of America 110, 16235-16240. [28] Soleas, G. J., Yan, J., Seaver, T., and Goldberg, D. M. (2002) Method for the gas chromatographic assay with mass selective detection of trichloro compounds in corks and wines applied to elucidate the potential cause of cork taint, Journal of agricultural and food chemistry 50, 1032-1039. [29] Kobilka, B. K. (2007) G protein coupled receptor structure and activation, Biochim Biophys Acta 1768, 794-807. [30] Chen, V. B., Arendall, W. B., 3rd, Headd, J. J., Keedy, D. A., Immormino, R. M., Kapral, G. J., Murray, L. W., Richardson, J. S., and Richardson, D. C. (2010) MolProbity: all-atom structure validation for macromolecular crystallography, Acta crystallographica. Section D, Biological crystallography 66, 12-21. [31] Morris, G. M., Goodsell, D. S., Halliday, R. S., Huey, R., Hart, W. E., Belew, R. K., and Olson, A. J. (1998) Automated docking using a Lamarckian genetic algorithm and an empirical binding free energy function, Journal of computational chemistry 19, 1639-1662. [32] Morris, G. M., Huey, R., Lindstrom, W., Sanner, M. F., Belew, R. K., Goodsell, D. S., and Olson, A. J. (2009) AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility, Journal of computational chemistry 30, 2785-2791. [33] Cerqueira, N. M. F. S. A., Ribeiro, J., Fernandes, P. A., and Ramos, M. J. (2011) vsLab—An implementation for virtual high-throughput screening using AutoDock and VMD, International Journal of Quantum Chemistry 111, 1208-1212.


Page 33 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


[34] Ribeiro, J. V., Tamames, J. A., Cerqueira, N. M., Fernandes, P. A., and Ramos, M. J. (2013) Volarea - a bioinformatics tool to calculate the surface area and the volume of molecular systems, Chemical biology & drug design 82, 743-755.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment

Page 34 of 34

Studying Haloanisoles Interaction with Olfactory Receptors - ACS

Recommend Documents