Comparative receptor surface analysis of octopaminergic antagonists for the locust neuronal octopamine receptor

Computational Biology and Chemistry 27 (2003) 531–540

Comparative receptor surface analysis of octopaminergic antagonists for the locust neuronal octopamine receptor Akinori Hirashima a,∗ , Eiichi Kuwano a , Morifusa Eto b a

Department of Applied Genetics and Pest Management, Faculty of Agriculture, Graduate School, Kyushu University, 6-10-1 Hakozaki, Higashi-ku, Fukuoka 812-8581, Japan b Professor Emeritus of Kyushu University, 7-32-2 Aoba, Higashi-ku, Fukuoka 813-0025, Japan Received 21 April 2003; received in revised form 25 July 2003; accepted 25 July 2003

Abstract In drug discovery, it is common to have measured activity data for a set of compounds acting upon a particular protein but not to have knowledge of the three-dimensional structure of the protein active site. In the absence of such three-dimensional information, one can attempt to build a hypothetical model of the receptor site that can provide insight about receptor site characteristics. Such a model is known as a comparative receptor surface analysis (CoRSA) model, which provides compact and quantitative descriptors which capture three-dimensional information about a putative receptor site. The quantitative structure–activity relationship (QSAR) of a set of 20 antagonists for octopamine (OA) receptor 3 in locust nervous tissue, was analyzed using CoRSA. Three-dimensional energetics descriptors were calculated from receptor surface model (RSM)–ligand interaction and these three-dimensional descriptors were used in QSAR analysis. The predictive character of the QSAR was further assessed using 24 agonists for OA receptor as test molecules. An RSM was generated using some subset of the most active structures and the results provided useful information in the characterization and differentiation of OA receptor. © 2003 Elsevier Ltd. All rights reserved. Keywords: Locusta migratoria; Quantitative structure–activity relationship; Comparative receptor surface analysis; Cerius2; Antagonist for octopamine receptor; Receptor surface model

1. Introduction Octopamine (OA, 2-amino-1-(4-hydroxyphenyl)ethanol) which has been found to be present in high concentrations in various insect tissues, is the monohydroxylic analogue of the vertebrate hormone noradrenalin. OA was first discovered in the salivary glands of octopus by Erspamer and Boretti (1951). It has been found that OA is present in a high concentration in various invertebrate tissues (Axelrod and Saavedra, 1977). This multifunctional and naturally occurring biogenic amine has been well studied and established as: (1) a neurotransmitter, controlling the firefly light organ and endocrine gland activity in other insects; (2) a neurohormone, inducing mobilization of lipids and carbohydrates; (3) a neuromodulator, acting peripherally on different muscles, fat body, corpora cardiaca, and the corpora allata; and (4) a centrally acting neuromodulator, influencing motor ∗ Corresponding author. Tel.: +81-92-642-2856; fax: +81-92-642-2858. E-mail address: [email protected] (A. Hirashima).

1476-9271/$ – see front matter © 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.compbiolchem.2003.07.001

patterns, habituation, and even memory in various invertebrate species (Evans, 1985, 1993). The action of OA is mediated through various receptor classes and three different receptor classes OAR1, OAR2A, and OAR2B had been distinguished from non-neuronal tissues (Evans, 1981), in which OAR2 is coupled to G-proteins and is specifically linked to an adenylate cyclase. Thus, the physiological actions of OAR2 has been shown to be associated with elevated levels of cAMP (Nathanson, 1985). In the nervous system of locust Locusta migratoria L., a particular receptor class was characterized and established as a new class OAR3 by pharmacological investigations of the [3 H]-OA binding site using various agonists and antagonists (Roeder, 1990, 1992, 1995; Roeder and Gewecke, 1990; Roeder and Nathanson, 1993). Recently, much attention has been directed at the octopaminergic system as a valid target in the development of safer and selective pesticides (Jennings et al., 1988; Hirashima et al., 1992; Ismail et al., 1996). Structure–activity studies of various types of agonists and antagonists for OA receptor were also reported using the nervous tissue of the

532

A. Hirashima et al. / Computational Biology and Chemistry 27 (2003) 531–540

migratory locust, L. migratoria L. (Roeder, 1990, 1992, 1995; Roeder and Gewecke, 1990; Roeder and Nathanson, 1993). However, information on the structural requirements of these agonists and antagonists for OA receptor with high receptor–ligand affinity is still limited. It is therefore of critical importance to provide information on the pharmacological properties of this OA receptor types and subtypes. Our interest in agonists for OA receptor was aroused by the results of quantitative structure–activity relationship (QSAR) studies using various physicochemical parameters as descriptors (Hirashima et al., 1999a; Pan et al., 1997a) and receptor surface model (RSM) (Hirashima et al., 1998a). Furthermore, molecular modeling and conformational analysis were performed in catalyst/hypo to gain a better knowledge of the interactions between antagonists and OAR3 in order to understand the conformations required for binding activity (Pan et al., 1997b). A similar procedure was repeated using agonists for OAR3 (Hirashima et al., 1999b). In drug discovery, it is common to have measured activity data for a set of compounds acting upon a particular protein but not to have knowledge of the three-dimensional structure of the protein active site. In the absence of such three-dimensional information, one can attempt to build a hypothetical model of the receptor site that can provide insight about receptor site characteristics. Such an analysis is known as a comparative receptor surface analysis (CoRSA), which provides compact and quantitative descriptors which capture three-dimensional information about a putative receptor site (Ivanciuc et al., 2000). Thus, the current work is aimed to perform three-dimensional CoRSA on a set of antagonists for OAR3 in thoracic nerve system of L. migratoria. 2. Materials and methods 2.1. Synthesis of agonists for OA receptor 2-(Aralkylamino)-2-thiazolines (AATs, 21–34) were synthesized by cyclization of the corresponding thiourea with concentrated hydrogen chloride (Hirashima et al., 1992). 3-(Substituted phenyl)imidazolidine-2-thiones (SPITs, 37 and 38) were synthesized by the cyclization of monoethanolamine hydrogen sulfate with the corresponding arylisothiocyanate in the presence of sodium hydroxide as described in the previous report (Hirashima et al., 1998b). 2-(Phenethylmercapto)-2-thiazoline (PMT, 44) was obtained by stirring 2-mercaptothiazoline and the corresponding phenethyl bromide overnight in pyridine (Hirashima et al., 1992). The structures of the compounds were confirmed by [3 H]- and [13 C]-NMR measured with a JEOL JNM-EX400 spectrometer at 400 MHz, tetramethyl silane (TMS) being used as an internal standard for [3 H]-NMR, and elemental analytical data. Data for other compounds 2-(arylimino)imidazolidine (AII, 35 and 36), arylethanolamine (AEA, 39 and 40), adrenalin (41), chlordimeform (CDM, 42), and 2-phenyl-2-imidazolidine (43) was cited from data by Roeder (1990).

2.2. Computational details 2.2.1. Molecular alignment All computational experiments were conducted with Cerius2 3.8 QSAR environment from Accelrys (San Diego, USA) on a Silicon Graphics O2, running under the IRIX 6.5 operating system. Multiple conformations of each molecule were generated using the Boltzmann Jump as a conformational search method. The upper limit of the number of conformations per molecule was 150. Each conformer was subjected to an energy minimization procedure to generate the lowest energy conformation for each structure. Alignment of structures through pair-wise superpositioning placed all structures in the study compounds in the same frame of reference as the shape reference compound, which was selected as a conformer of the most active antagonist for OAR3. The method used for performing the alignment was maximum common subgroup (MCSG). This method looks at molecules as points and lines, and uses the techniques of graph theory to identify patterns. It finds the largest subset of atoms in the shape reference compound that is shared by all the structures in the study table and uses this subset for alignment. A rigid fit of atom pairings was performed to superimpose each structure so that it overlays the shape reference compound. 2.2.2. CoRSA RSMs proposed by Hahn (1995), and Hahn and Rogers (1995) are predictive and sufficiently reliable to guide the chemist in the design of novel compounds. These descriptors are used for predictive QSAR models. This approach is effective for the analysis of data sets where activity information is available but the structure of the receptor site is unknown. Thus, activity data was used for the CoRSA, which attempts to postulate and represent the essential features of a receptor site from the aligned common features of the molecules that bind to it. This method generates multiple models that can be checked easily for validity. The RSM was tested for prediction with the leave-one-out cross-validation method. Once a reasonable RSM has been defined, a series of structures can be evaluated against the model. When a receptor model has been generated and the models have been aligned, a QSAR can be built using data from the receptor–structure interactions. The results of the minimization procedure were used as descriptors either to refine the model or to predict activity. For prediction, the molecules were minimized in the RSM. Three-dimensional energetics descriptors were calculated from RSM–ligand interaction. These three-dimensional descriptors were used in QSAR analysis. An RSM represents the global volume that can accommodate one or more molecules and can be seen as the shape of an active site built from the ligands that fit into it in their “active” conformation. The descriptors used in this study account for phenomena that occur at the contact surface between the ligands and the protein active site. An RSM represents essential information about the hypothetical

A. Hirashima et al. / Computational Biology and Chemistry 27 (2003) 531–540

receptor site as a three-dimensional surface with associated properties mapped onto the surface model. The location and shape of the surface represent information about the steric nature of the receptor site: the associated properties represent other information of interest, such as hydrophobicity, partial charge, electrostatic potential, and H-bonding propensity. The isosurface procedure produces a surface that entirely encloses the molecules over which it is generated. The surface has no holes and is known as a closed model. RSMs are best constructed from a set of the most active analogues that are chosen to cover the variety of steric and electrostatic variations likely to appear in the test data. The approach we took was to automatically build a set of different RSMs from different combinations of the most active analogues, and then use a variable-selection technique such as genetic partial least squares (G/PLS) to discover the RSM whose descriptors yield the best QSARs of the full training set. G/PLS allows the discovery and use of nonlinear descriptors by using spline-based terms. 2.2.3. G/PLS G/PLS, a variation of genetic function approximation (GFA), was run as an alternative to the standard GFA algorithm. G/PLS is derived from the best features of two methods: GFA and partial least squares (PLS). Both GFA and PLS have been shown to be valuable analysis tools in cases where the data set has more descriptors than samples. In PLS, variables might be overlooked during interpretation or in designing the next experiment even though cumulatively they are very important. This phenomenon is known as “loading spread”. In GFA, equation models have a randomly chosen proper subset of the independent variables. As a result of multiple linear regression on each model, the best ones become the next generation and two of them produce an offspring. This was repeated 10,000 (default 5000) times. For other settings, all defaults were used. Loading spread does not occur because at most one of a set of co-linear variables is retained in each model. G/PLS combines the best features of GFA and PLS (Cerius2 tutorial, Accelrys Inc., http://www.accelrys.com/cerius2) and actually G/PLS gave better results than in cases GFA or PLS was used. G/PLS retains the ease of interpretation of GFA by back-transforming the PLS components to the original variables.

3. Results A set of 20 molecules, whose inhibitory activities were tested elsewhere using the [3 H]-OA binding to OAR3 in the locust central nervous tissue, was selected from published data (Roeder, 1990) as the target training set. The molecular structures and experimental biological activities are listed in Fig. 1 and Table 1. Some models were statistically significant and were used to correctly predict the activities of a set of test molecules ranging over five orders of magnitude

533

(maximum pKi 8.99 and minimum pKi 4.09), indicating that these models could be useful tools to design active antagonists for OA receptor. This method generates multiple models that can be checked easily for validity. Maroxepine (13) showed the highest activity followed by mianserin (15) and hydroxymianserin (10) in study compounds. An RSM was generated (Fig. 2a and b) using some subset of the most active structures (10, 13 and 15). The rationale underlying this model is that the most active structures tend to explore the best spatial and electronic interactions with receptor, while the least active ones do not and tend to have unfavorable steric or electronic interactions. A rigid fit was performed to superimpose each structure so that it overlays the shape reference compound 13. Antagonists for OAR3 10, 13 and 15 were used to generate a virtual RSM with the van der Waals function (Fig. 2a) or the Wyvill steric function (Fig. 2b). The van der Waals steric function gives a hard receptor, very similar with the surface area of the compounds that generate the receptor, while the Wyvill steric function gives a soft receptor, a fuzzy representation of the molecular surface area, with much larger limits. A comparison between the shapes of the van der Waals and Wyvill receptors from Fig. 2a and b clearly shows the differences between them: while in the former the shape of the atoms can be easily recognized, the later has a fuzzy shape, with a larger distance between atoms and the surface points. The RSM is colored by H-bond contribution: a light-blue sign stands for a positive contribution of H-bond and a white sign stands for a negative contribution of H-bond. When this property is mapped, the color indicates the tendency for specific areas of the surface to act as H-bond acceptors (light-blue). Areas of the model with no H-bonding activity are colored white. The white regions are spread almost on the entire molecule, with the exception of hetero-atom regions colored in light-blue: the hetero-atoms such as oxygen and nitrogen contribute to H-bond acceptors. This color coding of the receptor–ligand interactions can offer a qualitative way of examining compounds, by introducing them into the virtual receptor and visually inspecting the favorable/unfavorable interactions; substituents that increase or decrease the binding affinity can be easily recognized, and one can make easily simple but accurate structure–activity estimations. The energies of interaction between the RSM and each molecular model were added to the study table as new columns, which were used for generating QSARs. Instead of one total number which is the sum of the interactions evaluated between each point on the surface and each molecular model, leading to one extra column in the study table, the energies at each surface point are available. Depending on the size of the drug molecules, this is potentially a great number of surface points. In order to quantitatively understand the dependence of biological activities on RSM parameters of antagonists for OA receptor, regression analysis was applied to representative 20 study compounds listed in Fig. 1 and Table 1. The best model generated using the descriptors from the closed RSM is given in Eq. (1), which is

534

A. Hirashima et al. / Computational Biology and Chemistry 27 (2003) 531–540

Cl N

N

N

N

N S

HN

1 Amitriptyline

2 Antazoline

N

3 Chlorpromazine

4 Cyproheptadine

N

H N

H N

HO O N

N H

N

N

NH N O

N

O

N

O 5 Demethylmianserin

7 Dihydroergotamine

6 Desipramine

8 Eresepine

OH N NH

OH

N

HO

N

OH

N 10 Hydroxymianserin

9 Gramine

H N

N

12 Isopropylarterenol

11 Imipramine

HN O O

N

O

N N

N H

N NH2

N

N

Cl

13 Maroxepine

14 Metoclopramide

OH

15 Mianserin

16 Phentolamine HO

N N

S N

N

O OH

17 Promethazine

O

N H

18 Propranolol

N

O HN

19 Triprolidine

20 Yohimbine

Fig. 1. Structures of antagonists for OAR3 used for regression analysis in study set.

similar with the 4D-QSAR of Hong and Hopfinger (2003). The number of variables for Eq. (1) was 1796. Ten percent of all new significant columns of variables were automatically used as independent X variables in the generation of QSAR:

+ 4.99813(ELE/3642) + 6.37002(ELE/3647) + 12.3139(ELE/4348) + 5.4795(ELE/4812) − 19.3455(VDW/2404)2 + 9.41944(VDW/3452)2 + 5.10814(VDW/4414)2 + 0.890613(VDW/5841) − 2.37094(TOT/804)2

pKi = 8.92612 − 10.1319(ELE/1959)

− 4.44977((TOT/3262) + 0.37734)

+ 5.82153((ELE/2006) − 0.003024) − 5.32257((ELE/2824) − 0.044163)

2

− 4.56253((TOT/3262) − 0.295216)

(1)

A. Hirashima et al. / Computational Biology and Chemistry 27 (2003) 531–540

535

Table 1 Regression analysis of structure–antagonist activities for OAR3 in study set Compound

MW

No.

R

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Amitriptyline Antazoline Chlorpromazine Cyproheptadine Demethylmianserin Desipramine Dihydroergotamine Eresepine Gramine Hydroxymianserin Imipramine Isopropylarterenol Maroxepine Metoclopramide Mianserin Phentolamine Promethazine Propranolol Triprolidine Yohimbine a b c

277.408 279.384 318.863 273.377 250.343 266.385 583.686 290.407 174.245 280.369 280.412 211.260 277.365 299.800 264.369 281.357 284.418 245.321 277.408 354.448

Vm

283.561 273.214 238.948 265.187 240.361 265.528 514.862 280.058 174.771 264.488 281.673 202.829 259.500 276.001 256.765 263.809 268.206 240.035 282.609 325.622

Ki (nM)a

1570 118 766 844 55 3210 272 474 1840 68.0 1450 18600 1.02 52601 1.20 19 32 29200 2900 82035

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

pKi

330 62 54 356 3 610 30 166 1012 0.5 333 4840c 0.18 6500 0.27 9 23 11920 145 22400

Observed

Calculatedb

Difference

5.80 6.93 6.12 6.07 7.26 5.49 6.57 6.32 5.74 8.77 5.84 4.77 8.99 4.28 8.92 7.72 7.45 4.53 5.54 4.09

5.73 7.00 5.92 5.93 7.74 5.79 6.56 6.59 5.55 8.77 5.64 4.75 8.56 4.38 8.80 7.78 7.58 4.54 4.58 4.03

0.07 −0.07 0.20 0.14 −0.48 −0.30 0.01 −0.27 0.19 0 0.20 0.02 0.43 −0.10 0.12 −0.06 −0.13 −0.01 −0.04 0.06

Cited from data by Roeder (1990). Calculated by Eq. (1). Cited from data by Roeder (1995).

where n = 20, r2 = 0.981, CV-r 2 = 0.712, PRESS = 11.917, and Bsr 2 = 0.848 ± 0.955. The descriptors ELE/1959, ELE/2006, etc. are the electrostatic interaction energy of the molecule with the receptor at point 1959, 2006, etc. The descriptors VDW/2404, VDW/3452, etc. are the Van der Waals interaction energy of the molecule with the receptor at point 2404, 3452, etc. The descriptors TOT/804, TOT/3262, etc. are added energy of both electrostatic interaction energy and Van der Waals interaction energy at point 804, 3262, etc. The term n means the number of data points; r-squared (r2 ), the square of the correlation coefficient, which is used to describe the goodness of fit of the data of the study compounds to the QSAR model; cross-validated r2 (CV-r2 ), a squared correlation coefficient generated during a validation procedure using the equation: CV-r2 = (SD − PRESS)/SD; predicted sum of squares (PRESS), the sum of overall compounds of the squared differences between the actual and the predicted values for the dependent variables; SD, the sum of squared deviations of the dependent variable values from their mean. The PRESS value is computed during a validation procedure for the entire training set. The larger the PRESS value, the more reliable is the equation. A CV-r2 is usually smaller than the overall r2 for a QSAR equation. It is used as a diagnostic tool to evaluate the predictive power of an equation generated using the G/PLS method. Cross-validation is often used to determine how large a model (number of terms) can be used for a given data set. For instance, the number of components retained in G/PLS can be selected to be that which gives the highest CV-r2 . Bootstrap r2 (Bsr2 ) is the average

squared correlation coefficient calculated during the validation procedure (Cerius2 tutorial). A Bsr2 is computed from the subset of variables used one-at-a-time for the validation procedure. It can be used more than one time in computing the r2 statistic. Table 1 depicts structures of antagonists for OAR3, their experimental Ki values, calculated pKi values using Eq. (1), and difference between observed and calculated pKi values. In case predicted activity is overestimated, deviation is obtained by calculating predicted activity subtracted by experimental value and indicated by minus. In case predicted activity is underestimated, deviation is obtained by calculating experimental activity subtracted by predicted value. The RSM was statistically significant and used to correctly predict the activities of a set of training molecules, indicating that these models could be useful tools to design active agonists for OA receptor. Residuals (observed versus calculated from Table 1 ) are plotted in Fig. 3 . Once the desired RSM has been constructed, all the structures in the test sets were evaluated against the model. The evaluation consists of computing several energetic descriptors that are based upon the interactions between ligand and model. By using receptor data to develop a QSAR model, the goodness of fit can be evaluated between a candidate structure and a postulated pseudo receptor. The predictive character of the QSAR was further assessed using 24 agonists for OA receptor as test molecules, whose structures are shown in Fig. 4, outside of the training set. The best statistically significant Eq. (1) was applied to access these agonists for OAR3. The predicted values of these molecules are listed in Table 2, which depicts agonists for OAR3, their

536

A. Hirashima et al. / Computational Biology and Chemistry 27 (2003) 531–540

Fig. 2. The top antagonists for OAR3 10, 13 and 15 embedded in an RSM generated from them computed with the van der Waals function (a) and the Wyvill steric function colored by H-bond (b): a blue sign stands for a positive contribution of H-bond and a white sign stands for a negative contribution of H-bond. The oxygen and nitrogen atoms contribute as H-bond acceptors.

A. Hirashima et al. / Computational Biology and Chemistry 27 (2003) 531–540

537

Fig. 3. Correlation of observed pKi values (horizontal) vs. calculated pKi values (vertical) from Table 1 using Eq. (1) in study set.

experimental Ki values, calculated pKi values using Eq. (1), and difference between observed and calculated pKi values. Some agonists for OAR3 were active according to Eq. (1) in inhibiting the binding of [3 H]-OA to OAR3. Residuals (observed versus calculated from Table 2) are plotted in Fig. 5. The number of terms in Eq. (1), incidentally, depends on the size of the molecule. This means that it is difficult to compare results of compounds that are very different in size. This does not seem to be a problem in the present case, although test compounds in Fig. 4 are on the whole a lot smaller than

N

N

R

S

N H 21-34 AAT OH HO HO 41 Adrenalin

H N

R

R

35-36 AII

N Cl

42 CDM

OH

S

HN

H N

those of the training set (Fig. 1). Molecular weight (MW) and molecular volume (Vm ) are included in Tables 1 and 2. Agonists for OA receptor is not likely to penetrate either the cuticle or the central nervous system of insects effectively, since it is fully ionized at physiological pH. Derivatization of the polar groups would be one possible solution to this problem in trying to develop potential pest-control agents. The above CoRSA studies show that phenyl ring substitution requirements for antagonists for OA receptor differ substantially from each other and other various types of

N

N

NH

R

NH2

37-38 SPIT

39-40 AEA

H N

S

N 43 2-Phenyl-2-imidazoline 44 PMT

Fig. 4. Structures of agonists for OAR3 used for regression analysis in test set.

S N

538

A. Hirashima et al. / Computational Biology and Chemistry 27 (2003) 531–540

Table 2 Regression analysis of structure–agonist activities for OAR3 in test set Compound

MW

Vm

No.

R

AAT 21 22 23 24 25 26 27 28 29 30 31 32 33 34

2-Cl-PhCH2 2-F-PhCH2 2-CH3 -PhCH2 2-CF3 -PhCH2 3-Cl-PhCH2 3-CF3 -PhCH2 3-NO2 -PhCH2 4-F-PhCH2 2-Cl,4-F-PhCH2 2,6-Cl2 -PhCH2 3-Cl,4-F-PhCH2 PhCH3 CH(D) 4-Cl-PhCH2 CH2 3-Pyridyl-CH2

226.723 210.269 206.305 260.276 226.723 260.276 237.276 210.269 244.714 261.168 244.714 206.305 240.75 193.266

189.945 180.856 191.515 206.706 189.433 206.012 192.672 180.703 194.393 203.048 194.043 191.906 205.46 170.914

AII 35 36

2,6-(CH3 )2 2,4,6-Cl3

226.723 264.541

187.846 193.192

SPIT 37 38

2,4-(CH3 )2 2,4,5-Cl3 -PhCH2

206.305 281.587

190.917 199.116

AEA 39 40 41 42 43 44

3-OH 3,4-(OH)2 Adrenalin CDM 2-Phenyl-2-imidazolidine PMT

153.180 169.180 183.207 196.679 144.176 223.350

145.195 153.724 171.004 180.432 134.773 198.905

a b c

Ki (nM)

pKi Observed

Calculateda

Difference

6.36 6.35 6.19 6.54 7.02 6.42 6.73 6.34 6.74 5.90 6.96 5.30 6.58 5.08

5.59 6.53 5.92 6.37 6.27 6.66 5.86 6.17 6.51 5.86 5.98 5.44 6.43 5.87

0.77 −0.18 0.27 0.17 0.75 −0.24 0.87 0.17 0.23 0.04 0.98 −0.14 0.15 −0.79

20 ± 7c 19 ± 3c

7.70 7.73

7.04 6.94

0.66 0.79

1660 ± 700b 1040 ± 730b

5.78 5.98

5.37 5.12

0.41 0.86

± ± ± ± ± ±

5.30 6.32 6.38 6.91 4.79 6.12

5.82 5.87 5.87 6.99 5.50 5.27

−0.52 0.45 0.51 −0.08 −0.71 0.85

440 447 650 290 95 380 185 460 184 1270 109 5000 264 8330

5050 475 416 137 16200 760

± ± ± ± ± ± ± ± ± ± ± ± ± ±

189b 125b 360b 203b 66b 266b 33b 124b 90b 635b 59b 2600b 95b 2500b

1860c 42c 75c 70c 4700c 243b

Calculated by Eq. (1). Personal communication (T. Roeder, Hamburg University, Germany). Cited from data by Roeder (1995).

antagonists for OA receptor could be potent, although the type of compounds tested here is still limited to draw any conclusions. These derivatives could provide useful information in the characterization and differentiation of OA receptor. The agonists for OA receptor showed reasonable predicted activities according to Eq. (1). The result may imply that the process of calculating an RSM treats these structures reasonably. The CoRSA could provide useful information in the characterization and differentiation of OA receptor. It may help to point the way towards developing extremely potent and relatively specific OA ligands, leading to potential insecticides, although further research on the comparison of the 3D-QSAR is necessary. In order to optimize the activities of these compounds as OA ligands, more detailed experiments are in progress.

4. Discussion RSM is quantitative and differs from pharmacophore models, which are qualitative, in that the former tries to capture essential information about the receptor, while the

latter only captures information about the commonality of compounds that bind. RSM tends to be geometrically overconstrained (and topologically neutral) since, in the absence of steric variation in a region, they assume the tightest steric surface which fits all training compounds. RSMs do not contain atoms, but try to directly represent the essential features of an active site by assuming complementarity between the shape and properties of the receptor site and the set of binding compounds. The CoRSA application uses three-dimensional surfaces that define the shape of the receptor site by enclosing the most active members (after appropriate alignment) of a series of compounds. The global minimum of the most active compound 13 in the study compounds (based on the value in the activity column) was made as the active conformer. It really is just one of possibly many self-consistent models that fit the biological activity data. This model ought to be predictive and sufficiently reliable to guide the chemist in the design of novel compounds. These descriptors were used for predictive QSAR models. This approach is effective for the analysis of data sets where activity information is available but the structure of the receptor site is unknown. CoRSA

A. Hirashima et al. / Computational Biology and Chemistry 27 (2003) 531–540

539

Fig. 5. Correlation of observed pKi values (horizontal) vs. calculated pKi values (vertical) from Table 2 using Eq. (1) in test set.

attempts to postulate and represent the essential features of a receptor site itself, rather than the common features of the molecules that bind to it. The agonists for OAR3 showed reasonable predicted activities according to Eq. (1). The result may imply that the process of calculating an RSM treats these structures reasonably, although antagonists may not interact with the same part of the membrane with which the agonists interact. The presence of some common structural elements, such as the phenyl ring, suggests the binding sites may have some features in common. Taken the part of the membrane with which the agonist interacts as the true receptor, the antagonist may well interact with an area surrounding the receptor. The CoRSA could provide useful information in the characterization and differentiation of OA receptor.

Acknowledgements This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, and Culture of Japan.

References Axelrod, J., Saavedra, J.M., 1977. Octopamine. Nature 265, 501– 504. Erspamer, V., Boretti, G., 1951. Identification and characterization by paper chromatography of enteramine, octopamine, tyramine, histamine, and allied substances in extracts of posterior salivary glands of Octopada and in other tissues of vertebrates and invertebrates. Arch. Int. Pharmacodyn. Ther. 88, 296–332. Evans, P.D., 1981. Multiple receptor types for octopamine in the locust. J. Physiol. 318, 99–122. Evans, P.D., 1985. Octopamine. In: Kerkut, G.A., Gilbert, G. (Eds.), Comprehensive Insect Physiology Biochemistry Pharmacology, vol. 11. Pergamon Press, Oxford, pp. 499–530. Evans, P.D., 1993. Molecular studies on insect octopamine receptors. In: Heller, S.R. (Ed.), Comparative Molecular Neurobiology. Birkhäuser Verlag, Basel, pp. 287–296. Hahn, M., 1995. Receptor surface models. 1. Definition and construction. J. Med. Chem. 38, 2080–2090. Hahn, M., Rogers, D., 1995. Receptor surface models. 2. Application to quantitative structure–activity relationships studies. J. Med. Chem. 38, 2091–2102. Hirashima, A., Yoshii, Y., Eto, M., 1992. Synthesis and octopaminergic agonist activity of 2-(substituted benzylamino)-2-thiazolines. Biosci. Biotechnol. Biochem. 56, 1062–1065. Hirashima, A., Pan, C., Tomita, J., Kuwano, E., Taniguchi, E., Eto, M., 1998a. Quantitative structure–activity studies of octopaminergic

540

A. Hirashima et al. / Computational Biology and Chemistry 27 (2003) 531–540

agonists and antagonists against nervous system of Locusta migratoria. Bioorg. Med. Chem. 6, 903–910. Hirashima, A., Kuwano, E., Taniguchi, E., Eto, M., 1998b. Synthesis and octopaminergic-agonist activity of 3-(substituted phenyl)imidazolidine-2-thiones and related compounds. Biosci. Biotechnol. Biochem. 62, 1179–1184. Hirashima, A., Shinkai, K., Pan, C., Kuwano, E., Taniguchi, E., Eto, M., 1999a. Quantitative structure–activity studies of octopaminergic ligands against Locusta migratoria and Periplaneta americana. Pestic. Sci. 55, 119–128. Hirashima, A., Pan, C., Kuwano, E., Taniguchi, E., Eto, M., 1999b. Three-dimensional pharmacophore hypotheses for the locust neuronal octopamine receptor (OAR3). 2. Agonists. Bioorg. Med. Chem. 7, 1437–1443. Hong, X., Hopfinger, A.J., 2003. 3D-pharmacophores of flavonoid binding at the benzodiazepine GABA(A) receptor site using 4D-QSAR analysis. J. Chem. Inf. Comput. Sci. 43, 324–336. Ismail, S.M.M., Baines, R.A., Downer, R.G.H., Dekeyser, M.A., 1996. Dihydrooxadiazines: octopaminergic system as a potential site of insecticidal action. Pestic. Sci. 46, 163–170. Ivanciuc, O., Ivanciuc, T., Cabrol-Bass, D., 2000. 3D quantitative structure–activity relationships with CoRSA: comparative receptor surface analysis. Application to calcium channel agonists. Analysis 28, 637–642.

Jennings, K.R., Kuhn, D.G., Kukel, C.F., Trotto, S.H., Whiteney, W.K., 1988. A biorationally synthesized octopaminergic insecticide: 2-(4-chloro-o-toluidino)-2-oxazoline. Pestic. Biochem. Physiol. 30, 190–197. Nathanson, J.A., 1985. Phenyliminoimidazolidines. Characterization of a class of potent agonists of octopamine-sensitive adenylate cyclase and their use in understanding the pharmacology of octopamine receptors. Mol. Pharmacol. 28, 254–268. Pan, C., Hirashima, A., Tomita, J., Kuwano, E., Taniguchi, E., Eto, M., 1997a. Internet J. Sci. Biol. Chem. 1, http://www.netsci-journal.com/ 97v1/97013/index.htm. Pan, C., Hirashima, A., Kuwano, E., Eto, M., 1997b. Three-dimensional pharmacophore hypotheses for the locust neuronal octopamine receptor (OAR3). 1. Antagonists. J. Mol. Model. 3, 455–463. Roeder, T., 1990. High-affinity antagonists of the locust neuronal octopamine receptor. Eur. J. Pharmacol. 191, 221–224. Roeder, T., 1992. A new octopamine receptor class in locust nervous tissue, the octopamine 3 (OA3) receptor. Life Sci. 50, 21–28. Roeder, T., 1995. Pharmacology of the octopamine receptor from locust central nervous tissue (OAR3). Br. J. Pharmacol. 114, 210–216. Roeder, T., Gewecke, M., 1990. Octopamine receptors in locust nervous tissue. Biochem. Pharmacol. 39, 1793–1797. Roeder, T., Nathanson, J.A., 1993. Characterization of insect neuronal octopamine receptors (OA3 receptors). Neurochem. Res. 18, 921–925.