JOURNAL HOME CME HOME THIS MONTH PAST ISSUES ETOC COLLECTIONS
AUTHORS REVIEWERS EDITORIAL BOARD FEEDBACK RSS HELP
A&A International Anesthesia Research Society
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a colleague
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sewell, J. C.
Right arrow Articles by Sear, J. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sewell, J. C.
Right arrow Articles by Sear, J. W.
Related Collections
Right arrow Mechanisms
Right arrow Pharmacology

Anesth Analg 2004;99:744-751
© 2004 International Anesthesia Research Society
doi: 10.1213/01.ANE.0000129978.92936.A2

ANESTHETIC PHARMACOLOGY

Derivation of Preliminary Three-Dimensional Pharmacophores for Nonhalogenated Volatile Anesthetics

Jason C. Sewell, PhD, and John W. Sear, PhD FFARCS

Nuffield Department of Anaesthetics, University of Oxford, John Radcliffe Hospital, Headington, Oxford, United Kingdom

Address correspondence to John W. Sear, PhD, FFARCS, Nuffield Department of Anesthetics, University of Oxford, The John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK. Address e-mail to john.sear{at}nda.ox.ac.uk Reprints will not be available from the authors.


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Discussion
 References
 
We investigated the molecular basis for the immobilizing activity of nonhalogenated volatile anesthetics by using comparative molecular field analysis (CoMFA). In vivo potency data (expressed as minimum alveolar anesthetic concentrations) for 38 structurally diverse drugs were obtained from the literature. The anesthetics were randomly divided into a training-set (n = 28) used to formulate the activity models and a test-set (n = 10) used to independently assess the models’ predictive power. The anesthetic structures were aligned to maximize their similarity in molecular shape and electrostatic potential to conformers of the most active drug in the group: hexanol. The individual conformers and alignments with maximum similarity (calculated with combined Carbo indices) were retained and used to derive the CoMFA activity models. The final CoMFA model explained 95.5% of the variance in the observed activities of the training-set anesthetics. The model had good predictive capability for both the training-set drugs (cross-validated r2 = 0.824) and the randomly excluded test-set anesthetics (r2 = 0.921). Pharmacophoric maps were derived by identifying the spatial distribution of key areas in which steric and electrostatic interactions are important in determining the immobilizing activity of the anesthetics considered.

IMPLICATIONS: We have derived an activity model for a group of structurally diverse nonhalogenated volatile anesthetics that correlates in vivo potency (minimum alveolar anesthetic concentration) with the spatial distribution of their molecular bulk and electrostatic potential. Our results suggest that there is a common molecular basis for the immobilizing activity of the anesthetics.


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Discussion
 References
 
Studies of the molecular basis of general anesthetic activity have been predominantly target oriented, focusing on the interactions of anesthetics with putative sites of action (1,2). For example, many studies have investigated the molecular specificity of anesthetic interactions with ligand-gated ion channels in vitro (3,4), yet the relative importance of these targets in achieving general anesthesia remains unresolved and controversial (1).

An alternative approach to investigating the molecular basis of activity is to focus on the anesthetic molecules themselves. The aims of such ligand-oriented approaches are to identify the molecular properties that determine activity and to formulate a model that correlates the magnitude of these properties with potency. One example of a ligand-oriented modeling technique is comparative molecular field analysis (CoMFA), a computer-aided drug design method frequently used when the detailed structure of a receptor is unknown (5). In CoMFA, the ligand compounds are aligned and placed in a grid consisting of regularly spaced lattice points. The steric and electrostatic interaction energies between the ligands and a charged probe atom are calculated at each point and correlated with potency to formulate an activity model. By identifying which lattice points make the greatest contribution to the model, three-dimensional pharmacophoric maps of the key steric and electrostatic features of the ligands that determine activity can be derived.

One of the major obstacles in applying CoMFA and related modeling approaches to general anesthetics is the structural diversity of the compounds, which precludes aligning them in the lattice grid by a common substructure. We have recently described how chemically diverse IV anesthetics can be aligned on the basis of similarities in their molecular shapes and electrostatic potentials (6,7), which overcomes this fundamental difficulty. In this study, we developed the molecular similarity approach to formulate a preliminary CoMFA activity model for a wide range of structurally diverse nonhalogenated volatile general anesthetics. We investigated the molecular basis for their immobilizing activity by using CoMFA to correlate in vivo potency with the spatial distribution of the molecules’ steric and electrostatic features.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Discussion
 References
 
A group of 38 volatile nonhalogenated general anesthetics were considered, consisting of alkanes, alkenes, alkynes, cycloalkanes, aromatics, alcohols, thiols, and an ether (Table 1). In vivo potency data, expressed as the minimum alveolar concentration of anesthetic that produces unresponsiveness to noxious stimulation in rats (MAC), were obtained from the literature (8–14). This group of anesthetics represents the widest range of nonhalogenated compounds for which comparable potency data were available and were suitable for the modeling techniques used (the limitations of which are outlined in Discussion).


View this table:
[in this window]
[in a new window]
  		Table 1. Anesthetic Potencies of the Nonhalogenated Volatile Compounds Considered in This Study
 
The anesthetics were subdivided into training- and test-sets (Table 1). The training-set (n = 28) was used to formulate the CoMFA activity models, and the test-set (n = 10) was used to independently evaluate the models’ predictive capability. The selection of the test-set anesthetics was random, except for the following compounds, which were forced into the training-set: hexanol (the most potent anesthetic in the group, which was used as an alignment template for the other anesthetic molecules), pentanol (the second most active drug), and decane (whose length determined the dimensions of the CoMFA lattice grid). The test-set was selected by ranking the anesthetics in order of increasing potency and dividing the compounds into 10 activity bins (Table 1). One compound was randomly picked from each bin and added to the test-set. This procedure ensured that the test-set compounds represented the full range of MAC values exhibited by the anesthetics.

Full details of the modeling procedures have been described elsewhere (6,7). In brief, computer-based models of the anesthetics in the gas phase were constructed by using the molecular modeling software SYBYL 6.7 (Tripos Inc., St. Louis, MO) on a Silicon Graphics O2 R10000 workstation. The structures were geometry-optimized by using molecular mechanics minimization with the default SYBYL force field (6). This process considers the molecules as a series of spheres (representing the atoms) connected by springs (representing atomic bonds). Potential energy functions defined in the force field describe the optimum bond lengths, bond angles, and torsion angles for the atoms of the molecule. These molecular features are adjusted during geometry optimization to minimize the total potential energy of the structure.

Most of the anesthetics are flexible structures that exist as an ensemble of interchangeable configurations called conformers. This flexibility was incorporated in the model by deriving a set of low-energy conformers for the anesthetics by using a random search procedure in SYBYL (6). The torsion angles of the molecules were randomly perturbed, and the resultant structures were subjected to full molecular mechanics geometry optimization. The optimized conformers with a potential energy within +4 kcal/mol of the lowest-energy conformer for a given anesthetic were retained. This limit ensured that only conformers with realistic geometries were used in the subsequent analyses. The process was repeated until each anesthetic had undergone 10,000 random perturbations or until each conformer had been found at least 12 times. The geometries of the conformers were further optimized by using quantum mechanics, in which a mathematical description of molecular structure is formed in terms of the nuclei and electron distribution. This provides a more accurate representation of molecular geometry but is computationally more intensive. The computation time was reduced by using semiempirical quantum mechanics, in which only the valence electrons are explicitly considered and in which experimentally derived variables are used to represent the nuclei and inner-shell electrons. Geometry optimization was performed with MOPAC 6 (Quantum Chemistry Program Exchange, IN) with the AM1 Hamiltonian. Atomic partial charges were assigned by using the Coulson method (6). After optimization, duplicate conformers (defined as conformers with a root-mean-square difference of <0.2 Å) were removed. A total of 625 conformers were retained for the 38 anesthetics.

The chemical diversity of the anesthetics precludes their alignment by a common substructure. Suitable alignments were obtained by using an unbiased molecular similarity approach (15) based on the local minimum method (16,17). In this process, the compounds are oriented to maximize their molecular similarity to the most potent drug in the group, the lead structure. Inspection of the potency data showed that the most active anesthetic was hexanol, which had 66 conformers. Each conformer of the lead structure was used as a separate alignment template.

Molecular similarity was calculated as combined shape and electrostatic potential Carbo indices, which range from 0 (totally dissimilar shapes and electrostatic potentials) to 1 (totally identical). The anesthetic conformers were prealigned by weighted molecular extent and atomic partial charge (by using the default weighting ratio of 1:10) before being translated and rotated in a rigid search (30° increments) with Simplex optimization to maximize their molecular similarity to the lead structure conformer acting as the alignment template. Carbo indices were calculated with an analytical method (18) by using ASP 3.22 software (Automated Similarity Package; Accelrys Inc., Cambridge, UK). The anesthetic conformers and alignments with the maximum similarity to the lead structure conformer were retained (7). The process was repeated until all the conformers of the lead structure had been used as the alignment template, producing 66 different sets of alignments for the anesthetics.

Separate CoMFA activity models were formulated for each alignment set by using the standard SYBYL variables for steric and electrostatic fields (see Ref. 19 for a description of these variables and their significance). The aligned anesthetics were placed in a rectangular grid consisting of lattice points at 1-Å intervals. This grid interval provides a good compromise between accuracy and the possible introduction of noise from sampling irrelevant data. The grid extended at least 4 Å beyond the surface of all the molecules and consisted of 3600 lattice points. An sp3 carbon probe atom with a unitary positive charge was placed at each lattice point, and the steric and electrostatic interaction energies between the probe and the anesthetic molecules were calculated (5,19). Steric energies were calculated by using Lennard-Jones potentials, which describe both the attraction between molecules due to van der Waals forces (dispersion, dipole-induced dipole, and dipole-dipole interactions) and the repulsion as a result of steric clashes. The electrostatic energies were calculated by using Coulomb potentials with a distance-dependent dielectric function. Cutoffs were applied to both the steric and electrostatic energies at 30 kcal/mol (5,19).

The interaction energies at each lattice point were block-scaled to unit variance and correlated with in vivo potency to formulate activity models. Because of the large number of variables produced (3600 steric and 3600 electrostatic) and their colinearity, partial least squares (PLS) regression was used for this purpose (20). This process derives one or more orthogonal components based on a weighted combination of the interaction energies at each lattice point. The weightings are adjusted so that each component explains as much covariance as possible. A regression-like activity model is formulated by using the orthogonal components as independent variables and anesthetic activity as the dependent variable. The number of orthogonal components used was determined by leave-one-out cross-validation (see below), with each additional component having to increase the value of the cross-validated r2 by >0.05 to be included (15).

The intrinsic predictive power of the activity models was assessed by using leave-one-out cross-validation (21). In this process, the model was repeatedly reformulated, but one of the training-set anesthetics was excluded at each stage. The revised model was used to predict the potency of the excluded anesthetic, and the process was repeated until all of the anesthetics had been excluded once and once only. The CoMFA model with the greatest cross-validated r2 was retained as the final model. The extrinsic predictive power of this final model was evaluated by predicting the potencies of the randomly excluded test-set anesthetics.

The possibility of a chance correlation occurring was evaluated by randomly reassigning the observed anesthetic potencies of the training-set anesthetics and repeating the CoMFA analysis. A total of 100 random perturbation cycles were used, and the intrinsic and extrinsic predictive powers of the distorted data sets were evaluated.


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Discussion
 References
 
The anesthetic structures were aligned to maximize their molecular similarity to conformers of the most potent anesthetic in the group: hexanol. A total of 66 individual conformers of hexanol were identified, and each hexanol conformer was used as a separate alignment template for the remaining anesthetics. All conformers were given equal weighting in the alignment process. Thus, 66 different sets of anesthetic alignments were obtained, for which 66 individual CoMFA models were derived. The CoMFA model with the greatest predictive capability (assessed with cross-validation) was retained as the final activity model. Files containing details of the molecular alignments and conformers used in this final model are available from the authors on request.

An equivalent procedure for halogenated volatile anesthetics (22) showed that their electrostatic potentials differ significantly from that of the lead structure, hexanol (Table 2). The Carbo electrostatic similarity index was only 0.178 ± 0.019 (mean ± SEM) (n = 58) for the halogenated anesthetics, compared with 0.682 ± 0.023 (n = 38) for nonhalogenated anesthetics. Hence, hexanol is not a suitable alignment template for halogenated anesthetics, and these drugs will be considered elsewhere.


View this table:
[in this window]
[in a new window]
  		Table 2. Molecular Similarity of Nonhalogenated and Halogenated Volatile Anesthetics to Hexanol
 
The anesthetic potencies derived from the literature for the 28 training-set and 10 test-set compounds are shown in Table 1. The final CoMFA model was based on 2 orthogonal components and explained 95.5% of the observed variance in the activities of the 28 training-set anesthetics (F2,25 = 263.529; P < 0.0001). The correlation between the observed anesthetic potency and the activities predicted with the final model is shown in Figure 1. It can be seen that the model is a good predictor of activity over the full range of anesthetic potencies. The mean ± SEM of the residuals [the absolute difference between the –log10(predicted) and –log10(observed) MACs] for the training-set was 0.230 ± 0.029, with 46% of the compounds having residuals <0.2 and 71% of the compounds having residuals <0.3. The model is capable of handling stereoisomerism and correctly predicted that the cis isomer of piperylene (Compound 20) was more potent than the trans isomer (Compound 21). However, notably weak predictions were obtained for two anesthetics: cyclopropane (Compound 16) and ethanol (Compound 4).



View larger version (35K):
[in this window]
[in a new window]
  		Figure 1. Correlation between observed minimum alveolar anesthetic concentration (MAC) and values predicted with the comparative molecular field analysis model for the training-set anesthetics. The numbers refer to the compounds listed in Table 1. The model explained 95.5% of the variance in the observed activities of the 28 training-set drugs (F2,25 = 263.529; P < 0.0001). Note that the model correctly predicted the potency order of the cis and trans piperylene isomers (Compounds 20 and 21). Weaker predictions were obtained for cyclopropane (Compound 16) and ethanol (Compound 4).

 
The CoMFA model possessed good intrinsic predictive power for the training-set anesthetics, with a cross-validated r2 of 0.824. However, a more rigorous test of predictive capability is to use the CoMFA model to predict the activities of the 10 randomly excluded test-set anesthetics (Compounds 29–38). There is a good correlation (r2 = 0.921) between observed and predicted anesthetic potency for these compounds (Fig. 2). The mean ± SEM residual for the test-set was 0.238 ± 0.066 (n = 10), with 50% of the compounds having residuals <0.2 and 60% having residuals <0.3. The activity model also correctly predicted that 2-cis-butene (Compound 35) was more potent than the trans isomer in the training-set (Compound 22). Poor predictions of anesthetic activity were obtained for acetylene (Compound 36) and 3-hexyne (Compound 33).



View larger version (28K):
[in this window]
[in a new window]
  		Figure 2. Correlation between observed minimum alveolar anesthetic concentration (MAC) and values predicted with the comparative molecular field analysis model for the 10 test-set anesthetics. The model was a good predictor for the randomly excluded anesthetics (r2 = 0.921) and correctly predicted that 2-cis-butene (Compound 35) was more potent than the trans isomer in the training-set (Compound 22). Weaker predictions were obtained for acetylene (Compound 36) and 3-hexyne (Compound 33).

 
The possibility of obtaining a chance correlation was tested by randomly perturbing the potency data and repeating the CoMFA analyses for the final alignment set. The activity models for the perturbed data sets exhibited no intrinsic predictive power for the training-set anesthetics: the mean ± SEM cross-validated r2 was negative, at –0.271 ± 0.027 (n = 100). The perturbed models were also poor predictors of activity for the excluded test-set anesthetics, with a mean correlation coefficient (r) ± SEM of –0.057 ± 0.038 (n = 100) for the relationship between observed and predicted activities.

For comparison, we also determined the effectiveness of a conventional activity model based on nonpolar solubility (olive oil/gas partition coefficients). This model explained only 73.2% of the variance in the observed activities of the training-set compounds (F1,26 = 71.138; P < 0.0001), with mean ± SEM residuals of 0.494 ± 0.087 (n = 28). The predictive capability of the nonpolar solubility model was also less than that of the CoMFA model for both the training-set (cross-validated r2 = 0.693) and test-set (r2 = 0.705) anesthetics.

The relative contributions of the electrostatic and steric interactions to the activity model were 74.4% and 25.6%, respectively. Analysis of the individual PLS weightings for each grid point within the orthogonal components of the model enables the identification of regions where steric and electrostatic interaction energies are important in determining activity. Pharmacophoric maps showing the spatial distribution of these regions can be derived by using isocontour plots. These plots were obtained by linking together lattice points in the CoMFA grid where the standard deviation of the interaction energies multiplied by the PLS weighting coefficients at that point (SD x coeff) exceed a certain value. Hence, the plots indicate the regions where the differences in either steric or electrostatic interaction energies are strongly associated with changes in anesthetic potency. We have used the same strategy for assigning the isocontour thresholds as described in our article for the IV anesthetics (7), linking lattice points that represent the greatest 40% of the individual positive and negative contributions to the activity model.

The pharmacophores derived from the final CoMFA model are shown in Figure 3. It can be seen from the electrostatic map (Fig. 3a) that there are two areas (A and B) where positive electrostatic potential is favored for high anesthetic potency (SD x coeff > +0.004) and two main areas where negative potential is favored (C and D; SD x coeff < –0.020). The positioning of these areas in relation to the lead compound hexanol is shown in Figure 3b. The arrows indicate areas where the electrostatic potential of the molecule qualitatively fits the pharmacophoric map. Thus, the electronegative oxygen atom of the hydroxyl group aligns with the red negative potential favored region (D), and most of the electropositive hydrogen atoms align with the blue positive potential favored regions. However, it is evident that there are areas where the electrostatic potential of hexanol does not match the pharmacophoric template. These regions, marked with crosses, correspond with the alignment of the hydrogen atoms with the negative potential favored area (C). On this basis, we speculate that substitution of these hydrogen atoms with more electronegative substituents (e.g., fluorine atoms) would result in a molecule with increased potency.



View larger version (71K):
[in this window]
[in a new window]
  		Figure 3. Pharmacophoric maps derived with comparative molecular field analysis. a, Spatial arrangement of the key areas where positive (blue; SD x coeff > +0.004) and negative (red; SD x coeff < –0.020) electrostatic potential are important for high anesthetic potency. b, Orientation of hexanol (the most potent anesthetic in the group) with the electrostatic pharmacophoric map. The arrows indicate areas where the electrostatic potential of the molecule qualitatively fits the template. The crosses signify regions where there is a discrepancy. Substitution of the hydrogen atoms with more electronegative components (such as fluorine) in these areas may lead to increased anesthetic activity. The steric pharmacophoric map is shown in (c). Regions where molecular bulk is favored (green; SD x coeff > +0.0049) and disfavored (magenta; SD x coeff < –0.0016) for high anesthetic potency are shown. The bulk favored regions form an area that encloses a molecule the size of cycloheptane (d). +ve = positive; –ve = negative.

 
The steric pharmacophoric map is shown in Figure 3c. The orientation and scaling are the same as for the electrostatic equivalent. There are four zones (E to H) colored green where molecular bulk is favored for high anesthetic potency (SD x coeff > +0.0049) and two regions (I and J) colored magenta where the presence of molecular bulk is disfavored (SD x coeff < –0.0016). If the molecule extends into one of the disfavored regions, anesthetic potency is reduced. Note that the arrangement of the bulk favored regions encloses a molecule such as cycloheptane (Fig. 3d), which is the largest of the cycloalkanes that exhibits immobilizing activity.


   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
References
 
The results of this study demonstrate that a single CoMFA activity model with good predictive capability can be formulated for nonhalogenated inhaled anesthetics, despite their structural and chemical diversity. We have included all of the nonhalogenated anesthetics for which we had comparable potency data and that were compatible with the modeling techniques used. Six nonhalogenated anesthetics were omitted: octanol, argon, krypton, xenon, nitrogen, and nitrous oxide. Octanol, a potential lead structure, was excluded because it exhibits a high degree of molecular flexibility and has an unmanageable number of conformers. Inert gases and nitrogen were omitted because they do not exhibit molecular electrostatic potentials and so cannot be aligned by using this property (only compounds that contain at least two atoms with different electron affinities can be tested with this procedure). Nitrous oxide was excluded because it is a resonant structure, interchanging between the N = N+ — O– and N– = N+ = O forms. An ab initio quantum mechanics approach (in which all the electrons of the molecule are explicitly considered) is required to adequately model the electrostatic properties of this anesthetic (23).

Halogenated volatile anesthetics were not considered because the electrostatic profiles of these drugs differ substantially from that of the lead structure, hexanol. The Carbo electrostatic similarity is only 0.178 ± 0.019 (mean ± SEM) (n = 58), compared with 0.682 ± 0.023 (n = 38) for nonhalogenated anesthetics. Hence, it is not possible to derive a suitable alignment for halogenated anesthetics by using hexanol as a template (22). Although this practical difficulty necessitates the separate modeling of nonhalogenated and halogenated anesthetics, it does not imply that the molecular basis of their immobilizing activity is different. This question can be addressed only when the independently derived pharmacophores for the two groups of anesthetics are quantitatively compared. Although the putative sites of anesthetic action are not considered in our ligand-oriented model at this stage, it is interesting to note that different receptor specificities have been proposed for nonhalogenated and halogenated volatile anesthetics (24).

How valid is the modeling approach for such a group of structurally diverse anesthetics? The predicted –log(MAC) values for 71% of the training-set anesthetics and 60% of the test-set anesthetics were within 0.3 of the observed values. This error is consistent with other published CoMFA models for structurally diverse systems (17,25) and a 4D-QSAR model for structurally homologous propofol analogs (26). Furthermore, the model is a good predictor of potency for unknown excluded test-set anesthetics (r2 = 0.921). It is unlikely that the predictive power of the model is due to a chance correlation, because random perturbation of the potency data resulted in models with no predictive capability. However, the CoMFA model poorly predicts the activities of some anesthetics—notably, the smaller molecules such as ethanol, acetylene, and cyclopropane. There may be some benefit in deriving a separate activity model for these compounds.

In general, the error associated with MAC estimations is <10%. Our activity model is not yet able to predict potency at this level of accuracy, hence its preliminary designation. The weaker predictions for some anesthetics are most likely due to an inappropriate alignment within the CoMFA lattice grid. The anesthetics were aligned by molecular similarity because of their structural diversity. However, this similarity was calculated by using whole-molecule Carbo indices based on an average of the shape and electrostatic potential similarities. This process assumes not only that the steric and electrostatic interactions are of equal importance in determining activity, but also that all regions of the molecules are of significance. However, our results indicate that the electrostatic interactions make a threefold greater contribution to the CoMFA model compared with steric interactions and that not all of the steric and electrostatic features of the anesthetic molecules are important for activity. We anticipate that improved MAC predictions will be obtained by using the preliminary pharmacophores presented in this article as alignment templates for the next-generation model.

Predictive capability might also be improved by allowing the geometries of the molecules to be flexible in the alignment process (thereby allowing a better fit to either the lead structure or pharmacophore templates) or by incorporating additional molecular properties in the activity model. Of particular interest would be the inclusion of an explicit polarizability field. This would enable anesthetics that do not exhibit molecular electrostatic potentials (nitrogen and inert gases) to be incorporated into the model. However, calculation of polarizability fields requires ab initio quantum mechanics, which is computationally intensive and would be impractical for the 625 anesthetic conformers considered in this study.

What does this study contribute to our understanding of the molecular mechanisms of volatile anesthetic action? Our results for the nonhalogenated anesthetics support the findings of previous studies that have shown that steric and electrostatic interactions are important in determining anesthetic activity (27,28). We have now developed this concept and derived a CoMFA model that has characterized the molecular basis of immobilizing activity in terms of the magnitude and spatial distribution of these steric and electrostatic interactions. The pharmacophoric maps illustrate the three-dimensional relationship of the key regions where such interactions are important in determining MAC. The predictive capability of the activity model demonstrates that these maps are applicable to a structurally diverse range of nonhalogenated volatile anesthetics. This raises the possibility that there might be a common molecular basis for the immobilizing activity of the anesthetics considered. However, it is important to note that a common molecular basis for immobilizing activity does not imply a common site of action. Further studies are required to establish whether the molecular basis of anesthetic specificity for various receptor systems (determined by using CoMFA with receptor-binding data as dependent variables) differs from the molecular basis of the immobilizing activity we have described.


   Acknowledgments
 
Supported in part by a British Journal of Anaesthesia project grant (JWS) and by departmental funds.

We thank Edmond I Eger II and his colleagues at the University of California, San Francisco, for the anesthetic potency data and for helpful discussions concerning our modeling approach.


   Footnotes
 
Presented in part at the annual meeting of the American Society of Anesthesiologists, Orlando, FL, October 2002 (Anesthesiology 2002;96:A90).


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 

  1. Sonner JM, Antognini JF, Dutton RC, et al. Inhaled anesthetics and immobility: mechanisms, mysteries, and minimum alveolar anesthetic concentration. Anesth Analg 2003; 97: 718–40.[Abstract/Free Full Text]
  2. Miller KW. The nature of sites of general anaesthetic action. Br J Anaesth 2002; 89: 17–31.[Abstract/Free Full Text]
  3. Krasowski MD, Harrison NL. General anaesthetic actions on ligand-gated ion channels. Cell Mol Life Sci 1999; 55: 1278–303.[ISI][Medline]
  4. Yamakura T, Bertaccini E, Trudell JR, Harris RA. Anesthetics and ion channels: molecular models and sites of action. Annu Rev Pharmacol Toxicol 2001; 41: 23–51.[ISI][Medline]
  5. Kubinyi H. QSAR and 3D QSAR in drug design. I. Methodology. Drug Discov Today 1997; 2: 457–67.
  6. Sewell JC, Sear JW. Can molecular similarity-activity models for intravenous general anaesthetics help explain their mechanism of action? Br J Anaesth 2002; 88: 166–74.[Abstract/Free Full Text]
  7. Sewell JC, Sear JW. Derivation of preliminary three-dimensional pharmacophoric maps for chemically diverse intravenous general anaesthetics. Br J Anaesth 2004; 92: 45–53.[Abstract/Free Full Text]
  8. Zhang Y, Trudell JR, Mascia MP, et al. The anesthetic potencies of alkanethiols for rats: relevance to theories of narcosis. Anesth Analg 2000; 91: 1294–9.[Abstract/Free Full Text]
  9. Fang Z, Ionescu P, Chortkoff BS, et al. Anesthetic potencies of n-alkanols: results of additivity and solubility studies suggest a mechanism of action similar to that for conventional inhaled anesthetics. Anesth Analg 1997; 84: 1042–8.[Abstract]
  10. Liu J, Laster MJ, Taheri S, et al. Effect of n-alkane kinetics in rats on potency estimations and the Meyer-Overton hypothesis. Anesth Analg 1994; 79: 1049–55.[Abstract/Free Full Text]
  11. Eger EI II, Laster MJ. The effect of rigidity, shape, unsaturation, and length on the anesthetic potency of hydrocarbons. Anesth Analg 2001; 92: 1477–82.[Abstract/Free Full Text]
  12. Fang Z, Sonner J, Laster MJ, et al. Anesthetic and convulsant properties of aromatic compounds and cycloalkanes: implications for mechanisms of narcosis. Anesth Analg 1996; 83: 1097–104.[Abstract]
  13. Taheri S, Halsey MJ, Liu J, et al. What solvent best represents the site of action of inhaled anesthetics in humans, rats and dogs? Anesth Analg 1991; 72: 627–34.[Abstract/Free Full Text]
  14. Zhang Y, Eger EI II, Dutton RC, et al. Inhaled anesthetics have hyperalgesic effects at 0.1 minimum alveolar anesthetic concentration. Anesth Analg 2000: 91: 462–6.[Abstract/Free Full Text]
  15. Jewell NE, Turner DB, Willett P, Sexton GJ. Automatic generation of alignments for 3D QSAR analyses. J Mol Graph Model 2001; 20: 111–21.[Medline]
  16. Demeter DA, Weintraub HJR, Knittel JJ. The local minimum method (LMM) of pharmacophore determination: a protocol for predicting the bioactive conformations of small, conformationally flexible molecules. J Chem Inf Comput Sci 1998; 38: 1125–36.[ISI][Medline]
  17. Baurin N, Vangrevelinghe E, Morin-Allory L, et al. 3D-QSAR CoMFA study on imidazolinergic I2 ligands: a significant model through a combined exploration of structural diversity and methodology. J Med Chem 2000; 43: 1109–22.[ISI][Medline]
  18. Good AC, Richards WG. Explicit calculation of 3D molecular similarity. Perspect Drug Discov Design 1998; 9–11: 321–38.
  19. Cramer RD III, DePriest SA, Patterson DE, Hecht P. The developing practice of comparative molecular field analysis. In: Kubinyi H, ed. 3D QSAR in drug design: theory, methods and applications. Leiden: ESCOM, 1993: 443–85.
  20. Dunn WJ III, Wold S, Edlund U, et al. Multivariate structure-activity relationships between data from a battery of biological tests and an ensemble of structure descriptors: the PLS method. Quant Struct Act Relat 1984; 3: 131–7.
  21. Cramer RD III, Bunce JD, Patterson DE, Frank IE. Crossvalidation, boot strapping and partial least squares compared with multiple regression in conventional QSAR studies. Quant Struct Act Relat 1988; 7: 18–25.
  22. Sear JW, Sewell JC. A preliminary CoMFA activity model for fluorinated inhalational general anesthetics [abstract]. Anesthesiology 2002; 96: A117.
  23. Frisch MJ, Del Bene JE. The electronic structure and electrostatics of nitrous oxide. Int J Quantum Chem Symp 1989; 23: 363–9.
  24. Campagna JA, Miller KW, Forman SA. Mechanisms of actions of inhaled anesthetics. N Engl J Med 2003; 348: 2110–24.[Free Full Text]
  25. Nicolotti O, Pellegrini-Calace M, Altomare C, et al. Ligands of neuronal nicotinic acetylcholine receptor (nAChR): inferences from the Hansch and 3-D Quantitative Structure-Activity Relationship (QSAR) models. Curr Med Chem 2002; 9: 1–29.[Medline]
  26. Krasowski MD, Hong X, Hopfinger AJ, Harrison NL. 4D-QSAR analysis of a set of propofol analogues: mapping binding sites for an anesthetic phenol on the GABAA receptor. J Med Chem 2002; 45: 3210–21.[ISI][Medline]
  27. Sewell JC, Halsey MJ. Shape similarity indices are the best predictors of substituted fluoroethane and ether anaesthesia. Eur J Med Chem 1997; 32: 731–7.
  28. Raines DE, Claycomb RJ. The role of electrostatic interactions in governing anesthetic action on the Torpedo nicotinic acetylcholine receptor. Anesth Analg 2002; 95: 356–61.[Abstract/Free Full Text]
Accepted for publication April 5, 2004.




This article has been cited by other articles:


Home page
 Anesth. Analg.Home page
E. I. Eger II, D. E. Raines, S. L. Shafer, H. C. Hemmings Jr, and J. M. Sonner
Is a New Paradigm Needed to Explain How Inhaled Anesthetics Produce Immobility?
Anesth. Analg., September 1, 2008; 107(3): 832 - 848.
[Abstract] [Full Text] [PDF]

Home page
 Anesth. Analg.Home page
J. M. Sonner
A Hypothesis on the Origin and Evolution of the Response to Inhaled Anesthetics
Anesth. Analg., September 1, 2008; 107(3): 849 - 854.
[Abstract] [Full Text] [PDF]

Home page
 Anesth. Analg.Home page
J. G. Bovill
Anesthetic Pharmacology: Reflections of a Section Editor
Anesth. Analg., November 1, 2007; 105(5): 1186 - 1190.
[Full Text] [PDF]

Home page
 Anesth. Analg.Home page
J. C. Sewell and J. W. Sear
Determinants of volatile general anesthetic potency: a preliminary three-dimensional pharmacophore for halogenated anesthetics.
Anesth. Analg., March 1, 2006; 102(3): 764 - 771.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a colleague
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sewell, J. C.
Right arrow Articles by Sear, J. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sewell, J. C.
Right arrow Articles by Sear, J. W.
Related Collections
Right arrow Mechanisms
Right arrow Pharmacology

Lippincott, Williams & Wilkins Anesthesia & Analgesia® is published for the International Anesthesia Research Society® by Lippincott Williams & Wilkins with the assistance of Stanford University Libraries' HighWire Press®. Copyright 2006 by the International Anesthesia Research Society. Online ISSN: 1526-7598   Print ISSN: 0003-2999 HighWire Press