September 1969
MO CALCULATIONS OF PHENYL CHOLINE ETHERS
Hey2* felt that the presence of a partial positive charge on the ether 0 (11) could explain his observations. On the other hand, Ormerod3 and Sekul and Holland4 suggested that a partial negative charge a t an appropriate distance from the onium head was essential for nicotinic activity. This negative charge was assumed to be in a position analogous to the partial negative charge assigned to the carbonyl oxygen in acetylcholine (111). I n support of this latter conclusion is a molecular orbital study of the preferred conformation of nicotine by Iiier.6 He concluded from this study that two principal atoms necessary for nicotinic activity in the nicotine molecule are a positively charged nitmgen atom and a negatively charged atom about 4.85 A removed. He further stated that its interaction with the receptor is similar to that of the carbonyl 0 of acetylcholine. Alolecular orbital calculations of Fukui, et al.,'j on Hey's ethers showed no correlation between charge density on the ether oxygen and pharmacologic activity. However, they did find a good parallelism between the frontier electron density at the ether oxygen and superdelocalizability at the ortho ring positions and nicotinic activity. Later Coleman, et U Z . , ~ studied ~ a new series of phenyl choline ethers exhibiting nicotinic activity. They attributed their findings to an iriductomeric or polarizability effect of the substituent on the ring T electrons. They suggested that the partially negatively polarized rings provide a secondary binding feature in the molecule, presumably comparable to the easily polarized carbonyl oxygen of acetylcholine. To further clarify the matter we have nom completed molecular orbital calculations on the ethers prepared and assayed by Coleman, et These calculations are summarized and discussed in the present communication.
765
TABLE I SEMIEMPlRICAL PAR.4METERS USED I N T H E HUCKEL Bond or a t o m
C (aromatic) C-N (amino)
C-N (aromatic) C-0 (ether)
Coulomb integral
hc
Resonance integral
0.0
=
hN = 1 . 5
hc h~ hc ho hc hF hc
c-C1
C-Br
G I /-N
1
-I
(-0-
No. of electrons
kc-c = 1.0 = 1.0
1
kc-h.
kC-N
= 0.0
0.4
=
= 0.0 = 2.0
0.0 = 3.0 = 0.2 hci = 2 . 0 hc = 0.1*5 hBr = 0.9 hc = 0 . 1
C -F
C-NOs
CALCULATIONS
=
1.0
1
kc-0 =
0.8
2
kc-F
=
0.7
2
=
0.4
2
0.3
2
=
h l
=
Irc hshc ho-
= = = =
0.5 0.0.5 0.6 0.0 1.6
kc-ci
kLBr =
kr-1
=
0.2
2
kC-N
=
1.0
1
kN+-o
=
1.0
1
exception of methylated onium head: free valence, iiet charge and superdelocalizability, and the energies of the highest filled (HOllO) and the lowest empty molecular orbital (LEMO). The nicotine-like activities of the ethers employed were taken from the data of Coleman, et al.2b These activities were obtained from dose-response curves and are expressed as relative ones on a molar basis with phenyl choline ether set equal to one. Values greater than 1 repre5ent enhanced nicotine-like activity.
Results
Molecular orbital calculations were made for a series of meta-substituted-phenyl choline ethers; their structures and relative nicotine-like activities are given in Table I . Since no correlation between free valence, T charge density, and LEAIO mere noted, these indices will not be considered. Methods I n our calculations, superdelocalizability at ring posiAll the calculatioiis were in the simple Huckel approxi~iialioti.~ tions 2 and 6 and the energy of the highest occupied 111the simple Huckel method, relative values in a series of closely molecular orbital showed parallelism with pharmacorelated molecules, such as the ones studied in t,his report, are logic activity. Superdelocalizability is a measure of the more significant than absolute values. The method has been ability of atoms in a molecule to form a weak T bond very successful in dealing with unsaturated systems. For this with an appropriate group in the receptor. The energy reason, we have not used the extended Huckel theory.6 This latter procedure takes into consideration both u and T electrons. of the highest occupied molecular orbital is a relative I t has had its greatest success in calculating the preferred conmeasure of the ability of an electron in the highest ocformations of hydrocarbons, both aliphatic and aromatic, as cupied orbital of a compound to be transferred to an acwell as predicting conformational energies. These parameters ceptor molecule. I n such studies molecular orbital would be of little value in the present study because the strucenergies are represented in p units (0is the resonance intures of the compounds examined are very similar. The semiempirical parameters are those recommended by tegral). The smaller the coefficient of p, the greater is Kriiger-Thiemer and Hansen.8 Theirs is a recent and extensive the electron-donating property of the molecule within a collection of parameters and, in addition, they take into conseries of closely related compounds. sideration the effects of the substituents on the aromatic carbon I n Table 11, we have also included assays on pyridyl atom. The parameters used are given in Table I. The dciilations were made rihig an IBRI system 360 ltodel choline ether and pyridylethyltrimethylammonium. 40 at the lIissisippi Research and 1)evelopmeiit Center, Jacksoil, Here again the correlatiori between nicotine-like stimulliss. All calculat,ions were doiie i n double-precision arithmetic lant action and the energy of highest occupied molecular with input' values having three significant figures. An I B l l orbital (HORIO) is reasonably good. P L / I program was employed. The following indices were calculated for each atom, with the JV. E. Ormerod, Brit. J . Pharmacol., 11, 267 (1956). (1) A. A . Sekul a n d W.C. Holland. J . Pharmacol. Ezptl. Therap., 132, 171 (3)
(1961); 133, 313 (1961). 15) L. B.Bier, M o l . P h ~ r m ~ c o 4, l . ,70 (1968). (8) F.Fiikui, ('. N a g a t a , a n d A. I m a m u r a . Science, 132, 8 i (1960). (7) .I. Streitwieser, J r . , "AIolecular Orbital Theory for Organic Chemi a t s , " .Jolin JViley a n d Suns, Inc., N e w York, N. Y., 1961. (8) E. Kriiger-Tliiemer a n d R . Hansen, Arzneimittel-Forsch., 16, 1453 (1966).
Discussion I n proposing a mechanism of binding one can consider a number of possible interactioiis hetweeii the ethers arid the receptor. It is now well linowii that onium head is essential for nicotine-like activity. The remaining part of the molecule contributes to binding at a neA?rbysite to
increwhe or decrease the intensity of activity. If such he the case, what is the nature of the intermolecular forces involved? The ether could sterically approxim i t e the receptor. be bound by electrostatic attraction, form H bonds, enter into complex formation, or act :IS ari electron donor or acceptor. While steric factors are frequently important, they cannot explain the observed difference in activity of these compounds since they have a constant spatial disposition. The lack of correlation with 8-charge distribution speaks against electro-
static interactioti. Iurthermore, if ;in elect roii-tr:iiisfvr mechanism is involved, the negative correlation with the energy of the lowest empty molecular orbital indicate. that the ethers do not :ict z iu electron :iccept:tiicr. Thus;. despite the crudenes:, of the theoretical (lata discussed here, the close relationship hetweeii HOAIO :ind superdelocalizability of atonis a t ihe 2 arid 6 ring po4tions arid nicotine-like activity iuggests that aromatic ring interacts with a secondary groiip(s) in the rwcptot hy formation of a charge-trmsfcr complex.
Comparison of Parameters Currently Vsed in t h e Study of Structure-Activity Relationships'
The reactivities of a large yroiip of ~iiiscellaiieousmolec:iile-, it< ineasured i i i f
