Conformational analysis of the dopamine-receptor agonist 5-hydroxy

Chem. , 1986, 29 (6), pp 917–924. DOI: 10.1021/jm00156a008. Publication Date: June 1986. ACS Legacy Archive. Cite this:J. Med. Chem. 29, 6, 917-924...
0 downloads 0 Views 2MB Size
J . Med. Chem. 1986,29,917-924 with minor modifications as described in ref 21.

Acknowledgment. We thank A. Mosimann and R. Labhart for their skillful technical assistance and our collegues and their collaborators in the Physical Chemistry Department and the Analytical Laboratory for their help in measuring and interpreting the spectra and performing the elemental analysis, respectively. We are further grateful to Dr. F. Braunschweiger and F. Seemann for the supply of ample quantities of intermediates. Registry NO.(*)-l, 69367-50-6; (&)-2,101403-04-7;(*)-2*HCl, 78598-56-8; (*)-3, 101403-05-8; (-)-(S)-3,88322-07-0; (+)-(R)-3, 101540-25-4;(*)-3*HC1,78598-52-4;(i)-4,101403-06-9;(*)-4*HC1, 78598-54-6; (*)-5, 78615-30-2; (*)-5~'/~nd,78615-31-3; ( i ) - 6 , 78598-46-6; (*)-6*HC1,78598-48-8;(*)-7,99755-60-9; (*)-7*HCl, 101403-00-3; (&)-8,101403-07-0; (*)-8*HC1, 101403-01-4; ( i ) - 9 , 101403-08-1;(*)-9*HCl,78598-61-5; (*)-lo, 101403-09-2;(*)-lo (5-mesylate), 101403-16-1; (f )- 10.HC1, 78598-62-6; (*)-l1, (20) Creese, I.; Schneider, R.; Snyder, S. H. Eur. J . Pharmacol. 1977, 46, 377. (21) Closse, A.; Frick, W.; Dravid, A,; Bolliger, G.; Hauser, D.; Sauter, A,; Tobler, H. J. Naunyn-Schmiedeberg's Arch. Pharm. 1984,327,95.

917

78598-65-9; (f)-llJ/2nd, 78598-66-0; (*)-12,7859847-7; (-)-(S)-l2, 7tmatx-3;(rt)-it.fu, m98-51-3; (f)-i3,ioi403-10-5; (*)-I~.HC~, 78598-63-7; (f)-14, 78598-58-0; (f)-15, 78598-59-1; (i)-15 (dimesylate), 101403-15-0; (f)-16, 101403-02-5; (f)-17, 78598-83-1; (f)-18,78598-79-5;(f)-18.1/2nd, 78598-80-8; (f)-19,101403-11-6; (&)-19*HCl,78598-53-5; (&)-20,101403-12-7;(&)-2O*HCl,7859878-4; (f)-21, 78598-72-8; (*)-21*pm, 101403-03-6; (&)-22, 101403-13-8; (*)-22*HCl,78598-70-6; (i)-23, 101403-14-9; (f)23.HC1,78598-71-7; (*)-I11 (R = Pr), 78598-91-1; (-)-(5')-I11 (R = CH2C6H5), 58349-23-8; (-)-(S)-III (R = Pr), 101403-24-1; (+)-(R)-III(R = Pr), 101403-25-2;(f)-IV (R = Pr), 78598-89-7; ( i ) - I V (R = Et), 101403-19-4; ( i ) - I V (R = Me), 101403-20-7; (-)-(S)-IV (R = Pr), 101470-23-9;(+)-(R)-IV (R = Pr), 10147024-0; (+V (R = Pr, X = NHz, n = 2), 101403-17-2;(*)-V (R = Pr, X = NH2, n = 3), 101403-18-3; ( i ) - V (R = Pr, X = NHS02NEt2,n = 2), 101403-21-8; (*)-V (R = Pr, X = SMe, n = 3), 78598-90-0; (+V (R = Pr, X = OH, n = 3)~'/~nd, 78598-886; ( i ) - V (R = Pr, X = C1, n = 3), 78598-49-9; ( i ) - V (R = Pr, X = C1, n = 3)e1/~nd,78598-50-2; (h)-V (R = Pr, X = S02Me, n = 3), 101403-22-9; (-)-(S)-V (R = C H ~ C ~ HXS ,= CH3), 101403-23-0; (-)-(S)-V (R = Pr, X = OH, n = 3), 101470-25-1;(-)-(S)-V (R = Pr, X = C1, n = 3), 101470-26-2; (-)-(S)-V (R = Pr, X = SMe, n = 3), 101470-27-3; C6H,CH20CONHCH2C02H,1138-80-3; C6H&H20CONHCH2CONH2,949-90-6; H2NCH&H2NH2,10715-3; Br(CH2)3CN,5332-06-9; I(CH2)30H,627-32-7.

Conformational Analysis of the Dopamine-Receptor Agonist 5-Hydroxy-2-(dipropylamino)tetralinand Its C(2)-Methyl-SubstitutedDerivative Anders KarlBn,+Anette M. Johansson,? Lennart Kenne,' Lars Erik Arvidsson,' and Uli Hacksell*' Department of Organic Pharmaceutical Chemistry, Uppsala Biomedical Center, University of Uppsala, S-751 23 Uppsala, Sweden, Department of Analytical Chemistry, KabiVitrum, S-112 87 Stockholm, Sweden. Received August 28, 1985 The conformational preferences of the dopamine (DA) receptor agonist 5-hydroxy-2-(di-n-propylamino)tetralin (1) and the DA-inactive 5-hydroxy-2-methyl-2-(di-n-propylamino)tetralin (2) have been studied by use of molecular mechanics (MMP2) calculations and NMR spectroscopy. A good agreement is demonstrated between the experimentally determined (by NMR) and the calculated (by MMP2) conformationaldistribution of 1 and 2. In addition, there is a good agreement between bond distances and bond angles in the X-ray structure of the hydrobromide of 1and those in the correspondingMMP2 conformation. Results obtained demonstrate that the energetically preferred conformations of 1 and 2 are different: Compound 1 preferentially adopts half-chair conformations with a pseudoequatorial nitrogen substituent whereas the low-energy conformations of compound 2 have a pseudoaxial nitrogen substituent. However, the results also indicate that the difference in conformational preferences is too small to account for the dopaminergic inactivity of 2. Therefore it is suggested that the steric bulk of the C(2)-methylgroup per se prevents a proper alignment of (2S)-2 with DA receptors.

5-Hydroxy-2-(di-n-propylamino)tetralin (I)' is a wellestablished dopamine (DA) receptor agonist in vivolab and in vitro.2 Due to its potency and selectivity for DA receptors, compound 1 has served as the lead compound in several structure-activity relationship (SARJ ~ t u d i e s . As ~ part of an ongoing investigation of the effects of introduction of methyl substituents in the nonaromatic ring of 1,4 we synthesized and tested racemic 5-hydroxy-2methyl-2-(di-n-propylamino)tetralin (2),5 the C(2)methyl-substituted derivative of 1. Compound 2 exhibits a complex pharmacological p r ~ f i l e :(a) ~ It reverses reserpine-induced akinesia, but this effect is not blocked by pretreatment with the DA-receptor antagonist haloperidol. (b) It increases the synthesis rate of 5-hydroxytryptamine but does not affect that of DA. (c) I t raises the body temperature in rats. Notably, neither racemic 2 nor the enantiomers of 26 appear to act on DA receptors. Thus, t

Department of Organic Pharmaceutical Chemistry. Department of Analytical Chemistry. 0022-2623/86/1829-0917$01.50/0

the introduction of a C(2)-methyl substituent in 1 completely changes the pharmacological profile.

2 1 ~

(1) (a) McDermed, J. D.; McKenzie, G. M.; Phillips, A. P. J . Med. Chem. 1975,18,362. The dopaminergicactivity resides almost entirely in the 2S4-1 enantiomer of 1: (b) McDermed, J. D.; McKenzie, G. M.; Freeman, H. S. Zbid. 1976,19,547 (synthesis). (c) Giesecke, J. Acta Crystallogr., Sect. E 1980, 36, 110 (X-ray crystallography). (2) (a) Tedesco, J. L.; Seeman, P.; McDermed, J. D. Mol. Pharmacol. 1979, 16, 369. (b) Seiler, M. P.; Markstein, R. Zbid. 1982,22, 281. (c) Zbid. 1984, 26, 452. 0 1986 American Chemical Society

918 Journal of Medicinal Chemistry, 1986, Vol. 29, No. 6

l

Karldn et al.

h

3 A

A

A

-3

A

-*

t

Figure 1. Plot illustrating the relationship between the tetralin inversion angle 4 and key intramolecular distancese in a (2S)-2amino-5-hydroxytetralin derivative. Distances are calculated from selected conformations along the tetralin inversion path. Arc is the center of the aromatic ring and ArP is the plane of the aromatic ring.

One or several of at least three conformational or steric differences between 1 and 2 might be responsible for the DA inactivity of 2: (a) the steric bulk of the C(B)-methyl group may prevent a proper aligqment of 2 with DA receptors; (b) the solution conformations of the nonaromatic ring of 2 may be different from that of 1; (c) actual lowenergy conformations of the dipropylammonium substituents of 1 and 2 may be different. In order to find out if any of these factors is more likely to dominate, we have now studied the conformational NMR specpreferences of 1 and 2 by use of 'H and troscopy and extensive molecular mechanics (MMP2) calculations. During the course of our work, Nichols et ala7 (3) See, for example: (a) Hacksell, U.; Svensson, U.; Nilsson, J. L. G.; Hjorth, S.; Carlsson, A,; Wikstrom, H.; Lindberg, P.; Sanchez, D. J. Med. Chem. 1979,22,1469. (b)Feenstra, M. G . P.; Rollema, H.; Dijkstra, D.; Grol, C. J.; Horn, A. S.; Westerink, B. H. C. Naunyn-Schmiedeberg's Arch. Pharmacol. 1980,313, 213. (c) Hacksell, U.; Arvidsson, L.-E.; Svensson, U.; Nilsson, J. L. G.; Wikstrom, H.; Lindberg, P.; Sanchez, D.; Hjorth, S.; Carlsson, A,; Paalzow, L. J. Med. Chem. 1981, 24, 429. (d) Arvidsson, L.-E.; Hacksell, U.; Nilsson, J. L. G.; Hjorth, S.; Carlsson, A.; Lindberg, P.; Sanchez, D.; Wikstrom, H. Ibid. 1981, 24, 923. (4) (a) Hacksell, U.; Johansson, A. M.; Arvidsson, L.-E.; Nilsson, J. L. G.; Hjorth, S.; Carlsson, A,; Wikstrom, H.; Sanchez, D.; Lindberg, P. J. Med. Chem. 1984,27, 1003. (b) Johansson, A. M.; Arvidsson, L.-E.; Hacksell, U.; Nilsson, J. L. G.; Svensson, K.; Hjorth, S.; Clark, D.; Carlsson, A.; Sanchez, D.; Andersson, 3.;Wikstrom, H. Ibid. 1985,28,1049. (c) Svensson, K.; Hjorth, S.; Clark, D.; Carlsson, A.; Wikstrom, H.; Andersson, B.; Sanchez, D.; Johansson, A. M.; Arvidsson, L.-E.; Hacksell, U.; Nilsson, J. L. G. J. Neural Transm. 1986,85, 1. ( 5 ) Hacksell, U.; Arvidsson, L.-E.; Johansson, A. M.; Nilsson, J. L. G.; Sanchez, D.; Andersson, 3.;Lindberg, P.; Wikstrom, H.; Hjorth, S.; Svensson, K.; Carlsson, A. Acta Pharm. Suec. 1985, 22, 65. (6) Andersson, B.; Sanchez, D.; Lindberg, P.; Wikstrom, H.; Svensson, K.; Hjorth, S.; Carlsson, A.; Arvidsson, L.-E.; Johansson, A. M.; Hacksell, U.; Nilsson, J. L. G. Acta Pharm. Suec. Suppl. 2, 1985, 139. ( 7 ) Nichols, D. E.; Jacob, J. N.; Hoffman, A. J.; Kohli, J. D.; Glock, D. J . Med. Chem. 1984, 27, 1701.

Figure 2. Tetralin inversion wheel that defines the relationship between tetralin ring conformation and the tetralii inversion angle 4. The eight inserted tetralin structures carrespond to conformations with 4 = Oo, 30°, SOo, 150°, M O O , 210°,270°, and 330°, respectively. Each of the tetralin conformations is characterized by the signs (inserted)of the relevaht torsion angles?2 Perspective drawings of eight conformations of a (ZS)-Z-aminotetralin moiety are shown outaide the corresponding tetralin conformations. It should be noted that, for (2R)-2-aminotetralin, a half-chair conformation with a pseudoequatorial amino group corresponds to

4=

MOO.

reported the synthesis and pharmacologic& evaluation of (4a),the racemic 6,7-dihydroxy-2-methyl-2-aminotetr'alin C(2)-methyl-substituted derivative of ADTN (3a). In contrast to 3a, compound 4a was found to be inactive as a DA,-type DA a g ~ n i s t . Therefore, ~ we have included MMPZ calculations also on compounds 3c and 4c (which serve as model compounds* for 8a and 4a, respectively) in the present study.

30, R = O H 3b. R.OCH3

4a, R = O H 4b, R=OCH,

3c, R=H

4c. R = H OH

w

'"{/

NHCH(C H ~ ) ~

OH 5

Results and Discussion Definition of Conformational Parameters. Conformational parameters of particular interest in the present investigation are (a) the conformation of the nonaromatic ring, to which certain key intramolecular distances9 are (8) Test calculations (MMPZ) performed on compound 3a indi-

cate that the two hydroxyl groups do not significantly affect the conformationof the nonaromatic ring. In addition, PCILO calculations on protonated phenethylamine, tyramine, and DA have revealed only minor differences in the conformational characteristics of these compounds: Pullman, B.; Coubeils, J.-L.; Courriere, Ph.; Gervois, J.-P.J. Med. Chem. 1972, 15, 17.

Journal of Medicinal Chemistry, 1986, Vol. 29, No. 6 919

5-Hydroxy-2- (dipropylamino)tetralin

related (Figure 1) and (b) the conformation of the dipropylammonium (or dipropylamino) substituent, that is, the relative orientation of the ammonium hydrogen (or the lone pair of electrons on the nitrogen)1° as well as the conformation and relative spatial orientation of each of the N-propyl groups.'l The construction of a tetralin inversion wheel12 (see Figure 2) enables one to define the conformation of the nonaromatic ring of any tetralin derivative by use of the tetralin inversion angle 4. Ideally, this parameter is simply calculated from eq 1where T~~ is the observed value and

4 = arc COS (7&d/7m,)

(1)

is the maximal value (64.73') of the torsion angle T ( C ~ , C ~ , C ~ ,However, C ~ ) . ~ ~in some conformations of 1 and 2, bond lengths and/or angles are slightly distorted and therefore, eq 1 is no longer strictly applicable. In such cases, an approximate tetralin inversion angle is estimated by comparison with relevant conformations of C(2)-unsubstituted tetralin (see Figure 2). Due to the inherent symmetry (C2J of the tetralin molecule, the inversion wheel contains two degenerate boats (4 = 90" and 4 = 270°, see Figure 2). Two pairs of degenerate skew boats (that is, conformations with all except one carbon atom in the same plane) with P (4 = 150" and 4 = 210") and M helicity (4 = 30" and 4 = 330°), respectively, are also present in the inversion wheel. However, introduction of a C(2)-substituent destroys the symmetry of the tetralin moiety and now eight different conformations emerge (see the conformational drawings outside the inversion wheel in Figure 2), all of which are defined by their positions on the inversion wheel, that is, by the tetralin inversion angle 4. It should be noted that the tetralin inversion angle is configuration dependent. Therefore, in the present study we have used 4 only to describe conformations of (2S)-2aminotetralin derivatives. Five parameters are needed to describe the conformation of the dipropylammonium (dipropylamino) group: The torsion anglei4 7N = 7(C1,C2,N,Hor electron pair) defines the relative direction of the N-H bond or the electron pair. The torsion angles 7~ = 7(C2,N,C,,C,) and 7~ = T (N,C,,C,,C,) define the conformation of the N-propyl group, which in a clockwise sense is next to the N-H bond (N-electron lone pair) when viewing along the C(2)-N bond. Similarly, the conformation of the second N-propyl Molecular Mechanics Calculations. Results from quantum mechanical and force field calculations on 2aminotetralin derivatives have already appeared in the 1iterat~re.l~However, in these previous studies, no atFor discussions on important intramolecular distances in DAreceptor agonists, see for example ref 3c and Seeman, P. Pharmacol. Rev. 1980,32, 229. The importance of the direction of the lone pair of electrons on the nitrogen has been frequently discussed. For a review, see: Kaiser, C.; Jain, T. Med. Res. Reu. 1985, 5, 145. One of the N-propyl groups of 1 appears to participate in receptor binding: see ref 3a. The term pseudorotation is not strictly applicable to the tetralin ring inversion process since it implicates a continuous change of dihedral angles to that each ring atom sequentially takes up each of the possible ring positions. See: Hendrickson J. B. J . Am. Chem. SOC.1964,86, 4854. Compare: Vanhee, P.;Tavernier, D.; Baas, J. M. A.; van de Graaf, B. Bull. SOC.Chim. Belg. 1981, 90,697. For definitions of torsion angle and related concepts, see: Klyne, W.; Prelog, V. Elcperientia 1960, 16, 521.

tempt was made to identify all available low-energy conformations. In the present study we have performed full energy minimization with respect to all internal coordinates. For the calculations we utilized the MMP2 program developed by Allinger;16specifically, the 1980 force field,l' to which parameters for the phenol groupla have been added and in which updated amine parameters19have been implemented, was used. Throughout, calculations were performed on the free bases although the 2-aminotetralins probably interact with DA receptors in their protonated form.20 There is, however, a good agreement between the geometry of (2R)-l#HBras observed by X-ray crystallography and the calculated geometry of the corresponding nonprotonated conformation (vide infra). In addition, the conformational preferences of protonated and nonprotonated 2-aminotetralins in solution appear to be very similar (compare for example 3b and 3c.HC1 in Table IV) and calculations performed on 2-aminotetralin~l~~ and phenethylamines21indicate that conformationaldifferences between protonated and nonprotonated species in the gaseous state are small. Compounds 1 and 2 possess a considerable amount of conformational flexibility; the tetralin ring is only semirigid; that is, a number of tetralin conformations are possible. In addition, the hydroxyl and amino groups can rotate around the C(5)-0 and C(2)-N bonds, respectively, and the mobility of the two N-n-propyl groups further increase the conformational flexibilities of 1 and 2. Thus, a very large number of potential low-energy conformations of 1 and 2 can be envisaged. To deal with this problem, we adopted the following strategy: To each of the eight tetralin conformations in Figure 2, to which had been added the proper C(2)- and C(5)-substituents, a 2-dimethylamino substituent was added in three different ways so that 7(Cl,C2,N,N - electron pair) = 60°, 180°, and -60". The starting geometry of the hydroxyl groups of 1 and 2 was always set at T ( C ~ ~ , C ~ , O=, H ) The resulting 24 "starting geometries" were then minimized and conformations with energies >3.4 kcal/mol above the global minimum were discarded. To each of the residual conformations, two methyl groups were then added so that the nine possible staggered "2-(diethy1amino)tetralin geometries" were formed from each 2-(dimethylamino)tetralin conformation. These new "starting geometries" were minimized and conformations with energies > 3.2 kcal/mol above the global minimum were discarded. The MIMIC programB contains a conformational mapping option and it would have been facile to construct conformational maps when searching for the low-energy 2-(diethyl(a) Kocjan, D.; Hadzi, D. J. Pharm. Pharmacol. 1983,35,780. (b) Kocjan, D.; Solmajer, T.; Hodoscek, M.; Hadzi, D. Znt. J . Quantum Chem. 1983,23, 1121. (c) Grol, C. J.; Rollema, H. J . Pharm. Pharmacol. 1977,29, 153. Allinger, N . L. J. Am. Chem. SOC.1977, 99, 8127. Allinger, N.L.; Yuh, Y. Quantum Chem. Program Exchange 1980,-12, 395. Dodziuk. H.: von Voithenbere. H.: AUineer. N. L. Tetrahedron 1982,381 2811. Profeta, S.,Jr.; Allinger, N. L. J. Am. Chem. SOC.1985, 107, 1907. The dimethylsulfonium analogue of DA possesses DA agonistic properties: Andersson, K.; Kuruvilla, A.; Uretsky, N.; Miller, D. D. J . Med. Chem. 1981,24,683. Weintraub, H. J. R.; Hopfinger, A. J. J. Theor. Biol. 1973,41, 53. Test calculations have shown that conformations with 7 (C,,CS,O,H) = Oo consistently have energies 0.2-0.5 kcal/mol above those of the corresponding conformations with r(Ck,Cb,O,H) 180'. Liljefors, T. Mol. Graphics 1983,1, 111. yl

I

920 Journal of Medicinal Chemistry, 1986, Vol. 29, No. 6

Karl&n et al.

amino)tetralin conformations. However, most of the enconclusion is further supported by the large Jsax,4ax value ergetically favored geometries would not have been idenin compounds l*HC1,3b,and 5.HBr. For compound 2.HC1 tified since the program locks the conformations of the the coupling constants in Table IV offer less information: rotating moieties when working in this mode. Actually, The C(2)-hydrogenis substituted with a methyl group and when certain rotamers of a 2-(dimethylamino)- or 2-(dithus the structurally informative coupling constants with ethy1amino)tetralin derivative are minimized, both the the C(1)- and C(3)-hydrogensare not available. However,. geometry of the tetralin ring and the value of rN change a four-bond coupling constant (4Jiz: 1 Hz) between one of considerably. However, neither the conformation of the the C(1)-hydrogensand a C(3)-hydrogen was observed in tetralin ring nor the value of rN is changed appreciably the spectrum of 2.HC1. Such long-distance couplings occur when the terminal methyl groups of a 2-(di-n-propylwhen two distant hydrogens are arranged in a W conforamino)tetralin derivative are rotated. Therefore, the mation.26 Thus, the presence of this W coupling estabconstruction of the 2-(di-n-propylamino)tetralin “starting lishes the two interacting hydrogens as Hleq and H,,, geometries”, which were used for the final energy minirespectively, and also the conformation of the tetralin mization, was based on information from conforinational moiety as a half-chair. The half-chair tetralin conformamaps of the low-energy 2-(diethy1amino)tetralin confortion of 2.HC1 is further inferred from the large value of mations to which terminal methyl groups had been added J3=,& (Table IV). However, on the basis of the above lH (one to six 2-(di-n-propylamino)tetralin “starting NMR data, no conclusion can be drawn regarding the geometries” were identified in each conformational map). preferred conformation (pseudoequatorial or pseudoaxial) By use of this approach we have identified 28 low-energy of the dipropylammonium substituent of 2.HC1. Therefore conformations (within 3.0 kcal/mol of the global minimum) a 2D-NOESY spectrum of 2.HC1 was recorded. In this of 1 and 33 low-energy conformations of 2 (see Table I). NMR experiment, the intensities of cross peaks are related Four additional low-energy conformations of (2S)-2 (FF, to the distance between nuclei. The 2D-NOESY spectrum GG, HH, and 11; Table I) were identified by minimization of 2.HC1 shows that the C(2)-methyl hydrogens interact of all possible staggered “2-(dipropy1amino)tetralin strongly with HI, and Hsax. This is consistent only with geometries” formed by addition of methyl groups to the a pseudoequatorial disposition of the C(2)-methyl sub24diethylamino)tetralin conformations of (2s)1 and stituent since the distance between a pseudoaxial C(2)(2s)-2, which were found to have energies less than 0.5 methyl and H1, and H3=, respectively, would be too large kcal/mol above the respective global minimum. to account for the observed strong interactions. Interestingly, the energetically most favorable conforIn the report on the DA, inactivity of compound 4a, mations of 1 and 2 differ considerably with respect to Nichols et al.7 presented ‘HNMR data for the 0,O-di1 apattained tetralin inversion angles; compound (2s)methyl derivative 4b. No dipseudoaxial coupling constants pears to preferentially adopt 4 values around Oo whereas were present in the spectrum of 4b. This was suggested compound (2S)-2 preferentially adopts values around to indicate the presence of two equally populated equili180° (compare Table I and Figure 3). An indication of the brating tetralin conformations (4 = Oo and 4 iz: 180°).7It quality of the present calculations is obtained by comis thus apparent that N,N-dipropylation changes the conformational equilibrium considerably in C(2)-methylparing observed geometrical parameters from X-ray substituted 2-aminotetralins whereas that of C(2)-noncrystallography of (2R)-l*HClwith structural data from the corresponding calculated conformation (Table 11). methylated analogues is much less affected (compare also data in Tables I and 111). Calculations were also performed on 2-aminotetralin (3c) and on 2-amino-2-methyltetralin (4c) by applying essenThe 13C NMR spectra of 1.HC1 and 2.HC1 in CD,OD tially the same strategy as that described above. Results were also recorded; chemical shifts for resonances due to obtained (Table 111) are in agreement with previous C(l)-C(4) are shown in Table V. The 13C NMR asQCFF/PI and PCILO calculations performed on 2signements were verified by use of 2D chemical shift correlation spectroscopy. Recently, Morin et al.n reported a m i n ~ t e t r a l i n ’but ~ ~ ,disagree ~ with the large difference in I3C chemical shift parameters for methyl substituents in energy (8.7 kcal/mol) between “axial” and “equatorial” tetralin derivatives. The parameters were obtained by a ADTN (3c) that has been reported by Grol and Rollema.15“ least-squares regression analysis of the 13C NMR spectra NMR Spectroscopy. High-resolution ‘H NMR specof a large series of methyl-substituted tetralin~.~’The tral data of compounds 1-HC1and 2.HC1 in CD30D are result of addition of the 13Cchemical shift parameters for shown in Table IV. Use of 400-MHz spectroscopy allowed a pseudoaxial C(2)-methyl group to the observed chemical analysis of the spectra by first-order approximations. shifts of 1-HC1is interesting (Table V); the calculated Assignments and coupling constants were verified by chemical shift of C(4) is 6 18.4 whereas the observed spin-decoupling experiments, COSY spectroscopy, and chemical shift of C(4) of 2.HC1 is 6 22.05; that is, C(4) spin-spin simulation. Also included in Table IV are preexperiences a much smaller shielding (y effect) than what viously reported spectral data for three other 2-aminowould be expected if the C(2)-methyl substituent was tetralin derivatives (3b, 3c.HC1, and 5-HBr), which assume mainly half-chair conformations in s o l ~ t i o n . ~ ~ ~ ~ , ~ pseudoaxially ~ located. This indicates that the dipropylIt is noteworthy that the coupling constants in 1.HC1, ammonium group of 2.HC1 predominantly is pseudoaxially 2-HC1, 3b, 3c.HC1, and 5-HBr (Table IV) are similar in oriented and supports our results from the NOESY exmagnitude regardless of ring substitution, N-substitution, periment and the molecular mechanics calculations. ionization state of the nitrogen, or the solvent used. comComparison of Conformational Distribution of 1, 2, 3c, and 4c. There is a good agreement between the pounds 1.HC1, 3b, 3c.HC1, and 5.HBr appear to preferexperimentally determined (by NMR) and theoretically entially assume half-chair conformations with pseudoecalculated (by MMP2) conformational preferences of quatorial nitrogen substituents as indicated by the large dipseudoaxial coupling constants Jlau.2an and J2ax,3an. This C # J

(24) de Jong, A. P.; Fesik, S. W.; Makriyannis, A. J . Med. Chem. 1982, 25, 1438. (25) Motohashi, M.; Mizuta, E.; Nishikawa, M. Chem. Pharm. Bull. 1981,29, 1501.

(26) See, for example: Jackman, L. M.; Sternhell, S. “Applications of NMR Spectroscopy in Organic Chemistry”, 2nd ed.; Pergamon: Oxford, 1969. (27) Morin, F. G.; Horton, W. J.; Grant, D. M.; Dalling, D. K.; Pugmire, R. J. J . Am. Chem. SOC.1983, 105, 3992.

Journal of Medicinal Chemistry, 1986, Vol. 29, No. 6 921

5-Hydroxy-2- (dipropy1amino)tetralin

Table I. Geometrical Parameters for Low-Energy Conformations of (2S)-1and (2S)-2" conf

hbdeg

A B C

344 344 344 345 0 353 357 356

D E

F G H

I J K

0'

15' 0' 0' 0'

L

M N

0'

0

342 343 343 15' 15' 15' 15' 15'

P

Q

R S T

U

v

X

0'

Y

0' 205' 205' 205' 205'

2 AA BB

cc

~ ~ degt e

T A , ~deg

deg

TA~,'

deg

(2S)-5-Hydroxy-2-(di-n-propylamino)tetralin ((2S)-1) 57 61 178 -178 56 60 179 179 64 177 171 -63 65 180 57 -63 65 43 175 -63 88 -176 167 -178 173 134 178 -171 172 137 177 -172 -52 -171 174 -176 178 -57 98 171 -54 -57 173 -177 -52 -170 173 -176 -53 -170 171 -177 -56 -56 170 -177 170 -67 -160 -170 -61 -174 -64 -162 -61 -57 -64 -162 83 -179 168 -51 170 49 -56 -176 -55 -173 55 50 -176 169 52 -55 53 55 -53 -171 172 -66 -31 -171 55 -64 -31 -170 60 64 177 -178 60 64 178 179 66 180 170 -62 66 -176 56 -62

~ 9 . deg t

-170 -57 -179 -179 170 -176 180 -60 -170 -168 -169 100 -54 -53 -171 -171 -171 -174 170 170 57 56 -174 -175 -171 -57 180 179

re1 steric energy, kcal/mol

0.5h 1.0 0.6 1.1 2.8 2.9 2.8 3.0 W',k

2.9 0.4 2.6 0.2 0.4 2.7 2.6 3.0 3.0 0.4 0.7 0.9 0.9 2.5 2.8 0.8 1.1 0.6 0.7

(2S)-5-Hydroxy-2-methyl-2-(di-n-propylamino)tetralin ((2S)-2) 61 81 176 171 -172 2.4 82 178 170 -58 2.8 61 174 -174 1.9 66 -129 180 -119 168 -175 1.9 60 180 D 0' 176 -173 2.2 -135 0' 66 59 E -98 -57 159 -175 2.8 0' 50 F 2.4 -63 -114 59 177 163 0' G 2.2 -175 180 56 175 129 H 350' -178 63 -163 62 115 2.8 I 350' 174 -178 -59 136 2.6 350' 54 J 170 -170 -173 77 -69 1.8 K 18 -173 2.0 18 L 55 -168 77 -68 2.1 172 -169 41 168 M 0' 67 66 2.3 173 -54 40 166 N 0' -171 -131 -175 0' 2.9 180 177 0 2.4 -170 -173 -70 -158 P 0' 170 56 2.7 -168 -174 -69 -158 0' Q 172 R 22 -76 69 2.6 -170 168 -77 22 68 173 S 2.8 -55 165 T 12 -175 179 175 129 -62 2.6 -170 -173 -40 170 12 2.2 -68 U -168 -174 -40 V 55 12 2.4 -67 -174 X 205' 62 87 -178 0.8 168 Y -175 -58 205' 62 88 0.4 165 -174 Z -138 179 205' 65 0.5 178 -136 174 63 AA 205' -61 1.0 169 BB 200' -167 -168 -56 74 1.9 178 -166 -57 cc 200' 74 172 -54 1.8 DD 195' 175 179 -166 59 119 0.0' 195' 64 EE -165 59 117 0.1 -178 -158 FF 64 52 195' 174 98 0.3 -173 55 GG 195' 170 138 0.4 -58 HH 65 195' 66 -159 94 0.9 51 195' I1 55 0.8 61 -60 -168 136 74 JJ -172 200' -169 0.1 170 -68 KK 75 -172 200' 55 -167 -68 0.0' LL 74 -171 200' -165 -102 -66 2.7 a Only conformations with T(C~,CS,O,H) 1 80' are included and conformations with energies larger than 3 kcal/mol above the respective global minima have been omitted. *Tetralin inversion angle (see Figure 2). T N = T(C1,Cz,N,electronpair). T* = T(C~,N,C~,C&. e T~ = T(N,C~,C~,CJ. f ~ =kT(C~,N,C&,C~). g ~ g= , T(N,C&,C~,C,). hConformationthat corresponds to the X-ray structure of the A molecule of (2R)-1.HCl (ref IC). Approximate 6 value estimated by comparison with relevant conformations of C(2)-unsubstituted tetralin. j Steric energy = 13.7kcal/mol. kconformation that corresponds to the B molecule of (2R)-l.HCl(ref IC). 'Steric energy = 19.5kcal/mol. A B C

0' 0' 0'

'

922 Journal of Medicinal Chemistry, 1986, Vol. 29, No. 6 Table 11. Comparison between ((2R)-1) bond length, A obsd 1.36 C(7)-C(8) C(E)-C(Ea) 1.40 C(4a)-C(8a) 1.41 C(I)-C(Ea) 1.50 C(4)-C(4a) 1.51 C(Z)-C(3) 1.51 C(3)-C(4) 1.54 C(I)-C(2) 1.55 C(2)-N 1.53 1.52 N-C, N-C,. 1.51 C,-% 1.52 G-C, 1.52 C#& 1.50 CE-CY 1.52

Karlen et al.

Observed' and Calculated' Geometrical Parameters of (ZR)-5-Hydroxy-Z-(di-n-propylamino)tetraIin

calcd 1.39 1.40 1.40 1.51 1.51 1.54 1.53 1.54 1.49 1.48 1.47 1.54 1.54 1.54 1.54

valence angle, deg obsd C(7)-C(E)-C(8a) 122 C(4a)-C(Ea)-C@) 118 C(l)-C(Ea)-C(E) 120 C(I)-C(Ea)-C(4a) 122 C(Z)-C(I)-C(Ea) 114 C(I)-C(Z)-C(3) 109 C(l)-C(Z)-N 108 C(3)-C(Z)-N 112 C(2)-C(3)-C(4) 110 C(3)-C(4)-C(4a) 113 C(Z)-N-C, 111 C(Z)-N-C,. 115 Ca-N-Cd 112 N-C& 114 N-CsC, 113

c,-q-c, cscE-c,.

113

calcd 120 119 118 122 116 107 112

113 111

113

torsion angle, deg obsd C(8)-C(8a)-C-C(2) 168 C(Ea)-C(l)-C(Z)-N 166 C(l)-C(Z)-C(3)-C(4) -64 C(Z)-W-C(4)-C(4a) 53 C(3)-C(4)-C(4a)-C(Ea) -21 C(l)-C(2)-N-Ca. -165 C(Z)-N-C,-C, -177 C(2)-N-C,&p -71 N-C,-CgC, -176 N-CsCgC,, -173

ealcd 167 168 -62 52

-19 -174 178 -61

170 -178

111

113 112

116 116 112

109 111 'Observed values from the X-ray crystal structure of molecule A of (ZR)-L.HCI (ref IC). 'Calculated values (MMP2) for the enantiomer of conformation A of (2S)-I (Table I). Table 111. Geometrical Parameters for Low-Energy Conformations of (2S)-3c and (2S)-4c re1 steric energy, conf 4," deg r ~ deg ? keal/mal A B C D E F G H I J K L A

B C

D E F G H

(ZS)-Z-Aminotetralin((2S)-3e) 354 60 358 -177 356 -59 80 63 82

76 191 195 195 290 288 288

-176

-58 59 176

b

0.0' 0.2 0.2

4.2 4.5 4.4 0.9 1.2

-68

1.1

61 -179

4.2 4.4

-61

4.2

(ZS)-Z-Amino-Z-methyltetralin((2S)-4e) 17 60 0.5 16 180 0.6 16 -59 0.6 106 64 4.1 105 -175 4.2 100 -54 4.2 58 O.Od 191 173

I J K L 282 -61 3.8 'Tetralin inversion angle (see Figure 2). T~ = v(C,,Cz,N,electronpair). 'Steric energy = 0.6 kcal/mol. dStericenergy = 1.8 kcal/mol.

compounds 1,2,3c(3b), and k(4b) (vide supra). Further, the agreement between bond distances and bond angles in the X-ray structure of l.HC1 and those in the corresponding MMP2 conformation (Table 11) is impressive. Thus, the following MMPZbased (Tables I and 111) comparison of the conformational distributions of 1,2,3c, and 4c can be made with considerable confidence: (a) Preferred r$ values for (2.9-1 and (2S)-2 are around 0" and 180°, respectively (Table I and Figure 3). However, a significant part of the conformer populations of ( 2 8 - 1 and (2.9-2 will assume 4 values around 180' and Oo, respectively, since these latter conformations represent local minima with energies < 2 kcal/mol above the respective global minimum (compare Figure 3). For compounds 3c

Figure 3. Cunfurmational distrihution of (2S)-I (unfilled bars) and ( 2 3 - 2 (shaded bars). The probability of existence of each cunformarion (at 37 "CJ was estimated from a Boltzman distri. burion hased on the steric energies in Table 1. la) Distriburion of dipropylamino p u p rotamen in mnformations having o \dues around Oo. rbj Distribution of conformations having o values around 180°. In this rase only one dipropylamino gmup rotamer 17. = fiO") appear* ru be populated. For definitions of T~ and o. see text. and 4c, the energy differences between tetralin conformations with 0 values around 0' and 180° are much smaller (