Quaternary and Tertiary Quinuclidine Derivatives as Inhibitors of Choline Uptake

Quaternary and Tertiary Quinuclidine Derivatives as Inhibitors of Choline Uptake GERALDH. STEHLING*§~, PETERH. DOUKAs**, FIORE J. RICCIARDI, JR.*, AND JOHN J. O’NEILL’ Received July 3, 1990, from the *Department of Pharmacology, Temple University School of Medicine, 3420 N. Broad Street, and the *Division of Medicinal Chemistry, Temple University School of Pharmacy, Philadelphia, PA 19140, and the §Department of Pharmacology, Hahnemann Accepted for publication November 30, 1990. University School of Medicine, Philadelphia, PA. ~~

~

Abstract 0The uptake of choline into cholinergic neurons for acetylcholine (ACh) synthesis is by a specific, high-affinity, sodium- and temperature-dependenttransport mechanism (HAChU). Of several qua-

ternary quinuclidinolderivatives tested, the N-ally1derivative proved to be most potent. Though the methyl, ethyl, and isopropyl derivatives were less potent at comparable concentrations, at higher concentrations they also maximally inhibited HAChU. The benzyl, hydroxyethyl, and methoxyethyl derivatives failed to inhibit HAChU by >50% at concentrations up to 100 pM. N-AIlyl-3-quinuclidinol (NAQ) proved to be a specific inhibitor of HAChLl (lCs0 = 0.9 pM) and a poor inhibitor of both sodium-independenttransport (lC50 = 680 pM) and choline acetyltransferase activity (K, = 200 pM). The NAQ exhibited noncompetitive type inhibition compared with N-methyl-3-quinuclidinol,a competitive inhibitor of HAChU. Thus, substitution at the N-functional group not only alters potency, but may change the mechanism by which inhibition is produced. The optical isomers of NAQ and several derivatives were prepared and employed to examine the stereochemical selectivity for inhibition of choline uptake. The S(+)-isomerof NAQ (lCs0 = 0.1 pM) had -100-fold greater inhibitory activity for HAChU than the corresponding R( -)-isomer (IC, = 10 pM). With all other quinuclidinols tested, the S(+)-isomers were also more potent thari the corresponding R(-)-isomers. In an effort to obtain a tertiary inhibitor of HAChU that would be expected to cross the blood-brain barrier following peripheral administration, 3-biphenyl3-quinuclidinol (BHQ) and 3-naphthyl-3-quinuclidinol (NHQ) were synthesized and evaluated. The BHQ was a fairly potent inhibitor of HAChU (IC, = 7.8 pM) which maximally inhibited HAChU at 100 pM. Its methylated derivative demonstrated a similar profile; in contrast, NHQ only inhibited HAChU by 52% at 100 pM.

The uptake of choline into cholinergic neurons for use in the synthesis of acetylcholine (ACh) is facilitated by a specific, high-affinity transport mechanism (HAChU) which is temperature and sodium dependent.’-5 Structural requirements for inhibitors of HAChU, which include choline analogues, hemicholinium derivatives such as HC-15 and HC-3, and quinuclidines, have been investigated by a number of laboratories.6-10 A comparison of these structural classes follows. Recently, the quinuclidine nucleus has been used as a struc-

Choline

Heniicholiniurn -15 (HC-15)

Quinuclidinols

L

A

fllkyl, flralkyl 9-11 -CH, 12-14 -H

Cpd # I - 8

0022-3549/9110800-0785$01 .OO/O 0 7997, American Pharmaceutical Association

2 -H

-flryl -flryl

tural backbone for the development of compounds affecting ACh synthesis, either as inhibitors of HAChU or of choline acetyltransferase (ChAT),gJ’ based in part on its known complementarity to cholinergic receptors.12 The ability of N-methyl-3-quinuclidinol(NMQ, l), a rigid analogue of choline, to inhibit HAChU was initially reported by Dowdallg and Kuhar and Murrin.10 Inhibition of HAChU by NMQ was shown to be stereospecific by Ringdahl et al.,13 with the isomer derived from S( + )-quinuclidinol exhibiting a 12-fold greater potency than the corresponding isomer derived from the R(-)-alcohol. The absolute configuration of these isomeric alcohols had originally been established by the X-ray diffraction studies of Baker and Pauling,14 and have more recently been confirmed by Saunders et al.15 Our laboratory previously reported the inhibitory activity of several quinuclidine derivatives, including that of N-allyl3-quinuclidinol (NAQ, 5 ) , the racemic mixture of which was shown to be more active than racemic NMQ.7 The quinuclidines were synthesized in a n effort to find selective therapeutically effective cholinolytic compounds with the objective of avoiding some of the undesirable effects known to be caused by the hemicholiniums. In the present study, stereoisomers of NAQ and several quaternary analogues (Table I, 2-8) were prepared and evaluated to further elucidate the structural requirements for inhibition of HAChU. In addition, lipophilic tertiary quinuclidinol derivatives (Table 11, 12-14) were also synthesized and evaluated for inhibition of choline uptake. Preliminary results have previously been reported.16

Experimental Section Materials-Male Sprague-Dawley rats (180-200 g) were purchased from Charles River Breeding Laboratories (Wilmington, MA). New England Nuclear (Boston, MA) was the supplier of Biofluor scintillation cocktail, [l4C1acetylCoA (specific activity, 4 mCi/mmol), and [3Hlcholine chloride (specific activity, 80-90 Ci/mmol). Sucrose, Ficoll (Type 4001, and various biochemicals were obtained from Sigma Chemical Company (St. Louis, MO). Glass microfiber filters and laboratory chemicals were from Fisher Scientific Company (King of Prussia, PA). Quinuclidinol Derivatives-All compounds evaluated for inhibition of HAChU were synthesized in this laboratory. Separation of the optical isomers of 3-quinuclidinol was accomplished using established literature methods, either by derivatization to acetoxyquinuclidine according to the method of Ringdahl e t a1.,17 or by resolution of the alcohol itself using p-chlorotartranilic acid by the method of Lambrecht.18 3-Aryl substituted quinuclidinols were prepared using standard Grignard reaction methods as previously reported.7 Preparation of Rat Brain Synaptosomes-Synaptosomal preparations (nerve endings) from rat cerebral cortex were obtained by the method of Cotman19 as described by Sterling e t a1.7 Animals were sacrificed by decapitation and the cerebral cortex was collected in ice-cold 0.32 M sucrose. Tissue was homogenized in 20 volumes of sucrose, then centrifuged at 1300 x g for 3 min. The supernatent was recentrifuged a t 14 000 x g for 20 min. The crude “P2”pellet was resuspended in sucrose and further purified by separation on a Ficoll Journal of Pharmaceutical Sciences I 785 Vol. 80, No. 8, August 1991

Table Clnhlbltlon of HAChU by Quaternary SQulnuclldlnol Derlvatlvea

aO” I +

A

Derivative Racemic Compound

R

I C ,

ma 1 (NMQ) 2 3

4 5 (NAQ) 6 7 8

4 %

-CH,CH, 4H2(CH3)2 -CH,CH,CH, -CH,CH=CH, -CH,CH,OH -CH,CH,OCH, --CHZ--C,H,

Methyl Ethyl lsopropyl Propyl Ally1 Hydroxyethyl Methoxyethyl Benzyl

S( +)-Isomer

Maximum Percent Inhibitionb

12.0 2.3 4.5 2.2 0.9

97 2 4 95 2 4 92 4 79 2 7 92 2 2 49 3 20 2 4323

* *

100

*

>100 >lo0

Maximum Percent lnhibition

C I,

9622 9223 100 f 3 94 f 1

1.1 0.95 0.9 0.1

-

-

-

R(-)-Isomer G o

>lo0 100 8.9 10.0

-

Maximum Percent lnhibition

*

38 5 50 2 6 69 2 2 51 * 3

-

a Concentration of compound (pM)producing 50% inhibition of high-affinity[3H)cholirre(1 pCi, 2 pM) uptake; at least eight concentrations of inhibitor from 10 nM to 100 pM were run (4-6 samples per concentration) to generate each C .I, Maximum percent inhibition at 100 pM.

Table ll-lnhlbltlon of HAChU by Phenyl, Biphenyl, and Naphthyl Qulnuclldlnol Derlvatlver

Maximum Percent Inhibitionb

Compound ~

Reference

~

9

9024

7

6.2

92 +- 3

7

10

10

11

43%

7.5

94 2 4

This study

12 (BHQ)

-H

7.8

95 2 2

This study

13 (NHQ)

-H

3.0

52 2 4

This study

14

-H

60.0

69 2 5

This study

CI a

Concentration of compound producing 50% inhibitionof high-affinity (3H]choline(1 pCi, 2 pM) uptake. Maximum percent inhibition at 100 pM.

gradient (48, 68, 13%) by centrifugation at 63 500 x g for 60 min. The synaptosomal band (interface between the 6 and 13% layers) was diluted with 20 mL of sucrow solution and recentrifuged at 50 000 x gfor 20 min; the pellet was resuspended in Krebs-Ringer bicarbonate buffer (0.7-1 mg protein/mL) for analysis of choline uptake. Procedure for Analysis of Choline U p t a k d y n a p t o s o m a l uptake of choline was determined as described by Sterling et a1.7 Synaptosomes were suspended in Krebe-Ringer bicarbonate buffer containing 120 mM NaCl, 5 mM KCI, 30 mM NaHCO,, 1 mM MgSO,, 1 mM KH,PO,, 3 mM CaCl,, and 10 mM D-glucose. To measure sodium-independent choline uptake, synaptosomes were incubated in 786 I Journal of Pharmaceutical Sciences Vol. 80, No. 8, August 7997

“low-sodium”(30mM) buffer; sodium chloride in excess of 30 mM was replaced by lithium ch1oride.m The synaptosomal suspension was aerated for 5 min with 96%:6% 02:C02and preincubated for 5 min at 37°C. [3H]Choline (1 pCi) was then added and each sample was incubated for 4 min. Total choline concentration was 2 p M in atudies evaluating potential uptake inhibitors. To measure inhibition, compounds to be tested were added just prior to the [3H]choline. Incubation was terminated by placing samples on ice and adding 1 mL of ice-cold buffer with 2 mM choline. Samples were filtered through presoaked Whatman GF/C glass microfiber filters and washed with buffer. Filters were placed in scintillation vials, tissue

was digested, Biofluor scintillation cocktail was added, and the samples were counted by liquid scintillation spectrometry. [3H]Choline in buffer without synaptosomes was passed through a filter to measure the filtration blank which was substrated from all samples. Sodium-dependent choline uptake was calculated by substrating sodium-independent from total uptake. Protein was measured by the BioRad protein assay.21 Statistical Analysis-The IC,, values, those concentrations which produce 50% inhibition of choline uptake, were obtained by plotting the percent probability of inhibition of choline uptake against the log concentration of inhibitor according to the method of Litchfield and Wilcoxon.22 Lineweaver-Burke double reciprocal plots were derived by computer fitting of kinetic data. The computer programs used were those described by Tallarida and Murray.23

Results and Discussion

*

100

0

8 0

. I

8

2 0

0 10-9

1 0 - 8

10-7

10-6

Pc

m

10-9

Enantiomeric pairs for NAQ and several analogues were prepared from the corresponding optically resolved 3-quinuclidinols and evaluated for inhibition of HAChU in vitro; certain quaternized racemates were also prepared and similarly evaluated. The inhibitory potencies of the various quaternary quinuclidinols are listed in Table I, with doseresponse relationships of racemic compounds shown in Figure 1. The control values for high affinity, sodium-dependent choline uptake (1 pCi, 2 pM [3H]choline) was 32.6 2.1 pmol/mg proteid4 min. In all cases where the activity of each of the isomeric pairs was determined, the S( + )-isomer exhibited greater inhibitory potency, in accord with a recent report concerning the stereospecificityof inhibition by NMQ (11.13In the present study, S(+)-NAQhad 100-foldgreater inhibitory activity than the corresponding R( -)-isomer, and was the most potent compound in the series (Table I, Figure 2). The stereospecificity exhibited by this series of compounds contrasts with the hemicholiniums, in which no stereospecificity has been reported. The NAQ proved to be a specific, as well as potent inhibitor of HAChU. The racemic compound inhibited HAChU with an IC,, of 0.9 pM compared with an IC,, of 680 pM for sodium-independent choline uptake. In additional studies in our laboratory, NAQ was found to be a weak inhibitor of choline acetyltransferase activity (Ki = 200 pM from Hofstee analysis). In separate studies on brain slices, we have employed this compound for its effect on high potassium ion and veratridine-stimulated ACh release.12 The compound failed

e

t .*

10-s 10-4

Concentration (M) Figure 1-Doseresponse relationship of racemic quinuclidinol derivatives on high-affinity choline uptake (HAChU). Synaptosomes were incubated with [3H]choline(2 pM, 1 pCi) with various concentrations of compounds to be evaluated. The IC, values are shown in Table I. Key to R group: (0)allyl; ( 0 ) Npropyl; (A) isopropyl; (0) hydroxyethyl;(0) methoxyethyl; (0)benzyl; (0) methyl; (El)ethyl.

10-8

10-1

10-6

10-5

10-4

IO-~

Concentration NAQ (M) Figure 2-Stereoisomers of Nallyl-3-quinculidinol were evaluated for inhibition of HAChU in synaptosomesincubated with [3H]choline(1 pCi, 2 pM). Each point represents the mean 2 SE of 4-6 observations. The IC, values are 0.9,O. 1, and 10.0 pM forthe racemic compound (0); the S(+)-isomer(W), and the R(-)-isomer (A),respectively.

to directly inhibit ACh release, though by inhibiting HAChU, subsequent release of both labeled and total ACh from cortical slices was markedly reduced. Certain general observations can be made regarding the structure-activity relationship of quaternary quinuclidinols. The addition of one methylene unit to the quaternizing moiety, in the transition from N-methyl (1) to N-ethyl (2), resulted in a fivefold increase in inhibitory potency (Table I, Figure 1).A further increase in lipophilicity, chain length, or volume, as with n-propyl (4) and isopropyl (3) quaternizing groups, did not result in a concomitant enhancement of activity. This differs from observations with simple quinuclidinium salts as inhibitors of acetylcholinesterase wherein transition from methyl, through ethyl, to propyl resulted in a stepwise increase in inhibitory potency.24 Quaternization of quinuclidinol with an allyl group (5), however, leads to a further enhancement of activity; the S( +)-N-ally1 derivative exhibited an inhibitory potency 10 times greater than the corresponding N-ethyl isomer (2). Although the lipophilic contribution of the allyl group does not differ significantly from that of the ethyl group,25 its steric and electronic configurations differ from those of ethyl, propyl, and isopropyl. The additional inhibitory activity afforded by this planar pi-electron-rich group suggests a complementary site on the binding surface in the biophase. A similar observation has been made by Benz and Long26 with regard to inhibition of neuromuscular activity by a series of dimethylalkylammonium derivatives of hemicholinium-3, itself a potent inhibitor of HAChU; specifically, the N-ally1 derivative exhibited greater inhibitory potency than the corresponding propyl and isopropyl analogues. The use of polar functions, such as N-hydroxyethyl (6) and N-methoxyethyl (7), substantially reduced activity, indicating that the quaternizing group in this series aligns with a lipophilic area that does not accept hydrophilic moieties. This pi-binding lipophilic pocket also has steric limitations, evidenced by the poor activity of the N-benzyl analogue (8);the N-benzyl analogue of choline has similarly been reported to be a poor inhibitor of HAChU.6 The more potent S(+)-isomer of NAQ was evaluated as to the nature of its inhibition of HAChU. Kinetic parameters for HAChU demonstrated a K, of 1.4 pM and V,,, of 54.5 pmol/mg proteid4 min. These values are in the range of those previously reported for HAChU.2,4,6,27-29A double reciprocal plot in the presence or absence of S(+)-NAQ is shown in Journal of Pharmaceutical Sciences I 787 Vol. 80,No. 8, August 1991

Figure 3. The S(+)-NAQ isomer reduced the V,,, to 32.5 pmoltmg proteid4 min, whereas the K, of 1.53 was not significantly affected, suggesting noncompetitive inhibition. A question that still remains to be resolved is the observation that the S-isomer displays a typical sigmoidal relationship to maximum efficacy in contrast to the less active R-isomer (refer back to Figure 2 and Table I). A possible explanation for this difference, which occurs in more than one of the isomeric pairs, is the concept of multiple binding sites. In this case, the S-isomer would bind to two postulated sites, leading to full activity, and the R-isomer, which differs spatially, may only interact with one such site, accounting for lower activity. Another explanation is that saturation of one allosteric site by the S-isomer causes a conformational change, leading to maximum inhibition, whereas the R-isomer causes only a partial conformational change, thus not fully altering uptake. Further studies need to be done to elucidate this mechanism. Observations to date have stressed the need for a quaternary nitrogen as essential for inhibition of HAChU, and many of the most potent inhibitors contain a quaternizing methyl group. This requirement, however, does not simply reflect a prerequisite for a positive charge; the methyl group also contributes a critical space-fillingand binding function. Many of the reported inhibitors of HAChU also possess lipophilic aromatic functions (phenyl, biphenyl) some distance from the quaternizing moiety, as found for example with the hemicholinium series. It is conceivable that with appropriate manipulation of a lipophilic center adjacent to the hydroxyl function in the quinuclidinol series, its relative contribution to the overall binding process between inhibitor and choline carrier may be enhanced, thereby overcoming the necessity for a quaternized nitrogen. This type of approach has been applied successfully in the development of tertiary amine inhibitors of ChAT by Gibson and Baker.30-31 A similar approach in the search for inhibitors of HAChU was utilized in our laboratory with a limited series of

3-quinuclidinols (Table 11). In our previous study, the quinuclidinyl isostere of HC-15, N-methyl-3-phenyl-3-quinuclidino1 (91, exhibited inhibition of HAChU with an IC,, of 10 pM;7 moderate enhancement of its activity was afforded by addition of a lipophilic para-chloro substituent (10). In the present study, the 3-biphenyl analogue (11) was synthesized and found to exhibit inhibitory activity similar to that of the para-chloro analogue. This indicated steric tolerance of the additional phenyl ring in the para-position. The nonquaternary analogue 3-biphenyl-3quinuclidinol(BHQ, 12) demonstrated activity equal to that of the N-methyl derivative. This latter observation suggests that the binding contribution of the biphenyl group may compensate for the lack of the quaternizing methyl group. The 3-naphthyl (13) and 3-(3’,5’-dichlorophenyl) (14) tertiax-y amine analogues also inhibited HAChU, although they were not as active as the biphenyl derivative. Though NHQ inhibited HAChU by 50% at low concentrations, it failed to further inhibit at concentrations up to 100 pM. In the present study there was no appreciable difference in inhibition between tertiary BHQ (12) and its quaternary salt (11). This is in contrast to an active bis-tertiary amine possessing HC-3-like activity in caudate slices previously reported.2’-29 Quaternization of the bis-tertiary amine compound was reported to enhance inhibitory activity 500-fold. The two compounds exhibit structural differences that may explain the variance in sensitivity to quaternization: (a)the bis-compound is considerably larger than BHQ; (b)the relative contribution of the biphenyl moiety to BHQ, with regard to overall molecular bulk, lipophilicity, and hydrophobic bonding, is greater when compared to the bis-compound; (c) whereas BHQ possesses one nitrogen, the bis-compound possesses two nitrogens, both of which may contribute to binding; and (d) the two compounds may not share the same binding domain. Further work needs to be completed to clarify these differences.

References and Notes 1. Yamamura, H. I.; Snyder, S. H. Science 1972,178, 626-628. 2. Barker, L. A.; Mittag, T. W.Phurmncol.Exp. Ther. 1975,192,86-94. 3. Simon, J. R.; Kuhar, M. J. Nature 1975,255, 162-163. 4. Simon, J. R.; Atweh, S.; Kuhar, M. J . J . Neurochem. 1976, 26, 909-922. 5. Jope, R. S. Brain Res. 1979, 1 , 313-344. 6. Fisher, A.; Hanin, I. Life Sci. 1980,27, 1615-1634. 7. Sterling, G. H.; Doukas, P. H.; Ricciardi, F. J.; Biedrzycka, D. W.; ONeill, J. J. J . Neurochem. 1986, 46, 1170-1175. 8. Long, J. P.; Schuler, F. W .J . Am. Pharm. Assoc. (Sci.Ed.) 1954, 43, 79-86. 9. Dowdall, M. J. In Cholinergic Mechanisms and Psychopharmacology; Jenden, D. J., Ed.; Plenum: New York, 1978; pp 359-376. 10. Kuhar, M. J.; Murrin, L. C. J . Neurochem. 1978,30, 15-21. 11. Doukas, P. H.; Sterling, G . H.; Sheldon, R. J.; O’Neill, J. J. In

Cellular and Molecular Basis of Cholinergic Function; Dowdall, M., Ed., Raven: New York, 1987, pp 355-360. 12. Schulman, J. M.; Sabio, M. L.; Disch, R. L. J . Med. Chem. 1983,

I

26, 817-823. 13. Ringdahl, B.; J o e, R. S.; Jenden, D. J . Biochem. Pharmacol. 1984,33,2819-2f;22. 14. Baker, R. W . ; Pauling, J. J . Chem. Soc., Perkin Trans. 2 1972, 2340-2343. 15. Saunders, J.; Showell, G. A.; Baker, K.; Freedman, S. B.; Hill, D.;

I

-0.5

0

0.5

1.0

1.5

2

hCholine], pM Figure &Effect of S(+)-N-allyl-3-quinuclidinolon the kinetics of HAChU analyzed on a double reciprocal plot (Vis expressed in units of pmol/mg proV4 min). Synaptosomes were incubated with various concentrations of [3H]choline(1-8 p M ) in the presence (A)or absence (0)of NAO (1 PM).

788 I Journal of Pharmaceutical Sciences Vol. 80, No. 8, August 7997

McKnight, A.; Newberry, N.; Salamone, J. D.; Hirshfield, J.; Springer, J. P. J . Med. Chem. 1987,30,969-975. 16. Sterling, G. H.; Doukas, P. H.; O’Neill, J. J. Pharmacologist 1986,28 (3), 130. 17. Ringdahl, B.; Resul, B.; Dahlbom, R. ActaPharm. Suec. 1979,16,

281-283. 18. Lambrecht, G. Eur. J . Med. Chem. 1979,14, 111-114. 19. Cotman, C. W . Methods Enzymol. 1974,31,445-455. 20. Rylett, R. J.; Ball, M. J.; Colhoun, E. H. Brain Res. 1983, 289, 169-1 75.

21. Bradford, M. Anal. Biochem. 1976 72,24%254. 22. Litchfield, J. T.; Wilcoxon, F. J . bharmacol. Exp. Ther. 1949, 9699. 23. Tallarida, R.J.; Murra R B. In Manual of Pharmacologic Calculations; Springer-+&lag: New York, 1981, pp 59-63. 24. Kellet, J. D.; White, C. J. Pharm. Sci. 1965,64, 883-887. 25. Hansch, C.; Leo, A. J. Substituent Constants for Correlation Analysis in Chemistry and Biology; Wiley: New York, 1981. 26. Benz, F. W.; Long, J. P. J. Pharmacol Exp. Ther. 1969, 168, 315-321. 27. Tedford, C. E.; Reed, D.; Bhattacharya, B.; Bhalla, P.; Cannon, J. G.; Long, J. P. Eur. J . Pharmacol. 1986, 128, 231-239. 28. Tedford, C. E.; Schott, M. J.; Flynn, J. R.; Cannon, J. G.; Long,

J. P. J . Pharmacol. Exp. Ther. 1987,240 (2), 41-85. : Flynn, J. R.; Bhatna ar, R. K.; Cannon, J. G. 29. Tedford, C. E Long, J. P. J.h m n a c o l . Exp. Ther. b88,247 (2),460465. 30. Baker, B. R.; Gibson, R. E. J. Med. Chem. 1971,14,315-322. 31. Baker, B. R.; Gibson, R. E. J. Med. Chem. 1972,15,639-642.

Acknowledgments The authors thank Ms. Diane Ciamaichelo for typing the manuscript and Mr. Remus Berretta for reparing the figures.This work was BU ported b the U.S.Army Zesearch and Development Command DAM6 11-864-6243.

180.

Journal of Pharmaceutical Sciences I 789 Vol. 80, No. 8, August 7997