INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

University Microfilms International A Bell & Howell Information C om pany 300 North Z eeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800 521-0600 Order Number 9120692

Part 1. Synthesis of fiuorinated derivatives as potential and thromboxane A 2 antagonists. Part 2. Synthesis of hydrazinium analogs of and antagonists

Markovich, Kimberly M., Ph.D.

The Ohio State University, 1991

Copyright ©1991 by Markovich, Kimberly M. All rights reserved.

UMI 300 N. Zeeb Rd. Ann Arbor, MI 48106 NOTE TO USERS

THE ORIGINAL DOCUMENT RECEIVED BY U.M.I. CONTAINED PAGES WITH SLANTED PRINT. PAGES WERE FILMED AS RECEIVED.

THIS REPRODUCTION IS THE BEST AVAILABLE COPY. Part 1: Synthesis of Fluorinated Catecholamine Derivatives as Potential Adrenergic Stimulants and Thromboxane A2 Antagonists Part 2: Synthesis of Hydrazinium Analogs of Dopamine Agonists and Antagonists

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

By

Kimberly M. Markovich, B.S., R.Ph.

* * it * *

The Ohio State University

1991

Dissertation Committee: Approved by

Duane D. Miller, Ph.D.

Robert W. Brueggemeier, Ph.D.

Dennis R. Feller, Ph.D.

Donald T. Witiak, Ph.D. Duane D. Miller, Ph.D., Advisor College of Pharmacy Copyright by Kimberly M. Markovich 1991 DEDICATION

To Steve

- ii - ACKNOWLEDGEMENTS

I would like to express my sincere thanks to the following individuals:

Dr. Duane D. Miller, for his guidance, support and friendship throughout my graduate career.

The members of my reading committee, Dr. Robert Brueggemeier, Dr. Dennis Feller and Dr. Donald Witiak, for their advice in the preparation of this manuscript.

Dr. Yoshiya Amemiya, for his invaluable advice on both chemistry technique and various synthetic aspects of this work and his friendship.

Dr. Akihiko Hamada, for his much appreciated advice with various aspects of this work and for synthesizing the racemic dibenzyloxy protected 8 -fluorotrimetoquinol.

Dr. Karl Romstedt, Dr. Gamal Shams, Katherine Doyle, Paul Fraundorfer and Dr. Dennis Feller, for the pharmacological data of the trimetoquinol derivatives and their help in the interpretations of the data.

Tahira Farooqui and Dr. Norman Uretsky, for their work on the biological studies of the dopamine antagonists.

Jack Fowble for his help in obtaining and interpreting various spectra, for sharing his pearls of wisdom and his friendship.

John Miller, for running my mass spectra on demand, for his help in obtaining other various spectra and for his support and friendship.

My fellow graduate students, especially Vimon Tantishaiyakul, Mustapha Beleh, Soheila Ebrahimian and Allen Hopper for their friendship, advice, and for their never ending support (and sarcasm).

Bruce Posey, for assistance in using the college computer system, especially for the preparation of this document.

- iii - Dr. Kurt Loening, for his assistance with the nomenclature of the hydrazinium analogs.

Carol Settles, for her friendship and continued support throughout my graduate career.

Most of all, to Dr. Stephen Beck-thanks for waiting.

- iv - VITA

June 9, 1963 Dorn - Chicago, Illinois

May, 1986 B.S. Pharmacy, Butler University, Indianapolis, Indiana

1986-present Academic Challenge Fellow The Ohio State University Columbus, Ohio

Sept. 1986 - Aug. 1988 Graduate Teaching Assistant College of Pharmacy The Ohio State University

Sept. 1988 - present Graduate Research Associate College of Pharmacy The Ohio State University

PUBLICATIONS

1. Gary L. Grunewald, Kimberly M. Markovich, and Daniel J. Sail, "Orientation of and Analogues in the Benzonorbornene and Benzobicyclo[3.2.1]octane Ring Systems at the Active Site of Phenylethanolamine N-Methyltransferase (PNMT)", J. Med. Chem. 1987, 30, 2191.

2. Kimberly M. Markovich, Akihiko Hamada, and Duane D. Miller, "Unexpected Aminoisoquinoline Formed Under Bischler-Napieralski Conditions Provides For A New Synthesis of 3-Aminoisoquinolines", J. Heterocyclic Chem. 1990, 27, 1665.

FIELDS OF STUDY

Major Field: Pharmacy

Studies in Synthetic Medicinal Chemistry TABLE OF CONTENTS

PAGE

D e d i c a t i o n ...... ii

Acknowledgements ...... iii

VITA ...... v

List of...... F i g u r e s ...... vii

List of T a b l e s ...... viii

List of Schemes ...... ix

CIIAPTER

Part 1: Synthesis of Fluorinated Catecholamine Derivatives as Adrenergic Stimulants and Thromboxane Aj Antagonists . . . . 1

INTRODUCTION ...... 2

1.1 Sympathetic Nervous System ...... 2 1.1.1 Catecholamine Biosynthesis ...... 3 1.1.2 Storage and Release of ...... 5 1.1.3 Metabolism and Reuptake ...... 6 1.2 Subclassification and of Adrenergic Receptors ...... 9 1.3 Structural Properties of Adrenergic Receptors ...... 31 1.4 Structure Activity Relationships of Adrenergic R e c e p t o r s ...... 38 1.4.1 Alpha Adrenergic Structure Activity Relationships ...... 38 1.4.2 Beta-2 Structure Activity Relationship ...... 47

- vi - 1.5 Physiological Functions of Adrenergic Agonists ...... 50 1 .6 Thromboxane A £ ...... 55 1.6.1 Hemostasis and Platelet Function ...... 57 1 .6.2 Biochemical Role of TXA« in Platelet Activation . . 61 1.6.3 TXA2 Involvement in Cardiovascular Disease .... 65 1.6.4 TXA2 Antagonists ...... 66

STATEMENT OF PROBLEMS AND OBJECTIVES ...... 69

2.1 Trimetoquinol Analogs ...... 69 2.2 Imidazoline Derivatives ...... 74 2.3 Derivatives ...... 77

RESULTS AND DISCUSSION ...... 79

3.1 CHEMISTRY ...... 79 3.1.1 Synthesis of 5- and 8 -Trifluoromethylated Trimetoquinol Analogs ...... 79 3.1.2 Separation of 8 -Fluorotrimetoquinol Enantiomers . . 97 3.1.3 Trifluoromethylated Catecholimidazoline Analogs . . 99 3.1.4 Trifluoromethylated Norepinephrine Analogs .... 103 3.2 B I O L O G Y ...... 109 3.3 S U M M A R Y ...... 113

EXPERIMENTAL ...... 115

Part 2: Synthesis of Ilydrazinium Analogs of Dopamine Agonists and Antagonists ...... 160

INTRODUCTION ...... 161

5.1 Location, Distribution and Function of Dopamine .... 161 5.2 Dopamine Subtypes ...... 163

STATEMENT OF PROBLEMS AND OBJECTIVES ...... 171

RESULTS AND DISCUSSION ...... 180

7.1 CHEMISTRY ...... 180 7.2 BIOLOGY ...... 197 7.3 SUMMARY ...... 197

EXPERIMENTAL ...... 198

BIBLIOGRAPHY ...... 209

- vii - LIST OF FIGURES

Catecholamine biosynthesis ...... 4

Metabolic pathways of catecholamines ...... 7

fij- and fi 2 _Adrenoceptor agonists and antagonists . . . . 11

Atypical adipocyte-selective beta-adrenoceptor agonists . 12 a^-Adrenoceptor agonists and antagonists...... 14

(^“Adrenoceptor agonists and antagonists...... 15 a 2fi-Adrenoceptor selective compounds ...... 22

Mechanism of signal transduction upon OC^-adrenoceptor stimulation ...... 25

Hydrolysis of PIP2 by ...... 26

Mechanism of coupled signal transduction . . . 28

Hydrolysis of GTP by G protein-GTPase...... 29

Mechanism of /^-adrenoceptor and a 2 _adrenoceptor signal transduction ...... 30

Conversion of ATP to cAMP by adenylate cyclase...... 30

Topography of the /^“ ...... 32

Extracellular view of the beta-adrenergic binding site 35

The Easson-Stedman hypothesis model ...... 39

Fisher projections of imidazolines used to test the Easson-Stedman hypothesis ......

Proposed conformations of norepinephrine and for alpha-adrenergic receptor binding ......

- viii - 19. Phenyl ring substituted clonidine-like compounds...... 44

2 0 . Imidazoline modified alpha-adrenergic agonists ...... 46

21. Selective/^"adrenergic agonists...... 49

22. Biosynthesis of TXA£ via the arachidonic acid cascade . . . 56

23. Intrinsic coagulation pathway...... 60

24. Role of TXA2 in platelet activation...... 62

25. Interrelationship of platelet receptor mechanism ...... 64

26. TXA2 Receptor Antagonists...... 68

27. Trifluoromethylated benzylimidazoline target compounds . . 76

28. Rotameric conformations of 2- and 6 -fluoronorepinephrine . 77

29. Initial retrosynthetic analysis of 8 -trifluoromethy1 trimetoquinol ( 6 4 ) ...... 84

30. Alternate Retrosynthesis of 6 4 ...... 90

31. *H NMR (250 MHz) spectrum of protoberberine 1 0 2 ...... 93

32. Mass spectra of 1 0 3 ...... 95

33. XH NMR (250 MHz) spectra of 103 in CDCI3/TMS ...... 96

34. Selective Dj receptor agonists and antagonists ...... 165

35. Selective D2 agonists and antagonists ...... 167

36. Permanently charged and uncharged analogs of dopamine a g o n i s t s ...... 172

37. N-Substituted norapomorphines ...... 174

38. Permanently charged and uncharged dopamine antagonists . . 175

39. Interaction of and permanently c h a r g e d ...... 176

40. Target hydrazinium and antagonist a n a l o g s ...... 179

41. Interaction of hydrazinium functionalities with a carboxylate i o n ...... 179

- ix - 42. Portion of the *H NMR spectrum of 159 in CDClg/TMS .... 186

43. Portion of the NMR spectrum of 160 in C D g O D ...... 187

44. NMR spectrum of 152 in CDgOD ...... 190

- x - LIST OF TABLES

TABLE PAGE

1. Characteristics of Alpha-1 Adrenergic Receptor Subtypes . . . 19

2. Physiological Functions of Alpha-Adrenoceptor Stimulation . . 51

3. Physiological Functions of Beta-Adrenoceptors ...... 54

4. Comparative Activities of TMQ, 5F-TMQ, and 8F-TMQ on Beta Systems (Guinea Pig Trachea and Atria) ...... 71

5. Comparative Inhibitory Activities of TMQ, 5F-TMQ, and 8F- TMQ in TXA2 S y s t e m s ...... 72

6 . Comparative Activities of TMQ (50) and 5CFo"TMQ (63) on Beta-Systems (Guinea Pig Trachea and Atria) ...... 110

7. Comparative Inhibitory Activities of TMQ (50) and 5CFg-TMQ Against U46619-Induced Human Platelet Aggregation ...... Ill

8 . Activities of Racemic TMQ (50), S(+)-62 and R(-)-62 on Beta-Systems (Guinea Pig Trachea and Atria) ...... 112

9. Inhibitory Activities of TMQ (50), S(+)-62 and R(-)-62 on U46619-Induced Human Platelet Aggregation ...... 113

- xi - LIST OF SCHEMES

PLATE PAGE

I. Synthesis of 78 and Related NMR Spectra in CDCI3 /TMS ...... 81

II. Synthesis of 5-trifluoromethyltrimetoquinol (63) .... 83

III. Synthesis of hydroxybenzoic acid 89 sideproduct .... 85

IV. Mechanism of formation of 89 . . . '...... 85

V. Synthesis of 8 8 ...... 86

VI. Initial Synthetic Approach to 8 -Trifluoromethyltrimetoquinol (63) 88

VII. Proposed mechanism of 3-aminoisoquinoline formation . . 89

VIII. Attempted synthesis of 1 0 0 ...... 90

IX. Attempted synthesis of 97 using isolated CFgCu ...... 92

X. Protoberberine formation using Mannich conditions . . . 94

XI. Separation of 104 via Preparative Chiralcel OD HPLC . . 98

XII. Deprotection to afford S(+)-62 and R(~)_62 ...... 99

XIII. Unsuccessful Synthesis of 5-Trifluoromethy1imidazoline 71 100

XIV. Synthesis of 5-Trifluoromethyl Catecholimidazoline 71 . 101

XV. Synthesis of Benzylnitrile 1 1 1 ...... 102

XVI. Synthesis of Imidazolines 70 and 7 2 ...... 103

XVII. Synthesis of the 6 -Trifluoromethylated Norepinephrine Analog 7 6 ...... 104

XVIII. Attempted Syntheses of 119 ...... 106

- xii - XIX. Synthesis of Catechols 121 and 1 2 2 ...... 108

XX. Synthetic Approach to 2-Trifluoromethy1 (74) and 5-Trifluoromethy1 (75) Norepinephrine Analogs ...... 109

XXI. Synthesis of chloropromazine hydrazinium analog 151 . . 181

XXII. Retrosynthesis of 1 5 2 ...... 182

XXIII. Synthesis of N,N-dimethylhydrazone 159 ...... 183

XXIV. Synthesis of N,N,N-trimethylhydrazonium iodide 160 . . . 185

XXV. Synthesis of Hydrazinium Derivative 1 5 2 ...... 189

XXVI. Attempted Synthesis of Dimethoxy dimethyldopamine (162) via Reductive Methylation ...... 192

XXVII. Synthesis of Dimethylhydrazonium 166 ...... 192

XXVIII. Attempts to the Synthesis of Hydrazinium 154 via Hydrogenation ...... 194

XXIX. Synthesis of Hydrazinium 167 and Attempted Deprotection ...... 195

XXX. Synthesis of Hydrazonium 170 and Attempts to Reduce with Platinum O x i d e ...... 197

- xiii - Part 1 SYNTHESIS OF FLUORINATED CATECHOLAMINE DERIVATIVES AS ADRENERGIC STIMULANTS AND THROMBOXANE A2 ANTAGONISTS

- 1 - CHAPTER I INTRODUCTION

1.1 SYMPATHETIC NERVOUS SYSTEM

The autonomic nervous system regulates the functions of the , glands, and at the unconscious level [1]. It is divided into two systems, sympathetic and parasympathetic, which are anatomically, physiologically, and biochemically distinct. Activation of the sympathetic nervous system leads to responses collectively known as the "fight or flight" reaction which includes tachycardia, increased inotropy, , , bronchodilation, and hyperglycemia [2]. The preganglionic neurons of the sympathetic pathway originate in the thoracolumbar segments of the spinal cord and in the sympathetic ganglia with postsynaptic neurons [1 ,2 ].

The postganglionic fibers then reach to all visceral structures of the thorax, abdomen, head and neck where they innervate end organs [3 ].

Communication between neurons as well as between neurons and effector cells in the autonomic nervous system is mediated via chemical . Norepinephrine, a catecholamine, is the secreted by sympathetic postganglionic nerve endings; therefore, these neurons are referred to as "adrenergic".

- 2 - The adrenal medulla is anatomically and functionally analogous to

the sympathetic ganglia [1]. The chromaffin cells of the adrenal

medulla are innervated by preganglionic fibers and the major product

which they secrete into the bloodstream is the catecholamine

epinephrine [1]. Epinephrine and norepinephrine produce very similar

biological effects. Thus, the sympathetic system and the adrenal

medulla are often collectively referred to as the "sympathoadrenal

system" due to their striking functional analogies in situations of

fright and stress [1 ].

1.1.1 Catecholamine Biosynthesis

The catecholamines, dopamine, norepinephrine and epinephrine, are synthesized in the brain, chromaffin cells, sympathetic nerves and sympathetic ganglia from the amino acid which is taken up from the bloodstream (Figure 1) [4]. First postulated by Blaschko [5] in

1939, the enzymatic sequence of catecholamine biosynthesis was confirmed by Nagatsu and co-workers [6 ] in 1964 with the discovery of the first step involving aromatic hydroxylation of tyrosine to

3,4-dihydroxyphenylalanine (DOPA) by the enzyme .

A component of catecholamine-containing neurons and chromaffin cells, tyrosine hydroxylase catalyzes the rate-limiting step in the sequence

91 and requires molecular C^, Fe and a tetrahydropteridine cofactor [4],

This enzyme is stereospecific and shows a high degree of substrate specificity oxidizing primarily L-tyrosine and to a smaller extent

L- [4]. L-DOPA is then converted to dopamine via

DOPA-decarboxylase which is a ubiquitous enzyme requiring pyridoxal phosphate (Vitamin Bg) as a cofactor. DOPA-decarboxylase removes

carboxyl groups from all aromatic L-amino acids; therefore, it is also

called aromatic amino acid decarboxylase. Dopamine /^-hydroxylase,

r% « located in the membrane of storage vesicles, is a Cu -containing

enzyme requiring O2 and ascorbic acid as a cofactor in the conversion

of dopamine to norepinephrine. Dopamine /?-hydroxylase does not show a

high degree of substrate specificity and in vitro will oxidize almost

any phenylethylamine to its corresponding phenylethanolamine [4]. In

the last step, norepinephrine is N-methylated to epinephrine in the

adrenal medulla via phenylethanolamine N-methyltransferase (PNMT) which

utilizes S-adenosylmethionine as a methyl donor. PNMT also shows poor

substrate specificity and will N-methylate a variety of /?-hydroxylated

[4].

HQ- ^ " C°’H hydroxyl.se HO^5^ "

L-Tyrosine DOPA H. OH

dopa dopamine

decarboxylase p-hydroxylase

Dopamine Norepinephrine

H OH phenylethanolamine H0v^ v 5 C s^NHCH3

N-methyltransferase HO^^^^

Epinephrine

Figure 1: Catecholamine biosynthesis 1.1.2 Storage and Release of Catecholamines

Norepinephrine is stored in synaptic vesicles (or granules) located in sympathetic nerve endings and adrenal chromaffin cells. The vesicles, which are a variety of sizes depending on location, are formed in the neuronal cell body and subsequently transported along the axon to the nerve terminal region [4,7]. These storage vesicles contain dopamine fi-hydroxylase and a high concentration of catecholamine complexed with triphosphate (ATP) in a 4:1 ratio. The anionic phosphate groups of ATP are thought to form a salt link with norepinephrine, which exists predominately as a cation at physiological pH, to bind the amines within the vesicle [4].

Chromagranin, an acidic protein also found in the storage vesicle, associates with the catecholamine:ATP complex to keep the neurotransmitter in a hypoosmotic state despite its high concentration

[7]. As well as a site for storage, the vesicles also protect norepinephrine from oxidation by monoamine oxidases.

Norepinephrine is released from the nerve terminal into the synapse upon nerve stimulation. The mechanism of catecholamine release has been studied primarily in the adrenal medulla. Here activity in preganglionic fibers causes release of which results in the influx of calcium. This influx of Ca causes the fusion of the storage vesicle and the chromaffin cell membrane resulting in exocytosis of catecholamines and all vesicle contents [7]. Whether the above applies to the sympathetic nerve endings is unclear but probable. Newly synthesized norepinephrine may be released preferentially,

which suggests that norepinephrine exists in more than one pool in the

neuron (4]. The concentration of catecholamines in the synapse

regulates its own release by interacting with presynaptic

. Other factors involved in catecholamine release

regulation include stimulation of inhibitory dopamine receptors,

inhibitory muscarinic receptors and inhibitory opiate receptors, all

found on presynaptic adrenergic nerve endings [8 ]. Prostaglandins of

the E series also inhibit norepinephrine release, although this action

is dissociated from any interaction with presynaptic receptors [4,8].

1.1.3 Metabolism and Reuptake

Once released into the synapse, catecholamines are initially

metabolized by monoamine oxidases (MAOs) and catechol-O-methyl-

transferase (COMT). The metabolic pathways are shown in Figure 2

[3,4]. Monoamine oxidases, flavin containing enzymes located in the

outer membrane of mitochondria, oxidize catecholamines to their

corresponding aldehydes. There are two types of MAO, termed MAO-A and

MAO-B, which differ in substrate affinities and specificity. MAO-A has a higher affinity for norepinephrine, such that the action of MAO-B becomes important only in the presence of high norepinephrine concentrations [9]. MAOs are typically considered intraneuronal enzymes; although, they are also found extraneuronally. The aldehyde product from MAOs may be further oxidized to the acid or reduced to the . COMT, an extraneuronal magnesium-dependent enzyme, catalyzes the transfer of a methyl group from S-adenosylmethionine to the m-hydroxyl group of catecholamines as well as various other catechol

con tain ing compounds.

OH CH,0 OH Sulfate or -► Glucuronide HO Conjugate

COMT

OH

HO .OH

HO Aldehyde Reductase

OH OH OH

HO HH HO NHCHHO MAO MAO

HO HO HO COMT COMT

OH OH OH

CH,0 NH HO CO.H

HO HO COMT

OH MAO MAO CH.O CO.H

HO Sulfate Sulfate or or Aldehyde Glucuronide Clucuronlde Dehydrogenase Conjugate Conjugate OH

CH.O .H

HO

Figure 2 : Metabolic pathways of catecholamines. (Modified from [3,4]) Conjugation is also an important metabolic fate of catecholamines.

Sulfoconjugation of the catecholamine metabolite 3-methoxy-4-hydroxy-

phenylglycol (MHPG) and other phenols is catalyzed via several types of

phenolsulfotransferase (PST) [9]. PST utilizes phosphoadenosine-

phosphosulfate (PAPS) as the sulfate donor [9]. And glucuronide

formation may occur via the transfer of active glucuronic acid to the

phenolic hydroxide of catecholamine metabolites catalyzed by uridine diphosphoglucuronyl transferase (UDPGT) [9].

The principal process for inactivating catecholamines released into

the synapse is a reuptake mechanism into the neuron. The reuptake is mediated by a Na+-dependent active transport mechanism as the neurotransmitter must enter the presynaptic terminal membrane and vesicle against a concentration gradient [4]. At least two types of uptake exists. The high affinity "Uptake l" is highly stereospecific for L-norepinephrine and operates after nerve transmission [10]. The low affinity "Uptake 2" is less stereospecific, reabsorbing amines structurally related to norepinephrine, and operates in the presence of higher concentrations of the neurotransmitter or structurally related compounds at the receptor [10]. Subsequent to reuptake, catecholamines may still be subjected to enzymatic degradation intraneuronally. 9

1.2 SUBCLASSIFICATION AND SIGNAL TRANSDUCTION OF ADRENERGIC RECEPTORS

In general, pharmacological specificity to a series of agonist and

antagonists is now the basis of receptor classification. Adrenergic

receptors were initially divided into two types, termed alpha "a" and beta "/[", as proposed by Ahlquist [11] in 1948 when he examined a series of sympathomimetic amines and the responses they elicited in a variety of tissues. The a-adrenoceptors were found to have the rank order potency of epinephrine > norepinephrine > a-methylnorepinephrine

> a-methylepinephrine > isoproterenol for stimulation of smooth muscle in the vasculature (vasoconstriction), stimulation of the uterus, nictating membrane, ureter and dilator pupillae, and inhibition of intestinal smooth muscle [11]- Conversely, the rank order potency observed for /[-adrenoceptors was isoproterenol > epinephrine > a-methylepinephrine > a-methylnorepinephrine > norepinephrine for inhibition of vascular, uterine and bronchial smooth muscle contraction and myocardial stimulation [11]. Subsequently, in 1967 Lands and co-workers [1 2 ] further separated /[-adrenergic receptors into /?^- and

/?2 ~adrenoceptor subtypes. /[^-Adrenergic receptors, found predominately in cardiac tissue, show the rank order potency of isoproterenol > epinephrine = norepinephrine. /^Adrenergic receptors show the rank order potency isoproterenol > epinephrine > norepinephrine and are found predominately in the smooth muscle of vasculature and bronchioles. There is; however, evidence for the coexistence of /[j- and /?2 _adrenoceptors in various tissues although 10

the relative percentage of the subtypes varies according to tissue

[13). The prototypic //-adrenergic agonist, isoproterenol (1), and the

//- prototype, (2 ), are both

non-selective (Figure 3). The synthesis of selective //-adrenergic

agonists and antagonists further supports the separation of ft and

/^-adrenoceptors (Figure 3).

In the past, //-adrenoceptors present in the adipose tissue were

classified as ft^, stimulation of which results in lipolysis (fatty

acid mobilization). Arch and co-workers [14] have described a group of

novel //-adrenergic agonists (Figure 4) which selectively stimulate

lipolysis in brown adipocytes. BRL 28410 (8 ), BRL 35113 (9) and

BRL 37344 (10) were found to be 21-, 28- and 400-fold more potent,

respectively, as agonists of brown adipocyte lipolysis than as

stimulants of atrial rate [14]. BRL 28410 (8 ), BRL 35113 (9), and BRL

37344 (10) also have more potent agonist activity on lipolysis than on

tracheal relaxation, whereas the opposite was true for isoproterenol

(1), (5) and (2) [14]. Selective ft and

/^-adrenergic antagonists, such as (4) and ICI 118,551 (6 ), were found to have a low affinity for the brown adipocyte

//-adrenoceptor as well [14]. These results conclude that atypical

//-adrenoceptors in rat brown adipocyte are of neither //j- nor ^ 2 “ subtypes [14].

Recently, Emorine and co-workers [15] have isolated a human gene which encodes for a third //-adrenoceptor they have termed 11

OH NH'^S ‘ w OH

non-selective 0-agonist non-salectlve 0-antagonist

2

OH Y'°'n^''— lmv ^' C 0 > r i W

selective 02-agonist selective 01-entegonist 3 4

OH

HO, NH OH NH CH. OH CH. OH

selective 02-agonist CH, 5 selective 02-antagonist OH OH 6

HO

selective 02-agonist

Figure 3 : /?^- and /^"Adrenoceptor agonists and antagonists

"/^-adrenergic receptor". Chinese hamster ovary (CHO) cells transfected with the cloned gene expressed binding sites which

[ I] labeled with a = 500 pM (almost ten times as large as that for fi^- or /^"adrenoceptors). BRL 37344 (10) is among the most potent /?-agonists in CH0-/?g cells and most of the classical 12

Ex E 2 8 - BRL 28410 H c o 2h

9 - BRL 35113 c f 3 c o 2h 10 - BRL 37 3 4 4 Cl o c h o 2h

Figure 4 : Atypical adipocyte-selective beta-adrenoceptor agonists

//-antagonists were ineffective against CHO-//g cells [15]. The

//^-adrenergic receptor was also found to have a higher sensitivity to

norepinephrine than to epinephrine suggesting that modulation of

//^-adrenoceptor activity via sympathetic innervation may be in

response to situations such as stress, high energy intake or cold

acclimatization [15]. This //g-adrenoceptor is suggested to be the

same as the atypical adipocyte //-adrenoceptor described by Arch and

co-workers [14].

In general, the //-adrenoceptors are postsynaptic receptors. The

presence of presynaptic //-adrenoceptors has been proposed which

provide a positive feedback mechanism for neurotransmitter release

during nerve stimulation [16]. Presynaptic //-adrenoceptors are

thought to exist on most peripheral adrenergic nerve terminals as well

as in some areas of the [17]. Circulating

epinephrine or epinephrine taken up and released as a cotransmitter with norepinephrine seem to activate the presynaptic //-adrenoceptors, as opposed to norepinephrine released from adrenergic nerve terminals 13

[17], Presynaptic /?-adrenoceptors are predominately /?2 -receptors

[17].

It was initially suggested by Lange [18] that a-adrenoceptors be

separated into OCj and otj subtypes based on their location in the

synapse. a^-Adrenoceptors were proposed to be postsynaptic

a-receptors which mediated the response of the effector organ, while

(^■adrenoceptors were proposed to be presynaptic and regulated

neurotransmitter release [18]. The above suggestion has not proven to

be accurate. Thus, Berthelsen and Pettinger [19] suggested that

a-adrenergic receptors be classified pharmacologically as opposed to

synaptic location and this is currently the basis of a-adrenergic

classification. In general, a^-adrenergic receptors are defined as

having a high affinity for the agonists (1J.),

(13), (14), and 6 -fluoronorepinephrine (1 2 ), potently

inhibited by the competitive antagonists (18) and

(15) and irreversibly blocked by alkylating agent (17)

(Figure 5) [20]. a 2 -Adrenergic receptors have a higher selectivity

for the agonists clonidine (.19), UK-14,304 (20), (21) and BHT

920 (£2) and are selectively antagonized by (24),

(25), (27) and (26) (Figure 6 ) [21]. ax Agonists: OH

a x Antagonists:

18

Figure 5: aj-Adrenoceptor agonists and antagonists. dj Agonists:

NH

Cl HN

19

Cl NH

H,C = > NH, ■CtH 21 22

(CH,),C

23

a 2 Antagonists:

NH

CH,0 CH.O.

OH OH 24 25 ocro 27 28

Figure 6 : (^-Adrenoceptor agonists and antagonists 16

Evidence suggests that a^-adrenergic receptors can be further

subdivided into pharmacologically distinct subtypes, although the exact

classification of these subtypes remains controversial. The following

is a discussion of the different methods utilized by several groups to

subclassify the a^-adrenergic system.

In 1982, McGrath [22] was the first to suggest that

(Xj-adrenoceptors might be comprised of two populations which he termed

a lfl and 0]^- His hypothesis was based primarily on the fact that dose

response curves produced by phenylethanolamines on rabbit basilar

artery [23] and rat anococcygeus [22] have a ’'shoulder” indicating two binding components. Conversely, imidazoline agonists produce no

response in rabbit basilar artery while producing a monophasic curve in rat anococcygeus. From this data, McGrath s\’.ggested that in rat anococcygeus the "low concentration" component of the phenylethanolamine and the imidazoline response is mediated by

(X ^-adrenoceptors while the response to high concentrations of phenylethanolamines and the lack of response by imidazolines was attributed to a^-adrenoceptors [22]. McGrath and co-workers [24,25] later elaborated this hypothesis, asserting that the ra^a-subtype was

24* 94- susceptible to Ca antagonists thus associated with opening of Ca channels while activation of the a^-subtype resulted in release of

2+ Ca from intracellular stores.

Looking at affinity studies of oc^-adrenoceptors in rabbit and rat vascular smooth muscle of blood vessels, Flavahan and Vanhoutte [26] 17

noted a variation in affinities for prazosin (lj>) and yohimbine (24)

They suggested the existence of two a-adrenoceptor subtypes, one

subtype had high affinity for both prazosin (.15) and yohimbine (24)

while the second subtype had a low affinity for both . Support

for this observation, however, remains uncertain [2 0 ].

Looking at the competition of the antagonists phentolamine (_18) and

WB4101 (.16) on [^HJprazosin (15) binding sites, Morrow and Creese [27]

have characterized a.^-adrenergic subtypes which they termed and

Otjg. The site with the highest affinity for phentolamine (1JJ) is the

ot ^-adrenoceptor versus the low phentolamine affinity site, the

a 1B-adrenoceptor. Additionally, [^H]WB4101 (16) will label only the

oc^ site. The rank order of antagonist affinity for was WB4101 >

prazosin > phentolamine > indoramine > . In

contrast, for the rank order was prazosin > indoramine >

dihydrocryptine > WB4101 > phentolamine [27].

Work from Minneman's laboratory also proved the existence of two distinct ot^-adrenoceptors by observing differences in inactivation of

125 [ I] BE (29) labeled a^-adrenergic receptor binding sites by the

irreversible alkylating agent chlorethylclonidine (CEC, 30) [28].

cl

29 30 18

They found that only half of the otj-adrenergic receptor binding sites

in rat cerebral cortex were CEC sensitive, whereas CEC pretreatment of I O C hippocampus membranes showed no inactivation of [ I]BE labeled

a-adrenergic binding sites [28]. Futhermore, CEC (30) caused a

virtually complete inactivation of binding sites in rat liver and

spleen, whereas 0£j binding sites in vas deferens were essentially

unaffected [29]. This suggests the presence of two oc^-subtypes in

mammalian tissues, one CEC sensitive and the other CEC insensitive

[29].

Han et al. [30] provided further evidence of two subtypes of

(Xj-adrenoceptors which also supports the hypothesis of Morrow and

Creese [27]. Moreover, Han and co-workers described functional differences attributed to the individual a^-subtypes. The site with high affinity to WB4101 (1^6), termed a^a , causes contractile responses by controlling the opening of dihydropyridine-sensitive channels via an OX unknown mechanism resulting in an influx of extracellular Ca [30].

The Ofjk subtype, with low affinity to WB410] (16), stimulates inositol phospholipid hydrolysis resulting in contraction independent of extracellular Ca^+ [30]. Recently, Cotecchia and co-workers [31] cloned the cDNA of the hamster a^-adrenergic receptor which has low affinity for WB4101 (1J>) and phentolamine (1JJ) and couples to inositol phospholipid metabolism; and therefore, is considered of the a ^ subtype. 19

In purifying a^-adrenergic receptors >500-fold in rat liver and

brain membrane, Terman and co-workers [32] again found 100% of

[ HJprazosin (15) binding sites in partially purified preparations of

rat liver to be inactivated by CEC, while only 50% of the brain

receptors were inactivated by CEC. And whereas the CEC-insensitive

125 sites in the brain were resistant to [ IJazidoprazosin photoaffinity

labeling («^a), photoaffinity labeling of partially purified liver and

brain samples not treated with CEC resulted in the specific labeling of

a Mr 80,000 protein (<*1 5 ) [32]. These reports provide additional

evidence that heterogeneity of aj-adrenergic receptor binding observed

in membrane preparations is due to structurally distinct receptor

subtypes and not due to differences in membrane components that

interact with the receptors [32]. The individual characteristics of the a^fl and subtypes are summarized in Table 1 [32].

Table 1

Characteristics of Alpha-1 Adrenergic Receptor Subtypes

Characteristics ‘lb

Tissue Distribution Brain, Aorta, Vas deferens Liver, Spleen Binding

VB 4101 high affinity low affinity Phentolamine high affinity low affinity

Methoxamine Activation + + + + CEC Sensitivity No Yes

[1 2 5 I]Azidoprazosin No Yes photolabeling 20

And last, Schwinn and co-workers [33] have recently reported a novel otj adrenoceptor subtype cloned from bovine brain cDNA. Localization of the bovine a^-adrenergic receptor to human chromosome 8 provided evidence that it is distinct from the hamster a^-adrenergic receptor which localizes to human chromosome 5 [33]. And although the bovine aj-adrenergic receptor subtype shares similar binding properties with the

(16) and phentolamine (18) and agonist (23) as compared to the Otjk receptor), it is sensitive to CEC inhibition and is not expressed in tissue where the 0Cja subtype exists (i.e. vas deferens and hippocampus) [33]. This data suggests a third a j-adrenergic. receptor subtype exists.

Bevan and co-workei;s [34,35] have suggested that differences in sensitivity of a^-adrenoceptors to agonists may be due to differences in agonist affinity as opposed to -structurally distinct receptor subtypes. Based on variations of norepinephrine affinity to aj-adrenoceptors in a variety of tissues, Bevan et al. have proposed the 'variable receptor affinity hypothesis' [34], This hypothesis suggests that affinity variation of both agonists and antagonists is a function of receptor modification induced by microenvironmental and intracellular processes [35].

aj-Adrenergic receptors exist postsynaptically, while (/^“adrenergic receptors are found both presynaptically and postsynaptically as well as extraneuronally and on organelles lacking [36], Presynaptic (^“adrenoceptors, found on adrenergic nerve terminals in

the periphery and the central nervous system, serve as a negative

feedback mechanism to modulate neurotransmitter release [21,36]. In general, postjunctional vascular a^-adrenoceptors are located at the

neuroeffector junction as opposed to the postjunctional vascular 0 - 2 adrenoceptors which are predominately located extrajunctionslly [37].

Ruffolo et al. [37,38] have pharmacologically discriminated between pre- and post-junctional (^-adrenoceptors both in vivo and in vitro using the antagonist SK&F 104078 (26). SK&F 104078 (26) selectively inhibited postjunctional vascular « 2 -adrenoceptors in a variety of systems, whereas it had no effect on prejunctional o^-adrenoceptors even at concentrations up to 10 //M [37], suggesting heterogeneity among a 2 _adrenergic receptors.

Bylund [39] and Nahorski et al. [40] both independently subclassified the C^-adrenergic system into anc* a2B receptor subtypes. The prototype tissue for the receptor is the human platelet where it has a low affinity (K^ = 240 nM) for prazosin (15) O inhibition of [ HJyohimbine (24) binding [39]. Alternately, the O^g receptor has a high affinity (K^ = 5 nM) for prazosin (15) in the rat neonatal lung [39]. Also, yohimbine (24) and rauwolscine (25) have similar affinities on the &2k subtype, whereas on the 0(23 subtype rauwolscine (25) is more potent than yohimbine (24) [40]. Looking at

O the ability of various drugs to inhibit [ H] antagonist binding on the human platelet, human cerebral cortex and HT29 cell line (tissues containing the c t ^ receptor subtype only) and on neonatal rat lung and

NG108-15 cell line (tissues containing the receptor subtype only),

Bylund et al. [41] found several subtype selective compounds. The

agonist oxymetazoline (23) is approximately 30-fold more selective for

(*2 ^ receptors, while the antagonists ARC-239 (3^), (32)

and 7-hydroxychlorpromazine (33) are 1 0 0 -, 18-, and 17-fold,

respectively, more selective for OC2 3 receptors (Figure 7) [41].

(CH HjC CHj (CH3)2N , 1 ^

31 32 13

Figure 7 : Otjg-Adrenoceptor selective compounds

The gene for the human platelet O^adrenergic receptor has been cloned by Kobilka et al. [42]. Expression of this gene in Xenopus laevis oocytes results in a receptor which binds with high and equal affinity to yohimbine (24) and rauwolscine (25) and has low affinity for prazosin (15) and thus classified as an «2 ^_adrenergic receptor

[42]. The human platelet a 2 _adrenergic receptor is derived from chromosome 10 and may be referred to as the OC2 "C1 0 receptor [42,44]. A third (X2 “adrenoceptor subtype has been proposed on the basis of

the characterization of C^-receptors in OK cells which are an opossum

cell line [43]. Although the « 2 “adrenoceptors in the OK cells

have a pharmacological profile similar to the «2 g subtype, the

affinity of the receptor to yohimbine (24) is in the range expected for

the a2A subtype and thus it is considered to be a distinct a.2 Q subtype

[43,44]. And last, the results from the cloned gene of an

O^-adrenoceptor from the human kidney may represent a fourth subtype

[45]. The human kidney (^-receptor gene has been localized to human

chromosome 4 and is referred to as a 2 “C4 [45]. The a2 “C4 has a much

higher affinity for prazosin (15) than to o^-ClO. Conversely,

oxymetazoline (23) has a much lower affinity for the 0C2 "C4 than for

a 2 -C10 [45]. These results initially suggested the

be of the O^j^subtype [45]. The correlation of affinities

between the expressed 0(2 “C4 receptor and o^g and receptors;

however, proves that the 0t2 “C4 receptor is of neither the O^g nor the

1*20 subtypes and may represent a distinct OC2 Q [41,44].

Recently evidence of an "imidazoline binding site" has been reported

[46,47]. The imidazoline binding site is a non-adrenergic site and although often associated with 0(2 ~a^rener8 ic receptors, it is pharmacologically distinct [46,48]. Many potent (^“adrenergic agonists and antagonists contain an imidazoline moiety making it

important to determine which, if any, of the physiological responses 24

induced by these compounds are due to activity on the imidazoline

binding site.

The mechanism by which adrenergic receptors mediate their functional

response varies. The signal transduction mechanism of a^-adrenergic Ol receptors involves a rise in intracellular free Ca (Figure 8 )

[20.49]. Stimulation of a^-adrenoceptors causes the activation of

membrane bound phospholipase C via a guanine nucleotide binding protein

(G-protein) termed Gp. Details of the Gp protein are uncertain;

however, it is known that Gp is not sensitive to Bordatella pertussis

toxin [20,50]. Activated phospholipase C catalyzes the hydrolysis of phosphatidylinositol-4,5-biphosphate (PIP2 ) to give the secondary messengers, inositol 1,4,5-triphosphate (IP3 ) and diacylglycerol (DAG)

(Figure 9) [49]. Via receptor activation, IP3 then mediates the 2+ release of intracellular Ca stores from the endoplasmic reticulum

[49] . IPg may also be phosphorylated via inositol triphosphate kinase to inositol 1,3,4,5-tetrakisphophate (IP^) [49]. IP^ may function as a secondary messenger which positively regulates the opening of cell

2+ 2+ surface Ca channels allowing the influx of extracellular Ca

2+ [20.49]. The free cytosolic Ca binds to calmodulin and the

2+ Ca -calmodulin complex then activates a variety of enzymes which act on proteins leading to various physiological responses [49] .

Diacylglycerol activates protein kinase C which phosphorylates specific cellular proteins also leading to physiological responses [49]. It is interesting to note that some aj-adrenergic agonists activate

2+ extracellular Ca influx only, while others also promote mobilization

2+ of intracellular Ca , although the reason for this is unclear [49]. 2 25 Ca

□ Oj agonist

1 i—i extracellular

PIP PIP

PLC intracellular (+)

Ca DAG

(+)

Protein Kinase C

Ca 2 +

endoplasmic reticulum Phosphorylation of Specific Proteins

n 2* L a /Calmoduli Kinases

Selected Responses

Substrates ^ Integrated Responses

Selected Responses

Figure 8 : Mechanism of signal transduction upon aj-adrenoceptor stim­ ulation. (Modified from [49]).

Activation of /J-adrenergic receptors results in stimulation of adenylate cyclase via the G protein termed G . Conversely, activation of (^-adrenoceptors causes inhibition of adenylate cyclase activity via coupling to the G protein termed G^. Both G proteins are membrane-bound, cytosolic heterotrimers comprised of an a, /?, and y subunit [51]. Typically, the oc subunit is the active component of the heterotrimer and it dictates the specificity of the interaction with 26 o ii OC(CH2)1sCH 0 II opo3 OC(CHj)16CH3 0 0 • • Phospholipace C O-P-O' II i R 1CO 0 HO OH

IP HO OPOj DAC 3 1 OPOj'

PIPJ

Rj - Arachldonic acid

Figure 9 : Hydrolysis of PIP2 by phospholipase C.

receptors and effectors [51,52]. Therefore, several different a

subunit subtypes exist including a small (GSa g , 45kDa) and a large

(GSw j, 52kDa) form of Gga [51] and several types of G ^ (40-41kDa)

[53]. The differences between Gga_g and Gga j are small; both couple

efficiently to /?-adrenergic receptors and both stimulate adenylate

cyclase and activate Ca^+ channels [51]. The three types of G^

collectively inhibit adenylate cyclase and activate atrial K+ channels

[50,53]. The Gga component contains a high-affinity binding site for guanosine triphosphate (GTP) and has intrinsic GTPase activity [54].

Gga is sensitive to ADP-ribosylation by Vibrio cholerae which results in irreversible activation of adenylate cyclase activity by inhibition of the G ga GTPase activity [50]. On the other hand, G^a undergoes

ADP-ribosylation by Bordatella pertussis which interferes with the coupling of G^b to the receptor resulting in increased cellular cAMP levels [50]. The (} (35-36kDa) and y (lOkDa) subunits form a tightly 27

bound component which is thought to be responsible for anchoring the G

protein into the membrane cytoplasmic surface [50]. The general

mechanism of signal transduction mediated by G protein coupled

receptors is shown in Figure 10. In the basal state, the a subunit is

bound to guanosine diphosphate (GDP) and furthermore the a subunit is

complexed with the fly subunits [50]. Agonist binding to the receptor

results in receptor coupling to the G protein complex [52]. The

receptor-G protein interaction causes a conformational change which

2+ allows GTP, in the presence of Mg , to replace GDP on the a subunit

[52]. The activated a-GTP subunit dissociates from the fly subunits

and interacts with the effector (such as adenylate cyclase) [52]. The

intrinsic GTPase activity of the a subunit hydrolyzes GTP to GDP

liberating inorganic phosphate (Pi) (Figure 11), and the a-GDP recombines with the fly subunits to end the activation cycle (Figure

10) [52].

With respect to both f l and /^-adrenoceptors, activation by

//-agonists results in receptor coupling to the G_ protein (Figure 12).

This results in stimulation of adenylate cyclase which catalyzes the conversion of ATP to cAMP (Figure 13). cAMP activates cAMP-dependent protein kinases which promote phosphorylation of specific proteins which leads to the functional responses (Figure 12) [50]. Conversely, agonist stimulation of (^-receptors causes coupling to the G^ protein

(Figure 12) [49]. This causes inhibition of adenylate cyclase and a resultant decrease in the basal concentration of cAMP [49]. A decrease 28

STEP 3

STEP 1

GOP

Ef factor

OTP GOP

OTP

STEP 2

Figure 10: Mechanism of G Protein coupled signal transduction. (Modi­ fied from [52])

In cAMP levels alone may not account for all responses attributed to

(^-adrenoceptor stimulation [49]. Stimulation of presynaptic

0C2 _adrenoceptors causes inhibition of neurotransmitter release by Ox blocking the influx of Ca through cAMP-dependent voltage-dependent

2+ Ca channels [49,55]. Human platelet O^-adrenoceptor induced aggregation and secretion is mediated by arachidonic acid release which may involve a Na+/H+ exchange process [55]. Postsynaptic a 2 _mediated Ol vasoconstriction may involve extracellular Ca influx. Other 29

CTPase + P. 0 —p— 0 —P— 0 HO — P— 0 — P— 0

OH OH OH OH

GTP GDP

Figure 11: Hydrolysis of GTP by G protein-GTPase.

mechanisms may also be involved in mediating the functional responses

of (^-adrenoceptors [49].

In summary, the adrenergic receptors belong to a family of receptors which are linked to G-proteins. The G-proteins link the receptor to secondary messengers which then mediate the effector responses. Both

/?j- and /^-adrenergic agonist activation stimulates adenylate cyclase activity. In contrast, agonist activation of 0C2 -adrenergic receptors inhibits adenylate cyclase activity. The a^-adrenoceptors stimulate phosphoinositide turnover. 30

P agonist Oj agonise Q a * *

Voltage Dependent run (N-type) Ca2* channel

AC

ATP 2 * cAMP Ca Phosphodiesterase^-' (+)

5'AMP cAMP dependent protein kinases ATP

ADP Phosphorylated -Protein Protein

1 Tissue Response

Figure 1 2 : Mechanism of /?-adrenoceptor and (^-adrenoceptor signal transduction. (Modified from [50]).

NH NH

Adenylate

0 —P— 0

OHOH OH

ATP cAMP

Figure 13: Conversion of ATP to cAMP by adenylate cyclase. 31

1.3 STRUCTURAL PROPERTIES OF ADRENERGIC RECEPTORS

The adrenergic receptors are homologous proteins, each composed of a

single polypeptide chain [56]. The a^-receptor polypeptide contains

515 amino acids [31], the a 2 -receptor contains 450 amino acids [42],

the /?j-receptor contains 477 amino acids and the /^-receptor is 413

amino acids in length [56]. This discussion will focus primarily on

/?2 ~adrenergic receptors as the most information is known about this

adrenergic receptor. The topography of the /^-adrenergic receptor

includes seven hydrophobic a-helical transmembrane regions which are

linked by alternating intracellular and extracellular hydrophilic loops

as shown in Figure 14 [57]. The positively charged amino acids Lys®®,

Arg*®*, Lys*^, Arg^*, Lys^®, and Arg®^®, are thought to demarcate

the cytoplasmic loops from the seven hydrophobic membrane-spanning

regions [56]. The /^-receptor contains two sites (Asn® and Asn*®) of

N-linked glycosylation near the amino terminus [56,57]. Although glycosylation is not important for ligand binding, it may play a role

in the expression of the receptor at the cell surface [57].

The extracellular surface is thought to be involved in ligand binding. Recently, Dohlman and co-workers presented evidence for the

involvement of disulfide bridging in ligand binding at the

/?2 _adrenoceptor [58]. They found that /^-adrenergic receptors treated with dithiothreitol (DTT) incorporated a much higher concentration of

[*^C]iodoacetamide versus untreated /^-receptors [58]. Thus, most of

the cysteines appear to be involved in disulfide bonds [58]. 32

HH. S-S

Extracellular

Intracellular

CO-H

Figure 14: Topography of the /^-adrenergic receptor. (Modified from [57J).

Furthermore, DTT was found to inhibit binding of IOC [ ‘ I]iodocyanopindolol to /^-adrenoceptors [58]. Using site-directed

mutagenesis in which valine was substituted for cysteine residues

located within the transmembrane segments and within extracellular

loops (eight substitutions in total), Dohlman et al. found that

replacement of cysteines (Cys^®^’ 150,191) wfthin the extracellular domains resulted in a marked decrease in ligand binding [58]. This is supported by earlier evidence reported by Dixon and co-workers in which of cysteine residues 106 and 184 caused a significant change

in agonist binding [59]. Taken together, this evidence suggests that 33

two extracellular disulfide bonds (one between Cys*®® and Cys*®^ and

the second between Cys*^® and Cys*®*) are important for the functional

integrity of the ligand binding site. This may be either by

contributing to binding specificity or via maintenance of the correct

structural topography of the receptor for ligand binding [58].

Through mutant hamster /^“adrenoceptors} Dixon et al. found that

most of the extracellular and intracellular hydrophilic residues could

be deleted without affecting ligand binding to the receptor [60] .

However, mutants lacking regions encompassing transmembrane helices six

and seven were completely devoid of binding capability [60]. Thus, it

is generally accepted that the seven transmembrane domains are arranged

in a cluster forming a hydrophobic core which acts as the ligand binding site [57,60]. Within this hydrophobic core, it is believed that the protonated moiety of ligands interact with the 113 carboxylate group of Asp in the third hydrophobic helix of the

/?2 "fldrenoceptor [61]. Substitutions of Asp**® resulted in a decreased affinity for both agonists and antagonists [61]. Conversely, 79 substitution of Asp in helix two resulted in a decrease agonist affinity but had no effect on antagonist binding, suggesting that A s p ^ may be involved in maintaining the agonist-bound conformation of the

/?2 receptor [61]. Further support for the hypothesis of the ionic

1 13 interaction between Asp and the ligand amino moiety is the fact that this Asp residue in helix three is conserved in all G protein coupled receptors which bind amine ligands [57]. Adrenergic agonists require the presence of a catechol ring for activity which may be involved in 34

both hydrogen binding and hydrophobic interactions in the ligand

binding site [57). Strader et al. [62] have identified two

residues, at positions 204 and 207 in the fifth helix, which are

critical for agonist binding and efficacy at the //-receptor. The data

suggests that the interaction involves two hydrogen bonds, one between

Ser^®^ and the meta-hydroxyl group of the ligand and the other between 207 Ser and the para-hydroxyl group of the ligand [62]. It is

interesting to note that these two serine residues are conserved among

all G protein coupled receptors which bind catecholamine agonists, but

are not conserved in G protein-coupled receptors whose endogenous

ligand lacks a catechol moiety [62]. Hydrophobic residues may also be

important for binding of the aromatic portion of adrenergic ligands.

non Again using mutant receptors, Strader et al. found that Phe and 290 326 Phe in helix six and Tyr in helix seven are important for

//-agonist binding [57,63]. The above information provides a model for the ligand binding site of the //-adrenergic receptor as shown in

Figure 15 [57].

Using computer-assisted molecular modeling of the hamster

/^■adrenoceptor, IJzerman and van Vlijmen [64] have described a putative binding site for S-propranolol which is contrary to the above mutagenesis studies. In this model, IJzerman and van Vlijmen suggested that the protonated amine moiety forms a reinforced ionic bond with 306 113 Glu in helix seven, as opposed to Asp in helix three proposed by

Strader et al [61]. The computer assisted model also proposes that the

//-hydroxyl group of S-propranolol forms a hydrogen bond with the 35

VII

,0H - -0

OH' *°

OH

Figure 15: Extracellular view of the beta-adrenergic binding site. (Modified from [57]).

backbone carboxamide of Ile^®^ in helix seven [64]. And the third

point of attachment in this model involves a charge transfer

*11 O interaction as well as a hydrophobic interaction between Trp in

helix seven and the naphthyl ring moiety of S-propranolol [64].

Creating a series of chimeric o^. "adrenergic receptors, Kobilka

et al. [65] found that the seventh hydrophobic domain or a combination of the sixth and seventh hydrophobic domains of the « 2 " and

/?2 _adrenergic receptors is the major determinant of both agonist and antagonist ligand binding specificity. Furthermore, several of the

first five hydrophobic domains may also confer agonist binding specificity [65]. The results of the chimeric receptors also suggested

that at least portions of the region spanning from the fifth 36 hydrophobic domain through the third cytoplasmic loop and to the carboxy-terminal portion of the sixth hydrophobic domain may be

required for determining the specificity of /^-adrenoceptor coupling to G [65]. Whereas hydrophobic domains 1, 2, 3, 4 and 7, the first two cytoplasmic loops and the carboxy-terminus have little effect on Gg coupling specificity [65] . Other studies have shown that the amino terminus of the third cytoplasmic loop as well as a portion of the carboxy terminal region of the third loop are involved in G protein interactions [56,66]. Similarly, Cotecchia et al. [67] have recently reported that the third cytoplasmic loop of the a j-adrenoceptor is involved in coupling to phosphatidylinositol breakdown.

There are several mechanisms of /7-adrenoceptor desensitization, which is the response attenuation upon prolonged exposure to agonist stimulation. The first mechanism includes sequestration of receptors to some intracellular component [57]. This occurs within minutes upon receptor stimulation. A second mechanism of rapid desensitization is an agonist-induced uncoupling of the receptor from the G protein as a result of receptor phosphorylation. There are two sites of phosphorylation by located on the human

^-adrenoceptor, one is located in the carboxy-terminal region of the third cytoplasmic loop and the second is in the proximal portion of the carboxy-terminal tail [68]. The distal portion of the carboxy-terminal tail is serine and threonine rich. It is at these sites that a novel

/7-Adrenergic Receptor Kinase (/7ARK) phosphorylates the receptor and conformationally alters the /7-receptor in such a way as to interrupt 37

Gs coupling [68]. Lefkowitz et al [68] have proposed that at low

agonist concentrations (nanomolar), the cAMP-dependent protein kinase A

is activated. And conversely, at high agonist concentrations

(micromolar) /?ARK is also activated [68]. Down-regulation, which

takes hours, is another mechanism of desensitization and may involve

receptor degradation [57].

Although the « 2 _adrenergic receptor (human platelet) is

structurally similar to the ^-adrenergic receptor (i.e. seven

hydrophobic transmembrane regions, two N-glycosylated sites near the

amino terminus, etc.), there are several distinct differences. The

third cytoplasmic loop of the (^-adrenergic receptor is relatively

large comprised of 156 amino acids. In contrast, the and

/?2 -adrenergic receptors which have comparatively short third

cytoplasmic loops of 78 and 52 amino acids, respectively [42]. And

although the O^-adrenoceptor is phosphorylated in an agonist-dependent

fashion by /?ARK [69], it has a very short carboxy terminal tail (21

amino acids versus 84 amino acids for /^-adrenoceptor) which is devoid of serine or threonine residues [70]. The third cytoplasmic loop of

the o^-adrenoceptor; however, contains a total of 19 serine and threonine residues, including a sequence of four contiguous , which may represent the phosphorylation sites [31]. So, in conclusion, the adrenergic receptors are very similar in structural topography; all believed to resemble the most characterized /^-adrenergic receptor described above. 38

1.4 STRUCTURE ACTIVITY RELATIONSHIPS OF ADRENERGIC RECEPTORS

1.4.1 Alpha Adrenergic Agonist Structure Activity Relationships

There are two major classes of a-adrenergic agonists,

/^-phenethylamines and imidazolines. Although many exceptions exist,

phenethylamines are generally nonselectlve or show selectivity for the

a^-adrenoceptor while imidazolines typically are either nonselective or show selectivity for O^adrenoceptors [71]. Furthermore, phenethylamines are generally full agonists, whereas imidazolines in most instances are partial agonists [72]. Again, exceptions to this

rule are found. The structure activity relationships (SARs) for

/?-phenethy lamines versus imidazolines are greatly different and therefore will be discussed separately. Although there are structure activity differences between a^- and o^adrenoceptors, the precise relationships are still largely unknown.

In adrenergic systems, it has been found that R(-)-norepinephrine is more potent than the S(+)-isomer which is equipotent to the analog (dopamine). The order of potency for all //-phenethylamines on a-adrenoceptors follows the Easson-Stedman hypothesis [73] which proposes a three-point attachment model of norepinephrine to the a-adrenergic receptor (Figure 16). (Note: Easson and Stedman's initial studies were performed using epinephrine.) The points of interaction include the substituted phenyl ring, the hydroxyl group at the asymmetric ft carbon and the amine moiety. It was proposed that 39

only R(-)phenethylamines have the optimum stereochemical configuration

of all three functional groups and thus are the most potent in the

series. Conversely, in the S( + )-isomer and in the /?-desoxy

derivative, the /?-hydroxyl group is either oriented away from the

binding site or absent. Thus, both the S(+)-isomer and the desoxy

derivative operate through a two-point attachment model which accounts

for their decreased potency relative to the R(-)-isomer as well as the

fact that they are equal in activity to each other. In summary, the

Easson-Stedman hypothesis predicts the order of potency for

/J-phenethy lamines to be R(-) > S( + ) = desoxy.

OH OH OH ,0H OH OH

HO H OH . .

HH HH

Figure 16: The Easson-Stedman hypothesis model. (Modified from [72]).

Interestingly, it has been proposed that o^-adrenergic receptors, but not oc^-adrenergic receptors, may also contain an additional 40

a-methyl recognition site [74]. Both 2S(+)-a-methyIdopamine and

lR,2S(-)-erythro-a-methylnorepinephrine were highly selective for the

(^-adrenoceptor; however, 2R(-)-a-methyldopamine was nonselective

[74], Furthermore, the 2S(+)-isomer of a-methyIdopamine was more

potent at 0£2 ~adrenoceptors than either the 2R(-)-isomer or dopamine,

which were equipotent [75]. This suggests that while

2R(-)-a-methyIdopamine and dopamine bind via two points of attachment

(Easson-Stedman) at a 2 _adrenoceptors, 2S(+)-a-methyIdopamine may bind

via three-points of attachment involving the catechol, the amine, and

the appropriately oriented a-methyl group [74].

With respect to aromatic substitution of //-phenethylamines, the

3,4-dihydroxy substituted derivatives are typically the most potent

[72]. The rank order potency for aromatic hydroxyl substitution is 3,4 diliydroxy > meta hydroxy > para hydroxy > nonphenolic and holds true at both a.j- and a 2 -adrenoceptors [72]. One difference to note is that relative to the catechol, the monophenolic //-phenethylamines have lower affinities but not lower efficacies on a^-adrenoceptors.

Conversely, on a 2 -adrenoceptors, monophenolic //-phenethylamines have both lower affinities and efficacies [72]. Thus the catechol moiety appears to be critical for a 2 _agonist activity. This trend is seen in methoxamine (13), containing a 2,5-dimethoxysubstituted aromatic ring as opposed to a catechol, which is a selective a^-agonist and essentially inactive on a 2 -adrenoceptors [72]. Sulfonamido and 41

halogens have been Investigated as hydroxyl replacements on the

aromatic ring. Replacement of the meta hydroxyl in epinephrine of

norepinephrine with a methanesulfonamido group retains potent

otj-adrenergic activity; however, replacement' of the para hydroxyl with

a methanesulfonamido group abolishes activity [72]. Furthermore,

increasing the bulk of the meta sulfonamido group to an

n-butylsulfonamido derivative dramatically decreases agonist activity

and activity is completely abolished with aryl sulfonamido substitution

[72]. Replacement of the catechol hydroxyls with either 3,4-dichloro or 3,4-difluoro substitution renders less active a^-agonists.

Substitution of aromatic hydrogens on norepinephrine with fluorine provided a series of compounds with very interesting biological activity which will be discussed later. And last, although epinephrine

(N-methylated) is a more potent a-adrenergic agonist than norepinephrine, as the amine is substituted ' with larger normal and branched groups, a^-agonist activity is decreased until activity is abolished with n-butyl substitution [75]. Activity is regained with the introduction of an aryl containing N-substituent as in

(34) which has oc^-agon.ist activity [75].

34 42

Imidazolines do not follow the Easson-Stedman hypothesis. Looking

at the enantiomers of 2-(3,4,a-trihydroxybenzyl)imidazoline and the

corresponding desoxy derivative 2-(3,4-dihydroxybenzyl)imidazoline on

a^- and a 2 _adrenergic systems, Ruffolo et al. [76] found the rank

order potency on a^-adrenoceptors to be desoxy R(-) > S( + ) (Figure

17). The same rank order potency was observed on (^"adrenoceptors;

however, the desoxy analog was found to a partial agonist [76].

desoxy

Figure 17: Fisher projections of imidazolines used to test the Easson- Stedman hypothesis. (Modified from [76]).

Although .it is believed that //-phenethylamines and imidazolines

interact differently at a-adrenergic receptors, there are similarities between the most energetically stable conformations of norepinephrine and clonidine as shown in Figure 18 [75,77]. Theoretical calculations

indicate that the preferred conformation of R(-)-norepinephrine is in the extended trans conformation [78]. It has been proposed that imidazolines such as clonidine can assume a conformation which mimics the distance of the phenyl ring and amino moiety of norepinephrine

[75]. 43

C l \ HO

HO OH 1.2-1.4 A 1.28-1.36 A 5.0-5.1 A

Figure 18: Proposed conformations of norepinephrine and clonidine for alpha-adrenergic receptor binding. (Modified from [75,78]).

Using clonidine (1JJ) as the prototype (a phenylaminoimidazoline),

structure activity relationships for the imidazoline focuses on the

modification of 1) the aromatic ring, 2) the nitrogen bridge and 3) the

imidazoline moiety:

1) Modification of the Aromatic Ring

Replacing the aromatic ring moiety of phenylaminoimidazolines with a

substituted thiophene ring, as in tiamenidine (35), reduces a-agonist

activity [75]. Conversely, replacing the phenyl ring with other

heteroaromatic rings such as a substituted quinoxaline and a

substituted pyrimidine as in UK-14,304 (20) and (36)

respectively, provides potent a-agonist activity (Figure 19) [75]. It

is also interesting to note that an aromatic ring moiety is not critical for activity as S3341 (37) has nonselective a-agonist activity and was recently marketed in France as rilmenidene [75].

With respect to halogen substitution on the phenyl ring of phenylaminoimidazolines, addition of a 2-chloro atom dramatically 44

CH Cl JL ^HH ^ H HH

H,C N OCH,

35 36 37

Figure 19: Phenyl ring substituted clonidine-like compounds.

increases (^“agonist activity [77]. The addition of a second chlorine

atom is most effective at the 6 -position, as in clonidine (19), although 2,3-, 2,4-, and 2,5-dichloro derivatives also increase potency

[77]. A 3,5- or 3,4- dichloro substituion pattern decreases activity, emphasizing the importance of the 2-substituted derivative [77]. In general, chlorine may be substituted with fluorine, bromine, ethyl or methyl groups with only minor losses of potency [75,77]. One important point to make is that although the alkyl-substituted derivatives retain potent a 2_agonist activity at the receptor level, the dihalogenated derivatives provide better candidates for antihypertensive agents.

This is due to their greater lipophilic properties and thus their ability to distribute across the blood brain barrier where the central o^-agonist activity mediates the peripheral hypotensive effects [7 5 ].

And last, like //-phenethylamines, 3,4-dihydroxy substitution of either phenylaminoimidazolines of benzylimidazolines provides the most efficacious aromatic hydroxyl substitution pattern [77]. 45

2) Substitution at the Nitrogen Bridge

With respect to a 2 _adrenergic agonist activity, a nitrogen atom

separating the phenyl and imidazoline rings provides the optimum

activity [77], Substitution of the nitrogen bridge with either a

methylene group or a sulfur atom slightly decreases 0 - 2 activity,

whereas substitution with an oxygen atom completely abolishes activity

[79]. With respect to a^-agonist activity, nitrogen may be replaced

with a methylene group with no loss of potency [72]. Increasing the

bridge length between the phenyl and imidazoline rings generally decreases agonist activity, with the exception of (38) [75].

Although the reason is unclear, lofexidine (38) is a non-selective a-agonist which is as potent as clonidine (JL9) [75].

Cl HN

CH Cl

38

In the benzylimidazoline series, hydroxylation of the methylene bridge results in a significant reduction in a-adrenergic agonist activity

[72]. This is in sharp contrast to hydroxy substitution of the analogous /J-phenethylamines as mentioned above.

3) Substitution at the Imidazoline Ring

Enlargement of the five-membered imidazoline ring to a six, seven, or eight membered ring system results in a dramatic decrease in both 46

0Cj- and « 2 _agonist activities [75]. There are; however, several novel

five membered mono- and bicyclic heteratom ring systems which when

replaced for the imidazoline ring provide very potent a-agonists

(39-41) as shown in Figure 20 [75].

X> CH.

39 40 41

Figure 20: Imidazoline modified a-adrenergic agonists

Next, replacing one of the imidazoline nitrogens with a carbon atom to

form a pyrrolidine ring results in only a slight reduction in tt2 “affinity [77]. This suggests that only one of the imidazoline nitrogen atoms is important for binding to the receptor [77]. Addi­ tionally, substitution of the imidazoline ring with an imidazole ring results in potent and selective central a 2 _agonist activity as shown with (42) [80,81].

CH

CH

NH

42 47

Interestingly, medetomidine (42) shows weak antagonist activity on the

human platelet otj-receptor, evidence of a 2 _receptor heterogeneity

[82]. The pharmacological activities of the Imidazole analogs,

attached at either the 2- or 4- position of the imidazole ring, show a

complex and high degree of structure dependence of whether the com­

pound will exhibit a^- or o^-adrenergic agonist or antagonist activity

[82,83]. Substitution of the imidazoline moiety on the nitrogens or on

carbons 4 or 5, or opening of the imidazoline ring all significantly

reduce (X-adrenoceptor affinity [77]. Finally, connecting the phenyl

and imidazoline ring systems as in 43 abolishes affinity, presumably by

forcing a planar structure which does not allow the appropriate confor­

mation needed for binding as shown in Figure 18 on page 43 [75].

NH

43

1.4.2 Beta-2 Adrenergic Agonist Structure Activity Relationship

Isoproterenol (1) is the prototype /?-adrenergic agonist. Although used clinically for its bronchodilating effects, it is a nonselective beta agonist and its use is limited by its cardiac (/?j) and skeletal muscle C/^) stimulatory effects. The structure of isoproterenol (1 ) possesses three important features critical for 48

/?2 -agonist activity: 1 ) a catechol ring, 2 ) an ethanol functionality

and 3) a secondary amine [84]. Typically the R(~) isomer at the /[

carbon displays the most potent /[-activity [84].

Modifications of the catechol moiety were made to circumvent the rapid metabolism by COMT, which methylate the meta hydroxyl as shown in

Figure 21. The resorcinol derivative, (44), is not a substrate for COMT but maintains potent /[-agonist activity, although it is also nonselective. Modifications of the catechol which are not

subtrates for COMT yet greatly enhance the P 2 /P 1 selectivity ratio include substitution of the meta hydroxyl for a sulfonamidfe (soterenol,

45, /[2 //[j=6 ), a hydroxymethyl group (albuterol ,7, /^//[j^SO, a methylamino group (46, /[2/^j=192) or incorporation of a meta amino

group into a carbostyryl ring system (47, P 2 /P 1 -IOUO') [7]. Many other similar derivatives have been made [84]. Replacement of the para hydroxyl results in a drastic decrease in activity.

Tertiary amines are not active as /[-agonists [7]. The presence of bulky amino groups abolishes a-adrenergic agonist activity and increases /[-adrenergic agonist activity. The optimum alkyl substitution on the secondary amine for selective /^-agonist activity is found with the t-butyl group [85]. Similarly, looking at a series of cycloalkylamine derivatives, the cyclobutylamino compound provided potent and selective /^“agonist action [85]. Extending the amino substituent with branched arylalkyl groups; however, provided the selective /^-agonist fenoterol, 5, which is as equipotent on 49

HH. HO HH. CH.SO,

HO

OH 45

OH

HH. HH

HO HO 47 46

OH

HO HH.

HO 48

Figure 21: Selective /?2 -adrenergic agonists.

/?2 -receptors (trachea) as isoproterenol (1), but is 45 times less potent on ff receptors (atria) than isoproterenol [8 6 ]. Compounds containing a-substituents which are incorporated into a ring as in provide selective /^"agonist activity [84]. Two compounds do not fit the above structure activity relationships for ft-agonist activity. Both (49) [7] and trimetoquinol (50) [87], have potent /7-agonist activity; however, they are nonselective for

(i-receptors. 50

NHCH

1.5 PHYSIOLOGICAL FUNCTIONS OF ADRENERGIC AGONISTS

The physiological functions of a-adrenergic receptor stimulation

are summarized in Table 2 [75,88]. a-Adrenergic receptors, found both

peripherally and centrally, are distributed throughout a variety of

organs including both arterial and venous vasculature and in renal

tissue. Thus cardiovascular effects of a-adrenergic stimulation are

mediated at several different levels. The antihypertensive action of clinically used a 2 ~agonists, such as clonidine (lj>), and azepexole, is mediated via a central mechanism which results in a peripheral . This hypotensive action is thought to be a result of postsynaptic a 2 _receptor stimulation in the brainstem causing decreased sympathetic outflow [89]. Central a 2 _adrenergic receptor stimulation may also enhance parasympathetic outflow, although the precise site of action in the CNS for this mechanism is still unknown [89]. Centrally acting antihypertensive a 2 ~agonists commonly 51 cause sedation as a side effect [8 8 ]. It has been established that this sedation is induced by an (^-adrenergic receptor mediated mechanism [89]. Presynaptic 0(2 “adrenoceptors, which provide a negative feedback mechanism to modulate transmitter release, have been found in all sympathetically innervated tissues examined to date [2 1 ].

The presynaptic (^-adrenoceptors are constantly under active tone as a result of endogenously released norepinephrine and are believed to play a role in the control of sympathetic flow [2 1 ].

Table 2

Physiological Functions of Alpha-Adrenoceptor Stimulation

a.2

-Smooth muscle contrection -Hypotensive effect mediated -Increased Inotropy through central mechanism

•Increased chronotropy -Arterial vasoconstriction age dependent: decreasing in •Inhibition of postganglionic geriatric; increasing in young norepinephrine release -Renal nerve stimulation (via presynaptic a }-adrenoceptors) -renal artery vasoconstriction •Stimulation of platelet aggregation -tubular sodium reabsorption and secretion -Inhibition of renin secretion •Inhibition of water and sodium -Increased release of adrenocorticotropic reabsorption at renal distal hormone (ACTH) from pituitary •Inhibition of gastric and intestinal -Extraocular muscle contraction motility and secretion -Bladder contraction •Inhibition of insulin release -Prostate contraction form pancreatic ialet cells •Inhibition of lipolysis •Ocular hypotension -Stimulation of growth hormone release from adenohypophysis Both postsynaptic a^- and (^-adrenergic receptors are present in

vascular smooth muscle and stimulation of these receptors leads to

vasoconstriction [75]. With respect to the arterial circulation,

postsynaptic vascular receptors are thought to be located at the

neuroeffector junction and therefore interact with endogenous

norepinephrine liberated from sympathetic nerves [21,89]. Thus, the

postjunctional vascular a^-adrenoceptor appears to maintain resting

vascular tone [89]. Conversely, the postsynaptic vascular

« 2 ”adrenoceptor is thought to be located extrajunctionally and

proposed to interact with circulating epinephrine [21,89]. The

extrajunctional (^-adrenoceptor, therefore, may contribute to

peripheral vascular resistance in times of stress or in certain

hypertensive disease states where circulating levels of catecholamines

are. abnormally high [89]. Both postsynaptic vascular a^_ and

(^-adrenoceptors also coexist in the venous circulation. In the

venous circulation; however, the postjunctional vascular

O^-adrenoceptor is located at the neuroeffector junction, whereas the postsynaptic vascular a^-adrenoceptor is found extraneuronally [2 1 ].

Topical and systemic a^-agonists are employed as nasal decongestants as stimulation of oij-adrenoceptors on nasal and pharyngeal mucosal surfaces results in marked vasoconstriction.

In the myocardium, a j-adrenoceptor stimulation causes increased

inotropy [75,88]. The actions of Of ^-adrenoceptor stimulation with respect to chronotropic effects are poorly understood [8 8 ]. It has 53

been suggested that activation of a^-adrenoceptors in the geriatric

population causes a decrease in chronotropy whereas an increase in

chronotropy occurs in the young [8 8 ]. There is also evidence of a

significant involvement of a-adrenergic induced coronary

vasoconstriction in the initiation and aggravation of myocardial

ischemia [90]. Human platelets contain a population of

a 2 ~adrenoceptors which induce aggregation and secretion [89].

a^-Adrenergic receptors located in the kidney cause renal artery

vasoconstriction and tubular reabsorption [89]. a^-Receptors are also

thought to inhibit renin secretion [8 8 ]. Conversely, renal

a 2 _adrenoceptors inhibit water and sodium reabsortion at the distal

nephron, possibly by inhibiting the action of [21,74,89].

Further actions of a-adrenoceptors may be found in Table 2 on page 51.

The prominent functions of the /[-adrenergic system are summarized

in Table 3 [3]. /[-Adrenoceptors in the heart mediate an increase in

both inotropy and chronotropy. Although fi ^-receptors are the predominant subtype in the heart, a pool of cardiac /^"adrenergic

receptors does exist. In congestive heart failure (CHF), cardiac

/[^-receptors are down-regulated whereas cardiac /?2 _receptors are not

[84]. Therefore, /^"agonists are used in the treatment of CHF to

improve cardiac function (particularly to increase inotropy), as well as their vasodilatory effects which cause a reduction in cardiac preload [84]. 54

Table 3

Physiological Functions of Beta-Adrenoceptors

-Heart •Bronchodilation -Lipolysis •increase chronotropy -Inhibition of chemical mediator release from mast cells -increase inotropy -Skeletal muscle contraction

-Vasodilation

-Uterine relaxation

-Increase insulin secretion

-Gluconeogenesis

-Glycogenolysis

/^"Adrenergic agonists are used predominantly in the treatment of

asthma due to their bronchodilating effects. In addition to relaxing

airway smooth muscle, /?2 _agonists also reduce the release of chemical

spasmogens (such as ) from mast cells, improve mucociliary

transport, reduce bronchial mucosal edema and decrease cholinergic

impulses which cause bronchoconstriction [91]. These actions make

/?2 -agonists the most popular anti-asthmatic drugs. The main side effects associated with selective /?2 ~agonists are due to the

/^-mediated vasodilation, which may cause reflex tachycardia and

/?2 -agonist induced skeletal muscle [84]. Tachycardia associated with the use of /?2 -agonists may be due to a non-selective agonist activity on /^-adrenoceptors as well. /^"receptor agonists also relax the uterus and are used clinically to control premature labor [8 8 ]. And last, although lipolysis was formerly categorized as a 55

/^-adrenoceptor response, characterization of the novel /?g-receptor

subtype in brown adipocyte provides evidence that lipolysis is actually a //^-mediated response [14].

1.6 THROMBOXANE A o

Thromboxane A2 (TXA2 ) is a potent inducer of platelet aggregation.

It also has powerful vasoconstricting properties. TXA2 is biosynthesized from arachidonic acid via the cyclooxygenase pathway as shown in Figure 22 [92]. Arachidonic acid is liberated from membrane phospholipids by the action of phospholipase A2 • The activation of phospholipase A2 may occur either nonspecifically by cell damage or injury, or by receptor-coupled cell activation [93]. Arachidonic acid may also be released by the action of diglyceride lipase on diacylglycerol, a product of phospholipase C action on phosphatidylinositol [93]. The formation of phosphatidic acid from diacylglycerol may also activate phospholipase A2 by releasing membrane bound Ca^+ [94]. 56

,CO,H Phospholipase C Phospholipids

Arachidonic Acid I Cyclooxygenase

Laukotrianes PCC, 1Peroxidase 0 fNM] Prostacyclin PCI2 Synthetase 0

PGH,

Thromboxane Synthetase PCE2

PCD, CO.H

PCF 2a

TXA,

OH 1

CO.H

HO

HO TXB,

Figure 22: Biosynthesis of TXA2 via the arachidonic acid cascade

Cyclooxygenase converts arachidonic acid to the cyclic endoperoxide

PGG2 which is converted to PGH2 by a peroxidase. Alternately, arachidonic acid may be converted to leukotrienes by the action of 57

lipoxygenases. P G ^ , which is the common intermediate for all

prostaglandins, is converted to PGE2 , PGD2 or FGF2 ,. Alternately via

prostacyclin synthetase, PGH2 is converted prostacyclin (PGI2 ), which

is potent inhibitor of platelet aggregation and a vasodilator.

Prostacyclin inhibits platelet activation by inhibiting phospholipase C

by raising intraplatelet cAMP levels [95]. Via the enzyme thromboxane

synthetase, PGH2 is converted to TXA2 . Although a very potent platelet

aggregator and vasoconstrictor, TXA2 is an unstable compound with a half-life of only 30 seconds. It then rapidly breaksdown, nonenzymatically, to the stable but inactive TXB2 [92].

1.6.1 Hemostasis and Platelet Function

Hemostasis, literally the arrest of bleeding, is initiated when the vascular endothelial lining is disrupted and blood is exposed to the subendothelial matrix [95]. The primary phase of hemostasis involves the initiation of platelet aggregation and the immediate formation of a platelet plug at the site of injury [95]. Secondary hemostasis requires a longer period of time and involves fibrin formation resulting from reactions of the plasma coagulation system [95]. In this phase, ultimately thrombin acts on fibrinogen to form a fibrin meshwork which acts to strengthen the primary hemostatic plug.

In resting state, platelets circulate as individual anucleate cells which exhibit minimal interaction with the vessel wall of other blood cells [96]. Expression of platelet adhesive properties is stimulated primarily by exposed collagen in the subendothelial matrix [96]. The primary hemostatic adhesive event comprises four components: 1 ) 58

attachment of platelet to a substratum such as subendothelial matrix,

2) granule release from the platelet, 3) platelet shape change from

normal disc shape to a more spherical form with pseudopod extension,

and 4) recruitment of additional platelets to form aggregates [95,96].

The von Willebrand factor, an adhesive glycoprotein, accomplishes

platelet attachment by forming a link between platelet receptor sites

called glycoprotein lb and subendothelial collagen fibrils [95]. The

adherent platelets then generate and secrete mediators from their

storage granules. These granule contents include several proteins such

as von Willebrand factor, fibronectin, thrombospondin, a heparin neutralizing protein (platelet factor 4) and platelet derived growth

factor (PDGF) from a granules [95,96]. Dense granules release ADP,

Ca 2+ , and , while lysosomes release endoglycosidases and a heparin cleaving enzyme [95]. The a granules are the first to secrete their contents following a weak stimulus. [97]. Dense granules require a moderate stimulus to secrete while a strong stimulus is necessary for

lysosomes to secrete their mediators [97]. Released ADP modifies the platelet surface so that fibrinogen can attach to a calcium-dependent membrane complex, glycoproteins Ilb/IJIa, and link adjacent platelets into a hemostatic plug [95].

The platelet plug is important in stopping blood loss from capillaries, and small arterioles and venules [97]. In the case of large vessel injury, fibrin clot formation is necessary to strengthen the platelet plug. As the primary hemostatic plug is formed, plasma coagulation factors are activated which stimulate secondary hemostasis 59

[95]. There are two pathways of coagulation, termed intrinsic and extrinsic pathways. Platelets are involved in the intrinsic pathway only. The intrinsic pathway of coagulation is initiated by autoactivation of factor XII by combining with injured subendothelial tissue of by contact with a negative surface such as on platelet membranes [98]. Activated factor XII (factor Xlla) activates the conversion of prekallikrein to kallikrein. Kallikrein can cause reciprocal activation of factor XII which is much faster than autoactivation [98]. Factor Xlla then causes the conversion of factor

XI to its active form (XIa) which through a similar proteolytic reaction leads to activated factor IX (IXa) (Figure 23). A complex of factor IXa, factor Villa, Ca and phospholipids (PL) results in the conversion of factor X to its active form (Xa). Factor Xa then cleaves two peptide bonds in prothrombin to form thrombin [99]. This action of

Xa is accelerated by the active form of factor V (Va), phospholipids 94* and Ca [99]. Thrombin then cleaves fibrinopeptides A and B from the amino termini of the Aa and Bfl chains of fibrinogen, respectively, to form fibrin monomers [99]. Fibrin monomers automatically polymerize to form an insoluble gel via hydrogen bond linkages. The active form of factor XIII (Xllla, activated by thrombin) catalyzes interchain transglutamination reactions resulting in a covalently linked stable fibrin clot [99].

Platelets are also involved in clot retraction by activating thrombasthenin and platelet actin and myosin proteins. These contractile proteins help compress the fibrin meshwork to pull the 60

xn — — ► x n a

XI — - ► XU */ IX ------► IXa

X ------► Xa c sa p lm of IX*. V IIIo. phospbollpldi. Cil *

Coaplii of X a , Va » Prothrombin —...» Thrombin phospholipid*. Co1* xm

Fibrinogen Fibrin monomers "PlbrlnopoptldAo A Aad B

Insoluble Fibrin polymers (H-bondod)

| Co1*, m u

Stable Fibrin polymer (eovolonc cross-Unking)

Figure 23: Intrinsic coagulation pathway. edges of the broken vessel together and express serum from the clot

[2]. Furthermore, PDGF released form platelet a granules encourages cellular growth to help repair the injury site.

The plasma coagulation system is tightly regulated so that only a small amount of each coagulation enzyme is converted to its active form and the clot does not propagate beyond the site of injury [95]. Normal blood flow dilutes the concentration of reactants and coagulation factors. Also, the presence of anticoagulant proteins such as antithrombin III and protein C inactivate several factors as a control 61 mechanism [98]. Furthermore, synthesis of prostacyclin (PGI2 ) inhibits coagulation.

1.6.2 Biochemical Bole of TXA2 in Platelet Activation

Present on the membrane surface of platelets are receptors for most of the aggregating agents, including thombin, ADP, epinephrine, platelet activating factor (PAF), PGH2 and TXA2 [97]. Although collagen also activates platelet aggegation, it appears unlikely that there is a specific collagen receptor on the platelet surface [97].

Platelet activation and secretion are mediated by changes in intracellular cAMP levels, the influx of calcium, hydrolysis of membrane phospholipids, and phosphorylation of critical intracellular proteins [95]. The binding of agonists to platelet surface receptors activates two membrane enzymes, phospholipase C and phospholipase A2 which catalyzes the release of arachidonic acid from membrane phopholipids, phosphatidylinositol and phosphatidylcholine [95]. A small amount of this arachidonic acid is converted to TXA2 via the cyclooxygenase pathway described above (Figure 22 on page 56).

Interaction of TXA2 with its putative TXA2 /PGH2 platelet receptor [1 0 0 ] activates phospholipase C (PLC) as shown in Figure 24 [101]. PLC catalyzes the hydrolysis of phosphatidylinositol 4,5-biphosphate (PIP2) to diacylglycerol (DAG) and inositol triphosphate (IP3 ) (Figure 9 on page 26) [95]. In the platelet, DAG, utilizing calcium and phenylserine as cofactors, activates protein kinase C [102]. Protein kinase C then phosphorylates a 45 KDa protein which regulates platelet secretion. The actions of DAG are terminated by its metabolism to 62

phosphatidic acid (PA) which plays a role in activation of

prostaglandin synthesis [94,102]. IPg mediates the mobilization of

2+ Ca into the platelet cytosol from internal sites of platelet calcium

sequestration suggested to include the dense tubular system and the

open canalicular system [97]. The resultant calcium-calmodulin complex

stimulates myosin light chain kinase to phosphorylate myosin light

chains [95]. The interaction of the phosphorylated myosin chains with

actin facilitates granule movement and shape change [95].

TXA./PGH, MwAnw I

n p ip PLC

PA DG IP, IP*

Protein Myosn Light, Kinase C Chain Kinase (MLCK)

43K— J L— 20K Prert 1 f Piet ©-43K Piot 0-2OK Plot

0m m Qranuk SECRETION AGGREGATION < SHAPE CHANGE

ADP* 5 - H T * * • Foodback Agontt*

Figure 24: Role of TXA2 in platelet activation, (diagram used by per­ mission of f’eller [1 0 1 ])

Figure 25 shows the interrelationship of mechanisms affecting platelet aggregation. Aggregation induced by many agonists occurs in two waves [103]. The primary wave occurs via direct stimulation of the polyphosphoinositide pathway and involves the processes of pseudopod extension and formation of small aggregates [104]. The primary wave is characterized by aggregation alone and is reversible [103,104]. The secondary wave also involves the activation of prostaglandin synthesis.

The secondary wave involves further aggregation, forming larger more dense aggregates, and granule secretion [103,104]. Furthermore, the secondary wave aggregation is irreversible [104]. Thrombin Collagen 64

Membrane s. S’ O 'f % J A :;raaiB6v#' mu'm m a m K ^ 'A <'" Phospholipids ( - ) ( + ) V > ' > ’ \ f <•— p l a 2 PLC (Late) (Early)

CAMP

AA«---DG

♦ — CO

i > PGGj PGH,

— PKC # Dense ♦ — TS § Tubules

P-45K W Prof TXA2* Ssssrf^' * — Calmodulin

MLCK

P-MLC (20K)

Dense Granule SECRETION AGGREGATION * ------SHAPE CHANGE i ADP, 5-HT * * Feedback Agonists

Figure 25: Interrelationship of platelet receptor mechanism, (diagram used by permission of Feller [101]). 65

1.6.3 TXA2 Involvement in Cardiovascular Disease

Platelets play a major role in thrombus formation and embolization.

There is also evidence of an important role of platelet aggregation in

the formation and progression of atherosclerotic plaques [97,105,106].

Atherosclerosis, which represents the leading cause of death in the

Western World, can result in a wide range of cardiovascular and

cerebrovascular diseases linked to platelet involvement [105]. These

include ischemic heart disease, transient ischemic attacks, and stroke

[105]. Other patients prone to thrombosis include those with

aortocoronary bypasses, chronic dialysis patients and diabetics [105].

The formation of the hemostatic platelet plug and the formation of

an arterial thrombus are very similar processes. Thrombosis can be

thought of as an inappropriate or unregulated form of hemostasis [95].

In a situation of repeated injury to the vessel wall, such as in the

diseased arteries of atherosclerosis, the thrombus tends to persist as

a continuous dynamic aggregation and deaggregation of platelets which

continually release mediators such as TXA2 [97,106]. The ultimate fate

of these thrombi can include partial or total obstruction of the vessel

or embolization leading to clinical manifestations such as ischemic heart disease [106].

Platelets from patients with arterial thrombosis, deep vein

thrombosis, or recurrent venous thrombosis, as well as from patients who have survived myocardial infarction, produce more TXA2 than normal

[107,108]. There is evidence that TXA2 is released into the coronary 66

venous effluent in animals during coronary occlusion and reperfusion

causing vasoconstriction thus impairing coronary circulation [109].

TXA2 mimetics have been shown to cause potent coronary vasoconstriction

leading to myocardial ischemia and finally to sudden death [1 1 0 ].

Enhanced TXA2 release during myocardial infarction has been shown to

lead to cell damage and extension of the ischemic myocardial tissue

[111]. Furthermore, inhibition of the actions of TXA2 via synthetic

antagonists has been found to protect the myocardium during acute

ischemia [111,112]. Lastly, glomeruli from spontaneous hypertensive

produce more TXB2 , a stable metabolite and thus a biochemical

marker of TXA2 , than glomeruli from normotensive rats [114], It has

been suggested that increased renal TXA2 synthesis leading to increased

renal vascular resistance may play a causal role in the development of

[113,114]. The above discussion provides evidence that

TXA2 receptor antagonists may prove clinically useful as antithrombotic

agents and in the treatment of diseases involving ischemia and excess vasoconstriction.

1.6.4 TXA2 Antagonists

Inhibition of TXA2 at the receptor level is desirable for the

following reasons:

1 . PGH2 and TXA2 share the same receptor, although PGH2 is less

active. Thus, antagonists at this receptor results in inhibition

of both endoperoxide pro-aggregatory agents and mediators of vaso­

constriction . 67

2. Inhibition of TXA2 synthesis leads to accumulation of prostaglan­

din endoperoxides which can activate PGH2 /TXA2 receptor.

The PGH2 /TXA2 receptors have been subclassified into two types: the

platelet a receptor, responsible for aggregation, and the vascular t

receptor, responsible for contractile responses or tone of vascular

smooth muscle [115]. Two TXA2 mimetics, U44069 (51) and U46619 (52)

[116], are commonly used in the biological evaluation of TXA2 antago­

nists .

0 MM*mi

52 51

The initial design of TXA2 antagonists revolved around modifications of the basic prostaglandin backbone and included antagonists such as

13-azaprostanoic acid (53) [117] and pinane TXA2 (54) [118]. Recent advances in the design of prostanoid TXA2 antagonists has resulted in potent and long acting compounds including SQ33552 (55) [119], GR32191

(56) [120] and ICI 192605 (57) [121]. Relatively few nonprostanoid

TXA2 antagonists exist. These include trimetoquinol (50) [122] and sulotroban (BM13.177, 58) [123], which is currently being tested in clinical trials. The tetrahydrocarbazolyl acetic acid derivative,

L-670,596 (59) [124] was also found to have potent and selective TXA2 properties, representing another type of nonprosta­

noid antagonists. Recently an azulene derivative of sulotroban,

KT2-962 (60), has been reported to be 30 times more potent than sulo­

troban (58) [125]. Research in this area is actively being pursued in

search of potent, long acting, orally effective TXA2 antagonists for

the treatment of thrombotic and ischemic disorders.

HO 53 54

58 57

c h ,s o 2 59

Figure 26: TXA2 Receptor Antagonists CHAPTER II STATEMENT OF PROBLEMS AND OBJECTIVES

2.1 TR1MET0QU1N0L ANALOGS

Trimetoquinol (THQ, 50) is a nonspecific //-adrenergic agonist

[87,126], as well as a competitive inhibitor of human platelet

aggregation and rat aortic contraction induced by the TXA2 mimetic,

U46619 (52) [127], In //-adrenergic systems, S(-)-TMQ is more potent

than R(+)-TMQ [128]. Currently, the S(-) isomer of TMQ is marketed in

Japan as Inolin for its bronchodilating effects. The anti-aggregatory activity of TMQ was first reported by Shtacher and co-workers [122] when TMQ was found to inhibit platelet aggregation induced by collagen,

ADP, and epinephrine. This activity is unrelated to a- or

//-adrenergic mechanisms and is independent of prostaglandin biosynthesis or cAMP [122]. Subsequently, Mukhopadhyay et al. [127] showed that TMQ is an antagonist of PGH2 /TXA2 “mediated responses in rat aorta and human platelets. The stereoselectivity for TMQ antagonism of putative PGH2 /TXA2 receptors is opposite to that required for

//-adrenergic activity [129,130]. The R(+)-isomer of TMQ was found to be 40- and 22-fold more potent than S(-)-TMQ as an inhibitor of

U46619-Induced platelet aggregation and serotonin secretion, respectively [131]. Although there was early controversy as to whether

- 69 - 70

TMQ was a PGH2 /TXA2 antagonist [132,133], Ahn et al. [131,134] have

proved that R(+)-TMQ acts as a selective and highly stereospecific

PGH2 /TXA2 receptor antagonist in platelets.

Recently, Clark et al. [135] reported the synthesis of the 5-fluoro

(61) and the 8 -fluoro (62) derivatives of TMQ.

H HO

NH HO

OCH, OCH

OCH OCH, OCH, OCH, 61 62

These analogs have been evaluated for f t and /J2-adrenergic activity

(Table 4) [135] and TXA2 antagonist activity (Table 5) [136]. In \ producing tracheal relaxation (/?2 )» TMQ (1 2 ) an<^ the fluoro analogs

are essentially equal in activity [135]. However, on guinea pig

trachea (ftj) the order of chronotropic activity is TMQ (50) > 5F-TMQ

(61) > 8 F-TMQ (62) [135]. The changes in /?j-activity were correlated

to differences in phenolic pKfl attributed to the electronic influence of fluorine [135]. As the pKfl decreases from 8.77 for TMQ (50) to 7.86

for 8 F-TMQ (62), the /^-activity decreases [135]. This converts to an enhanced /?2//Jj selectivity as follows: 8 F-TMQ > 5F-TMQ > TMQ [135].

The fluoro analogs are also selective relative to TMQ for inhibition of

U46619 (52) actions on rat thoracic aorta (t receptors) versus human 71

platelets (a receptors) [136]. Although 8 F~TMQ (62) is less active

than TMQ (52) on human platelets, it is of interest to note that 8 F-TMQ

is more active than TMQ on aorta [136]. This corresponds to a f/a

selectivity ratio of 35 for 8F-TMQ (62) when compared to TMQ (50)

[136].

Table 4

Comparative Activities of TMQ, 5F-TMQ, and 8F-TMQ on Beta Systems (Guinea Pig Trachea and Atria)

Trachea * Atria (px)- Selectivity Potency Potency Ratio Drug P*CSok Ratio P*c50k Ratio P2/0i

TMQ (50) 7.24 * 0.02 1.00 7.52 ± 0.21 1.00 1.00 5F-TMQ (61) 7.26 t 0.11 1.05 6.95 ± 0.13 0.27 3.88 8F-THQ (62) 7.15 s 0.14 0.81 6.53 a 0.11 0.10 7.94

'Values are Che naan £ SEN of M - 4-6 kpEC50 - -log EC,0 (M)

The first objective of this research was to design and synthesize a selective /^"agonist, which retains the bronchodilating actions but lacks the /^-attributed side effects. The second goal was to gain a better understanding of the structural requirements for optimum activity at a and t PGH2 /TXA2 receptors with TMQ analogs. Because introduction of an electron withdrawing fluorine atom at the 5- and 8 - position of TMQ provided very desirable biological profiles on both

/[-systems and TXA2 systems, the target compounds for this study 72

Table 5

Comparative Inhibitory Activities of TMQ, 5F-TMQ. and 8 F-TMQ in TXAq Systems -

Huaan Platelets' Bat Thoracic Aorta' (Aggregation) (Contraction)

Potency Potency Selectivity Ratio Drug Eatlo p V Ratio Aorta/Platelets

TMQ (50) 6.36 x 0.13 1.00 5.84 * 0.12 1.00 1.00 5F-TMQ (61) 5.13 ± 0.11 0.06 5.50 ± 0.29 0.46 7.67 8F-TMQ (62) 5.41 * 0.18 0.11 6.43 s 0.14 3.89 35.36

'Values are the naan * SEM of N - 4-12 6picso - -log ic50 (H) cpKj - -log [A]/(CK • 1) vhere [A] - solar concentration of antagonist and CR - concentration ratio - ECS0(plus antagonist)/ECJ0(control)

involved the introduction of the highly electron withdrawing trifluoromethyl group. The

0.06, respectively [137]. For the trifluoromethyl group, (Tffl equals

0.43 and a equals 0.54 [137] which corresponds to significantly greater electron withdrawing ability of the trifluoromethyl group compared to fluorine. The rationale for the target compounds was that a further increase in the electron withdrawing properties at the 5 - and

8 - positions of TMQ (50) may further separate /?£" and /7^-adrenergic activities as well as provide a greater separation of PGH2 /TXA2 t and a receptor activities. Thus, the target compounds were

5-trifluoromethylTMQ (63) and 8 -trifluoromethylTMQ (64). 73

CF HO HO

NH NH HO HO OCH CF OCH

OCH OCH OCH3 OCHg 63 64

The complementary relationship between a receptor and a drug molecule may result in a distinction between enantiomers of a chiral drug molecule [138]. Often only one enantiomer possesses full activity. As discussed above, on /^-systems, S(-)-TMQ was more potent than R(+)-TMQ [128]. Conversely, on TXA2 systems, R(+)-TMQ was more potent than S(-)-TMQ [131]. Since 8 F-TMQ (62) retained potent activity as well as showed a marked separation of activities on /J-systems

(Table 4 on page 71) and TXA2 systems (Table 5 on page 72), the third objective of this study was to separate and evaluate the biological activities of R-8 F-TMQ (R-62) and S-8 F-TMQ (S-62).

HO HO

NH NH HO HO OCH OCH

OCH OCH OCH, OCH, R-62 S-62 2.2 IMIDAZOLINE DERIVATIVES

As mentioned above, imidazolines represent a major class of

compounds which possess a-adrenergic activity. Ruffolo et al.

[76,139] found 3 ,4-dihydroxytolazoline (65) to have potent full agonist

activity at a^-adrenoceptors and partial agonist activity at

0t2 -adrenoceptors. Furthermore, 65 had weak (i ^-adrenergic agonist

activity and even weaker activity at /^"adrenoceptors [139].

Introduction of a benzyl hydroxyl group on 65, to give

3,4,a-trihydroxytolazoline (6 6 ), results in a decrease in activity on

Of2 — and (^-adrenoceptors [76].

OH

HO HO NH NH

HO HO 65 66

This is interesting because benzyl hydroxylation of the corresponding

/?- (to give norepinephrine) results in increased

a-adrenergic activity.

Ring fluorination of norepinephrine resulted in a series of

compounds with drastically different activities [140].

2-Fluoronorepinephrine was a selective /?-adrenergic agonist,

5-fluoronorepinephrine possessed both a- and /Nadrenergic agonist

activity and the 6 -fluoro analog (1 2 ) was a potent selective a^-adrenergic agonist [140] (Note: the IUPAC nomenclature for the 75

6 -fluoro analog (1 2 ) is numbered as "2 -fluoro" but, for clarity in the

following discussions, all substitution at this position in 12 and

related compounds will be commonly referred to as the 6 -fluoro or

6 -trifluoromethyl analogs.) This separation of activity induced by

fluorine substitution prompted the synthesis of benzyl ring fluorinated

3 ,4-dihydroxytolazoline derivatives [83].

F

F 67 68 69

Fluorinated catocholimidazolines 67, 6 8 , and 69 were found to be

30-fold more selective as agonists on a^-adrenoceptors versus

« 2 “adrenoceptors [83,141]. The rank order agonist potency on or. j-adrenoceptors was as follows: 2-fluoro analog (67) > 5-fluoro analog (6 8 ) > 3,4-dihydroxytolazoline (65) [141]. The 6 -fluoro analog

(69) was a partial agonist on a^-adrenoceptors [141]. With this in mind, the objective in this study was to synthesize a series of benzyl ring trifluoromethylated analogs of 3,4-dihydroxytolazoline (shown in

Figure 27) to examine the effects of going from fluorine substitution to the highly electronegative trifluoromethyl substitution.

This idea is further substantiated by the report of the trifluoromethylated phenylaminoimidazoline St587 (73) which has potent and highly selective a^-adrenergic agonist activity [142]. DeJonge et 76

HO NH

HO

R2

70 - 2-Trifluoromethyl analog: CF3 H H

71 - 5-Trifluoromethyl analog: H CF3 H

72 - 6-Trifluoromethyl analog: H H CF3

Figure 27: Trifluoromethylated benzylimidazoline target compounds al. [142] have hypothesized that for phenylaminoimidazolines, increasing the size of the substituent at the 5-position on the phenyl ring determines selectivity towards aj-adrenoceptors. Thus, the target compounds 70, 71, and 72 may impart a highly selective aj-adrenergic activity by virtue of electronic and/or steric bulk contributions of the trifluoromethyl group.

NH

CF HN

73 2.3 NOREPINEPHRINE DERIVATIVES

As discussed above, Kirk et al. [140] have synthesized the ring

fluorinated analogs of norepinephrine. Although 5-fluoronorepinephrine

possessed both a- and /J-adrenergic agonist activity,

2 -fluoronorepinephrine was a selective /^-adrenergic agonist and

6 -fluoronorepinephrine (1 2 ) was a selective a-adrenergic agonist

[140]. The separation of activity with these compounds was

rationalized by different rotameric conformations formed via hydrogen

bonding with the fluorine substitution at the 2 - or 6 -position and the

/?-hydroxyl group (Figure 28) [140,143].

NH

HO NH HO

HO HO

"Beta" Conformation "Alpha" Conformation

Figure 28: Rotameric conformations of 2- and 6 -fluoronorepinephrine. (Modified from [143])

Alternately, it was suggested that the differences in activity may be due to perturbations in phenolic ionization induced by fluorine substi­ tution [140]. The differences may also be due to local lipophilic per­ turbations at the site of fluorine substitution [140]. To further examine this issue, the objective was to synthesize and evaluate the 78

activity of the 2- (74), 5- (75), and 6 - (76) trif luoromethylated

derivatives of norepinephrine.

OH OH OH HONH HONH HO. NH

HOHOHO CF CF,

74 75 76

The trifluoromethyl group is more electron withdrawing than fluorine, as described above. Furthermore, the trifluoromethyl group is more lipophilic as the = 0.88 in comparison to fluorine in which 7Tp =

0.14 [144]. This substitution should provide more information concern­ ing the interaction of norepinephrine with a- and /^-adrenoceptors. CHAPTER III RESULTS AND DISCUSSION

3.1 CHEMISTRY

3.1.1 Synthesis of 5- and 8-Trifluoromethylated Trimetoquinol Analogs

The key intermediate in the synthesis of the 5-trifluoromethyl derivative of trimetoquinol (63) was 3,4-dimethoxy-

2-(trifluoromethyl)benzaldehyde (78). A part of the challenge in the synthesis of this compound was to obtain an aldehyde ortho to the trifluoromethyl group which is a potent meta director. The synthesis of 78 was achieved using 2-iodo-3,4-dimethoxybenzaldehyde (77) > previously synthesized in our laboratory [145], and applying the method of Matsui and co-workers [146] in which aromatic halides are trifluoromethylated. Thus heating a mixture of 77, 2 eq. of cuprous iodide and 4 eq. of sodium trifluoroacetate in N-methy1-2-pyrrolidinone

(5% w/v) to 175°C for 4 hours gave 78 in 40% yield (Scheme I). It is essential that this reaction be carried out under strict anhydrous conditions as the inclusion of moisture favors the reduction of the aromatic halide [146]. The trifluoromethylated product was confirmed first by the NMR spectrum of 78 in which the aldehyde and aromatic peaks were shifted downfield compared to the starting material 77.

- 79 - Furthermore, the aldehydic proton of the starting material 7_7 appeared as a sharp singlet at 10.0 ppm, whereas in the *1 NMR spectrum of 78, the aldehydic peak appeared as a broad quartet at 10.2 ppm as a result of splitting by the 3 fluorine atoms (Scheme I). Also, a single peak 1 9 at -52.5 ppm appeared in the F NMR spectrum of 7 I which is indicative of an aromatic trifluoromethyl group [147]. IHieCUL Synthesis of 78 and Related -H NMR Spectra in CDCl^/TMS in Spectra NMR -H Related and 78 of Synthesis CH-0 CH,0 77 cee I Scheme Cul m CH-0 CH-0 78 F 0 CF- Beginning with benzaldehyde 28, the remaining synthesis of

5-trifluoromethyltrimetoquinol (63) is shown in Scheme II.

Benzaldehyde 78 was reduced with sodium borohydride in THFrf^O (9:1

v/v) to afford benzyl alcohol 79. The alcohol was converted to benzyl

bromide 80 via phosphorus tribromide in dichloromethane and subsequent

displacement with sodium cyanide in dimethylformamide gave

benzy lnitrile 8_1. Demethy lation using boron tribromide liberated

catechol 82 which was reprotected by reaction with benzylchloride to

produce the dibenzyloxy protected benzylnitrile 83. The cyano moiety

of 83 was reduced to the amine using diborane in tetrahydrofuran (THF).

Without purification, the resultant phenethylamine was condensed with

trimethoxyphenylacetic acid 84 by heating to reflux in toluene with

azeotropic removal of water via a Dean-Stark trap to afford the phenylacetamide 85. Applying Bischler-Napieralski conditions of phosphorous oxychloride in acetonitrile to phenylacetamide 85 resulted

in the dihydroisoquinoline which was immediately reduced with sodium borohydride to afford the dibenzyloxy protected tetrahydroisoquinoline

86 isolated as the hydrochloride salt in 40% yield. Deprotection of the catechol moiety with a refluxing equivolume mixture of HC1 and methanol gave the desired 5-trifluoromethyltrimetoquinol (63).

Looking at the initial retrosynthetic analysis of the

8 -trifluoromethylated derivative, it was envisioned that the tetrahydroisoquinoline nucleus could be formed using the appropriately substituted amide (Figure 29). Although the trifluoromethyl group is a meta director, it was envisioned that the amide precursor would be able 83

Scheme II

Synthesis of 5-trifluoromethyltrimetoquinol (63)

CF, o CH ?' CH NaBH. PBr, n X T m CH,0 THF/HjO CH-0 CHjClj CH-0

78 9:1 v/v 79 80

CF BBr, NaCN HO 01130 CHjCl, DMF CH3O HO 81 82

CF PhCH-Cl PhCH-0 1) BH, • THF PhCH-0 CN

K jCOj .KI PhCH-0 2) PhMe PhCH-0 OCH Acetone CH-0 CO-H OCH. CH-0 OCH CH-0

CF 1) POCI3 PhCH-0 HO HCl/MeOH HC1 •HC1 CH3CN NH ■> NH PhCH-0 HO OCH. OCH. 2) HaBH4 reflux

3) 3N HCl/MeOH OCH. OCH. OCH, OCH

to cyclize to the ortho position because in the synthesis of

5-fluorotrimetoquinol (61) [135] the amide readily cyclized to the meta

position although fluorine is an ortho-para director. This

retrosynthetic analysis led to 2-hydroxy-3-(trifluoromethyl)anisole as

the starting material.

The initial approach to phenol 88 involved the use of a typical aromatic hydroxylation procedure. Following the method of Ladd and 84

HO PhCHjO

NH 4 > HO PhCH-0 CF OCH CF3 OCH

OCH OCH \ OCH OCH

Figure 29: Initial retrosynthetic analysis of 8 -trifluoromethy1 trime- toquinol (64)

Weinstock [148], 3-(trifluoromethyl)anisole 87 was deprotonated with n-butyllithium in THF, followed by trimethoxyborate and then hydrogen peroxide with a basic then acidic workup. With this procedure, the desired trifluoromethylated phenol 88 was not formed, but rather the hydroxybenzoic acid derivative 89 was isolated in 13% yield (Scheme

III). The proposed mechanism of formation of the hydroxybenzoic acid derivative 89 involves the phenolic anion of the desired product as shown in Scheme IV.

Subsequent to this work, Backstrom and co-workers [149] reported the synthesis of phenol 88 in 18% yield by treatment of 85

Scheme III

Synthesis of hydroxybenzoic acid 89 sideproduct

CH 1) nBuLi thf ch30 V ^ “ HO^P cf3 2) B(OCH3 )3 T 3) H 20 2 3 «2 88

Scheme IV

Mechanism of formation of 89

F

HO OH NaOH

CH.O CH.O CH.O 88

F .OH OH HO

Repeat x 2 HC1

CH.O CH.O 89 86

3-(trifluoromethyl)anisole (8 8 ) with n-butyllithium and

N,N,N'jN'-tetramethyleneethylenediamine (TMEDA), followed by

trimethylborate, then hydrogen peroxide and again basic then acidic

workup. Previous to this publication, an efficient synthesis of

trifluoromethylated phenol (8 8 ) was achieved applying an alternative

aromatic hydroxylation method described by Lambert and co-workers

(150]. 3-(Trifluoromethyl)anisole (87) was lithiated using

n-butyllithium in THF, followed by transmetalation by addition of 1 eq.

cuprous bromide (Scheme V). Dry air was then bubbled though the

reaction mixture and the reaction was quenched with HC1. Purification

via column chromatography afforded the desired phenol (8 8 ) in 5 9 %

yield.

Scheme V

Synthesis of 88

1) nBuLi THF 2) Cu(I)Br

3) dry air CF CF 3 3 4) HC1 87 88 (Y: 59%)

The initial attempt to synthesize 8 -trifluoromethyltrimetoquinol

(64) is shown in Scheme VI. Aminomethylation of phenol 88 was carried

out by Mannich conditions using formaldehyde and N,N-dimethylamine to yield benzylamine 90, isolated as the oxalate salt. Benzylnitrile 91 was easily formed by preparing the quaternary ammonium salt of the free 87

base of benzylamine 90 with and, without purification,

displacement with sodium cyanide. Deprotection of 91 with boron

tribromide in dichloromethane gave the catechol 92 which was protected

with benzyl chloride to afford the dibenzyloxynitrile 93. Diborane

reduction of benzyl 93 gave the phenethylamine 94, which was

isolated as the hydrochloride salt. The free base of phenethylamine 94

was then condensed with trimethoxyphenylacetic acid 84 to give

phenethylacetamide 95. By applying Bischler-Napieralski conditions to

phenylacetamide 95, using phosphorus oxychloride in acetonitrile

followed by reduction, the desired tetrahydroisoquinoline 97 could not

be isolated. However, an unusual 3-aminoisoquinoline derivative 96 was

isolated in 11% yield (Scheme VI). The proposed mechanism of the

3-aminoisoquinoline side product involves the reaction of acetonitrile with the intermediate nitrilium salt as shown in Scheme VII [151].

Performing the Bischler-Napieralski reaction on 95 with phosphorus oxychloride in toluene or benzene completely eliminated the presence of the aminoisoquinoline side product; however, the desired compound 97 could not be isolated from the complex mixture.

The alternate approach to 64 involved the trifluoromethylation of the 8 -iodinated protected tetrahydroisoquinoline precursor (Figure 30).

The trifluoromethylation method of Matsui et al. [146] did not mention whether the presence of a basic amine moiety would affect the reaction.

Thus the amine of dibenzyloxy protected 8 I-TMQ 98, previously synthes­ ized in our laboratory [145], was protected as the trifluoroacetamide using trifluoroacetic anhydride, dimethylaminopyridine (DMAP) and pyri- 88

Scheme VI

Initial Synthetic Approach to 8 -Trifluoromethyltrlmetoqulnol (63)

1) CHjI CHjO CH-0 CH20 Y rV ^ N ( C H J)2 oitci2 CN

HO HNHe, 2) N.CN ^ CF, CF, DMSO CF, 88 90 91

BBr, HO, PhCH^Cl ^ ™UHPhCH.Oj0 > v ^ ^ Y ^ C N

ch,ci2 HO' KjCO,K2C0,.KI phCH20 'VJ CF, Acetone CF,

92 93

PhCH,0t-ncMjO BH3 • THF PhCHjO ^NH HC1 2) PhMe PhCHjO PhCH.O S^S ' O C H , CF CH.O CF, CO,H

CH.O 95 s X » , CH.O OCH,

1) POClj PhCHjO • HC1 CHjCN PhCH.O CH. HC1 ,NH 2) NeBH4 PhCH.O CF .OCH. 3) 3N HCl/MeOH CF OCH

OCH. OCH. OCH. OCH.

Desired product not obtained dine in dichloromethane to afford 99 (Scheme VIII). Applying the meth­ od of Matsui et al. [146] to 99 did not result in the desired trifluo- romethylated product 100, but rather the reduced product 101. This may have been a result of the trifluoroacetyl restricting free rotation of the trimethoxyphenyl ring forcing this moiety to cause a steric impe- dence to the approach of the trifluoromethy1 copper complex. Evidence for this hypothesis is in the NMR spectrum of 99. The aromatic 89

Scheme VII

Proposed mechanism of 3-amlnolsoqulnoline formation

95 0CH3

poclg CH3CN 1 r

BnO

BnO CF OCH

OCH OCH.

peaks (excluding the dibenzyloxy aromatic peaks) appeared as doublets in a ratio of 1:3 which close to singlets upon heating. Alternately, the reduced product may be a result of a presence of excess copper in the reaction mixture. 90

PhCHjO PhCHjO

NH NH PhCHjO CF OCH OCH

OCH OCH

Figure 30: Alternate Retrosynthesis of 64

Scheme VIII

Attempted synthesis of 100

o o PhCH.O PhCH.O CF NH PhCH.O PhCH.O DMAP OCH. •OCH. CH2C12

OCH. pyridine OCH. OCH. OCH.

CF.CO.N* PhCH.O Cul

PhCH.O PhCH.O OCH. CF. OCH.

OCH OCH. 100

PhCHjO

PhCHjO OCH.

101 OCH. OCH. 91

The method of Kobayashi and co-workers [152] was attempted next

(Scheme IX). In this case, the presence of basic amines did not affect

the reaction as reported by Kobayashi et al. [152] and much earlier by

McLoughlin and Thrower [153], thus eliminating one potential steric

problem. Furthermore, in this method the CF^Cu complex is isolated

before using, eliminating any excess copper. In this reaction, the

CF3Cu complex was prepared heating a mixture of trifluoromethyliodide

and copper powder in hexamethylphosphoramide (HMPA) in a steel tube to

o p 120 C for 2.5 hours and cooled. In an AtmosBag (Aldrich) purged with

argon, the solution containing the CFgCu complex was filtered through

celite to remove the excess copper. The filtrate and 98 [145] were

heated to 75°C under argon (Scheme IX). After 1 week, the absence of

starting material was detected by TLC. Upon workup, the isolated

compound was not the desired trifluoromethylated compound 9 7 .

Elemental analysis and mass spectral analysis suggested the product was

the iodinated protoberberine hydrochloride 102 isolated in 14% yield.

The presence of the berberlne system was also indicated by the NMR

spectrum of 102 which showed the typical AB doublets at 4.64 and 4.45 ppm indicative of the C -8 protons and geminal coupling (Figure 31)

[154]. 92

Scheme IX

Attempted synthesis of 97 using Isolated CF^Cu

PhCH.O PhCH.O 1) CFjCu • HC1 NH NH PhCH.O PhCH.O OCH. HMPA CF OCH. 75*C OCH3 2) Chromatography OCH. OCH. OCH. 3) 3N HCl/MaOH

PhCH.O ■ HC1 PhCH.O OCH.

OCH OCH,

103 • HC1

To confirm the structure of protoberberine 102, the hydrochloride salt of tetrahydroisoquinoline 98 [145] was treated with Mannich conditions of formaldehyde, ethanol and water, a typical method for protoberberine formation (Scheme X) [155]. The NMR and mass spectra of the protoberberine free base (1 0 2 ) isolated from this method were identical to the free base of the protoberberine structure obtained using CFgCu. When the free base from the Mannich method was treated with methanolic HC1, the expected hydrochloride salt of 102 was not obtained, but rather protoberberine 103 was obtained. The structure of compound 103 was confirmed by mass spectral analysis where the molecular ion appeared at 699 (37C1) and 697 (33 C1) (Figure 32).

Fragmentation peaks 469, 230 (37C1) and 228 (3 5 C1) indicate that chlorination has occurred on the trimethoxyphenyl ring as shown in 93

/M JW . J ^ U l J V m a m

7.5 7.1 6.1 5.5 4.5 1.5 2.5

Figure 31: NMR (250 MHz) spectrum of protoberberine 102

Figure 32. This was further proven by the NMR spectrum of 103 where the aromatic peak at 6.45 ppm is absent (Figure 33) in comparison to the 1H NMR spectrum of 102'HCl (Figure 31). Elemental analysis was 94

consistent with tha structure of compound 103. The mechanism of

formation of 103 is unclear. Interestingly, when the hydrobromide salt

of 102 was formed, no aromatic addition of bromine occurred and the

NMR and mass spectra were identical to 102HC1. To date, the synthesis of 64 has not been achieved.

Scheme X

Protoberberine formation using Mannich conditions

1 ) c h 2o PhCH.O EtOH PhCH.O H,0

Hethanolic HC1 ^ Hethanolic HBr 95

469

PhCHjO

PhCHjO OCH,

OCH,

230 (*7Cl) 228 (* * 0 1 )

; i t H- i»| » -1 "■* I } i i » « t— -r 100 000 300 700

Figure 32: Mass spectra of 103 96

1 /»\AAa V ^ Aft AMA

5.5 5.« 4.5 2.5

Figure 33: XH NMR (250 MHz) spectra of 103 in CDCI3 /TMS 97 3.1.2 Separation of 8-FluorotrimetoquinoI Enantiomers

The approach to the separation of the enantiomers of

8 -fluorotrimetoquinol (62) began with the free base of the dibenzyloxy

protected 8 -fluoro derivative 104 which was previously synthesized in our laboratory [135]. The free base of 104 was resolved using a D preparative Chiralcel OD (Diacel) column as shown in Scheme XI. The

chiral stationary phase (CSP) of the Chiralcel OD column is a cellulose tris(3,5-dimethylphenylcarbamate) polymer. The chromatographic separation of enantiomers on chiral HPLC columns involves the formation of diastereomeric complexes formed between the enantiomers and CSP bound to an inert support [156,157]. For the to be resolved, there must be a difference in stability of the resultant diastereomeric complexes [156,157]. The Chiralcel OD column was chosen because similar tetrahydroisoquinolines, such as laudanosine, have been resolved on this column [158]. Using a mobile phase of 70/30 hexane/isopropanol, the first enantiomer of 104 eluted at 88.79 min. and the second enantiomer eluted at 114.71 min.. The optical purity of each enantiomer was >99% as determined on an analytical Chiralcel OD

HPLC column. The dibenzyloxy enantiomers were crystallized as their hydrochloride salts. The dibenzyloxy protected R and S enantiomers of

104 were deprotected via catalytic hydrogenation with palladium on carbon to afford S(+)-63 and R(-)-63, respectively (Scheme XII). The configuration was determined from the CD spectra of S(+)-63 and R(-)-63 by analogy to the CD spectrum of authentic R(+)-TMQ (R(+)-50). 98

Scheme XI

Separation of 104 via Preparative Chiralcel OD HPLC

1) Preparative Chiralcel OD PhCH.O (2 ca x 50 ca) PhCH.O NH HPLC NH • HC1 PhCH.O - ► PhCHjO OCH OCH, 6 al/aln 104 OCH.OCH, 2) 3 N HCl/HeOH OCH, OCH.'3 OCH,

Enantloner 1 of 104 Ketention tiae - 88.79 aln [a]” - +15.6*

PhCH.O • HC1 NH PhCH.O .OCH

OCH. OCH,

Enantloaer 2 of 104 Retention tlae > 114.71 aln [a]25 - -14.7* D 99

Scheme XII

Deprotection to afford S(+)-62 and R(-)-62

PhCH.O HO

HH HC1NH • HC1 HH HC1NH PhCHjO Pd/C HO OCH. OCH.

OCH. OCH. OCH, OCH.J S(+)-62 Enantloner 1 of 104 [«] 23 +4.4* S configuration (CD (pectrum opposite to CD spectrum of authentic R(+)THQ)

PhCHjO HO

NH • HCl NH HCl PhCHjO Pd/C HO OCH. OCH.

OCH. OCH, OCH, R(->-62

Enantiomer 2 of 104 («]” - -4.2*

R configuration as determined by analogy

to CD spectrua of authentic R(+)TMQ

3.1.3 Trifluoromethylated Catecholimidazoline Analogs

The 5-trifluoromethylimidazoline 7J. was the first synthesized in the series. The initial synthesis (Scheme XIII) began with the formation of imidate 105 from the dibenzyloxy benzylnitrile 93, described above, via a pinner reaction using excess HCl gas and 1.1 eq. absolute ethanol in benzene [159]. Reaction of imidate 105 with ethylene diamine in chloroform [159] resulted in the dibenzyloxy protected imidazoline 106, isolated as the hydrochloride salt. Imidazoline 106HC1 was hygroscopic and difficult to crystallize, resulting in a poor isolation yield of 33%. Deprotection of dibenzyloxy imidazoline 106HC1 by 100

refluxing in an equivolume mixture of HCl and methanol gave

catecholimidazoline 2 1 ’HC 1 as a hygroscopic viscous oil which was

resistant to crystallization.

Scheme XIII

Unsuccessful Synthesis of 5-Trifluoromethylimidazoline 71

CF, CF,

“ '■xL.-=r- ■ ' m - PhCHjO benzene PhCH,0

93 105

CF / \ HCl/HeOH 1) H,M HH, PhCH,0 HO.

2) NeHCO, PhCHjO HO HH

3) 3 N HCl/HeOH *HC1 • HCl 106 • HCl 71 . HCl

The alternate route to 5-trifluoromethylimidazoline 71 involved isola­ tion of 106, obtained by treating imidate 105 with ethylene diamine, as the oxalate salt in 80% yield (Scheme XIV). Deprotection of dibenzy­ loxy imidazoline 1 0 6 (C0 2 H )2 via catalytic hydrogenation with palladium on carbon gave the target compound 2 1 which crystallized as the half oxalate salt in 50% yield from methanol/diethyl ether.

The starting material for the synthesis of the

6 -trifluoromethylimidazoline 72 was prepared as shown in Scheme XV.

Commercially available iodobenzaldehyde 107 was trifluoromethylated using the procedure of Kobayashi et al. [152] by heating a mixture of 101

Scheme XIV

Synthesis of 5-Trlfluoromethy1 Catecholimidazoline 71

P,C„,0 m |ici | ^,0

PhCHjO NailC0S PhCHjO 3) (COjH)j • 2HjO (COjH),

105 106 •(C0jH)2

CF. HO >- Pd/C HO NH • H(C02H)

71 • H (C02H)2

107, CFgl, and copper powder in HMPA in a steel tube at 110°C for 48 hours. Purification via column chromatography gave trifluoromethylbenzaldehyde 108 as white needles in 56% yield.

Benzaldehyde 108 was reduced with sodium borohydride to give benzyl alcohol 109. which was converted to benzyl bromide 110 via phosphorus tribromide. Subsequent displacement with sodium cyanide in dimethylformamide gave benzylnitrile 1 1 1 .

The synthesis of trifluoromethylated imidazolines 70 and 72 was the same as shown in Scheme XVI. A pinner reaction was performed on benzylnitriles 83 and 111 to yield imidates 112 and 114, respectively.

Treatment of the imidates with ethylene diamine in chloroform and isolation as the oxalate salts gave imidazolines 113 and 115.

Deprotection of imidazolines 113 and 115 via catalytic hydrogenation 102

Scheme XV

Synthesis of Benzylnltrlle ill

PhCH.O P h C H .O CF HMPA

steel tube PhCH.O 110*C, 48 hours PhCHjO 0 0 107 108

HsbH ^ P h C H ^ v ^ Y ^ s PBr, t PhCHi°

THF :H.O CHjCIj JL jl Br * PhCHjO 2 * PhCHjO ^ W Bf 9 :1 v/v 109 110

PhCH.O . y v .CF HaCN ► DHF J k ^ ^CH PhCHjO X 3 C 111

using palladium on carbon gave the target compounds 70, triturated from diethyl ether as the oxalate salt, and 72, crystallized from methanol/ether as the half oxalate salt. Scheme XVI

Synthesis of Imidazolines 70 and 72

PhCH.O HCl

CH EtOH PhCHjO benzena

X X

83 - X - CF,. Y - H 112 - X - CF, Y - H 111 - X - H. Y - CF, 114 - X - H, Y - CF,

• (COjH), PhCH.O

2) NaHOO, PhCHjO 3) (COjH)2 2H,0 X X

113 - X - CF,, Y - H 70 - X - CF,, Y - H • (COjH)j 115 - X - H. Y - CF, 72 - X - H, Y - CF, • H (C0,H),

3.1.4 Trifluoromethylated Norepinephrine Analogs

The synthesis of the 6 -trifluoromethyl norepinephrine derivative

(76) was carried out first because the dibenzyloxy benzaldehyde

precursor 108 was available as described above. The synthesis, as

shown in Scheme XVII, utilized a modified procedure from Kirk et al.

[160] which initially involved the formation of the cyanohydrin

trimethylsilyl ether of benzaldehyde 108 using THSCN and a catalytic amount of Zn^ [161]. Without purification, the cyanohydrin trimethylsilyl ether was reduced with diborane [145] and isolated as the hydrochloride salt to afford the dibenzyloxy protected amino alcohol 116. The initial method of deprotection of amino alcohol 116, using a refluxing equivolume mixture of HCl and methanol, did not result in the desired norepinephrine analog T6 but rather the f t -methoxy derivative 117 confirmed by *H NMR and mass spectra and

elemental analysis. To alleviate this problem, amino alcohol 116 was

deprotected via catalytic hydrogenation with palladium -on carbon to

afford the target compound 76. Interestingly, 76 crystallized with

0.15 mole of ether when recrystallized with methanol/ether as

determined by NMR and elemental analysis. The pure compound was

obtained via recrystallization from methanol/dichloromethane.

Scheme XVII

Synthesis of the 6-Trifluoromethylated Norepinephrine Analog 76

1) THSCN PhCH.O Znl,

PhCH.O 2) BH, ■ THF 0 3) 3N HC1/M«0H

PhCH.O HC1/M«0H

The starting material for the synthesis of the 5-trifluoromethylated derivative 75, was the appropriately substituted benzaldehyde 118 which had been prepared in 32% yield by Backstrom et al. [149] by treating trifluoromethylated phenol 88 with 1 eq. of hexamethylenetetramine

(HMTA) and trifluoroacetic acid. Modifying this procedure by performing it with 2 eq. of HMTA increased the yield of benzaldehyde 105

118 to 42%. Both Kirk et al. [160] and Tantishaiyakul [145] found

methoxy groups to be Inefficient catechol protecting groups in the

synthesis of norepinephrine analogs as boron tribromide demethylation

in the last step resulted in impure compounds which decomposed upon

attempted purification. Therfore, the synthesis of the dibenzyloxy

protected catechol benzaldehyde 119 was attempted (Scheme XVIII) which

can be deprotected by catalytic hydrogenation as described above.

Demethylation of benzaldehyde 118 followed by dibenzylation of the

catechol using benzyl chloride in ethanol gave the protected

benzaldehyde 119 in only 4% yield. By *H NMR, it appeared that much of

the benzyl chloride was spent by reaction with the solvent ethanol.

Acetone cannot be used as the solvent in the benzylation step in the

formation of 119 because proton abstraction on acetone by potassium

carbonate results in nucleophilic attack on the aldehyde moiety by the

acetone anion. When THF was used as the solvent in the dibenzylation

step, none of the desired product 119 was obtained; however, the free

catechol was detected by *H NMR. It was hypothesized that the THF

dielectric constant was too low; therefore, the reaction was performed

using DMF as the solvent. Again, the desired product was not obtained.

Next, a modified procedure from Adejare [162] was tried which involved demethylation of phenol 118. followed by acetal formation using Dowex

50x8-100 and methanol, then dibenzylation using benzyl chloride in acetone, and last hydrolysis of the acetal with 1 N HCl. This resulted

in a 10% yield of the desired trifluoromethylated benzaldehyde 119, as well as a 5% yield of the methyl derivative 120 (structure 106

1 1 ^ confirmed by H and C NMR, IR, and mass spectra). This does not appear to be a synthetically useful route.

Scheme XVIII

Attempted Syntheses of 119

CF PhCH.O CF. 1) BBr, HO HMTA CH.Cl PhCH.O

CF,CO,H CH.O CH.O 119

88 118 EtOH PhCH2OCHjCHj 1) BBr3 ca2c i2 not desired product 2) PhCHjCl KjCOj. KI THF

1) BBr, CH2C12 "► not desired product 2) PhCH2Cl K 2CO,, KI DMF

PhCH20 1) BBr, ch2ci2 PhCHjO 2) Dovex

MeOH 119 (Y: 10%) 3) PhCHjCl r 2CO,. KI CO,CH. Acetone PhCH,0

O 120 (Y: 5%) 107

The alternate synthetic approach to trif luoromethy lated analogs lk_

and 75 began with the appropriately substituted catechols 121 and 122 prepared by boron tribromide demethylation of benzaldehydes 78 and 118,

respectively (Scheme XIX). Catechol 122 had been previously synthesized by Backstrom et al. [149] in 68% yield by treating benzaldehyde 118 with 5 eq. of 1M boron tribromide in dichloromethane

for 2 hours then quenching with HCl followed by recrystallization from toluene. This method was not reproducible by this investigator, as the brown solid obtained was insoluble in toluene and it was not possible to purify with other recrystallization solvent systems. Catechol 122 was prepared by treating benzaldehyde 118 with 3 eq. of 1M boron tribromide in dichloromethane overnight, quenching with water and after workup, recrystallization from ethyl acetate/hexane to yield the desired compound as a white solid in 75% yield. In preparing the iodo analogs of norepinephrine, Tantishaiyakul [145] protected the catechol of the benzaldehyde precursors as trimethylsilyl ethers using trimethylsilyl chloride and triethylamine. In this investigation, the catechols 121 and 122 were also protected as the trimethylsilyl ethers; however, this was achieved in the same step as cyanohydrin trimethylsilyl ether formation by treating the catechol benzaldehydes with 4 eq. of TMSCN as well as a catalytic amount of Z n ^ (Scheme XX).

Tantishaiyakul [145] reduced the trimethylsilyl ether cyanohydrins with diborane and obtained the iodonorepinephrine analogs as oxalate salts.

In this investigation, the resultant trimethylsilyl protected cyanohydrins (123 and 124, structures confirmed by NMR) were reduced 108

without further purification with diborane. However, after quenching

with methanol, attempts to isolate 74 and 75 as the oxalate salts

resulted in a mixture of at least 3 products per example as shown in

each respective NMR spectrum. It was hypothesized that the

resultant mixtures consisted of partially hydrolyzed boron adducts,

which required stronger conditions than methanol followed by oxalic

acid to cleave. Therefore, trimethylsilyl ether cyanohydrins 123 and

124 were reduced with diborane, quenched with methanol and isolated as

the hydrochloride salts. The *H NMR spectra of 74 and 75 show the desired 2- and 5-trifluoromethy1 norepinephrine analogs, respectively.

Attempts are currently being pursued to purify the target compounds.

Scheme XIX

Synthesis of Catechols 121 and 122

CF.

c h 3o HO BBr

CH.Cl CHjO HO 121

CF CF

HO HO. BBr,

CH.Cl 2 CH.O HO

0 0 118 122 109

Scheme XX

Synthetic Approach to 2-Trifluoromethyl (74) and 5-Trlfluoromethyl (75) Norepinephrine Analogs

H A eq. THSCN aixturs of

3 products

2) 3M HCl/MoOH

3.2 BIOLOGY

Trimetoquinol 50 and 5-trlfluoromethyl trimetoquinol 63 have been

examined for comparative activities on /?2 " and //^-adrenergic receptors

in guinea pig trachea and atria, respectively (Table 6). The

5-trifluoromethyl analog 63 has neither agonist nor antagonist activity on trachea C/^)- On atria (//^); however, the 5-trif luoromethyl derivative (63) has weak antagonist activity with a pKg of 5.32 (Table

6). Therefore, substitution of a trifluoromethyl group at the

5-position of TMQ completely abolishes agonist activity on

//-adrenoceptors while imparting weak antagonist activity on

//^-receptors. 110

Table 6

.nvj.ui.co w j . a.iiv/ v«/vy a n u Xiiv/ u (Guinea Pig Trachea and~Atria)

Trachea (02)* Atria (P1)*

Treatnent pECJ0k I.A.e PKCS0‘ I.A.* PKBd

TMQ (SO) alone 7.09 ± 0.11 0.94 * 0.01 .... 7.91 a 0.09 0.86 a 0.02

TMQ plus 6.89 * 0.14 0.91 ± 0.01 .... 5.55 a 0.08 0.80 a 0.08 5.32 ± 0.06 10‘4 M 5CFj-TMQ (63)

'Values are the aean a SEM of N - 3-5 bpEC50 - -log ec50 (M) eI.A. - - aaxinua response of drug relative to the response of 10"s M ISO *pKB - -log [A]/(CR - 1) where [A] - solar concentration of antagonist and

CR - concentration ratio - ECJ0(plus antagonist)/ECso(control)

The 5-trif luoromethyl analog (63) has also been evaluated for its

anti-aggregatory effects on U46619-induced human platelet aggregation

(TXA2 a receptor). 5-Trifluoromethyl TMQ (63) has a pIC^Q of 3.08

(Table 7). This converts to a potency ratio of 5.25 x 10~^ in

comparison to the inhibitory activity of TMQ (50). Thus, introduction

of a trifluoromethyl group at the 5-position of TMQ also dramatically

decreases anti-aggregatory activity.

The enantiomers of 8-fluoroTMQ (S(+)-62 and R(-)-62) have been

evaluated for their /J-adrenergic activity in guinea pig trachea and atria in comparison to TMQ (50). The pEC^g values are 7.41, 6.56, and

5.45 on trachea (/J2) and 7.52, 6.88, and 5.22 on atria (/Jj) for TMQ

(50), S(+)-'62 and R(-)-62, respectively (Table 8). Thus, S(+)-62 is

14-fold and 46-fold more potent than R(-)-62 on /?2~ and

/?^-adrenoceptors, respectively. This follows the same trend as found Ill

Table 7

Comparative Inhibitory Activities of TMQ (50) and 5CF3 -TMQ Against U46619-Induced Human Platelet Aggregation"

Platelets' (Aggregation)

Potency Drug PlC50‘ Ratio

TMQ (50) 6.36 ± 0.13 1.00

5CF3-TMQ (63) 3.08 ± 0.10 5.25 x IQ-4

'Values are the mean ± SEM of N - 3-10 bpIC50 - -log IC50 (M)

for TMQ, that Is S(-)-TMQ is more potent than R( + )-TMQ on /?-systems

[128]. The potency ratios for S(+)-62 are 0.14 on trachea and 0.23 on

atria. This results in a /^//^i selectivity ratio of 0.61 for S(+)-62.

The potency ratios for R(-)-62 are 0.01 on trachea and 0.005 on atria

which corresponds to a /?2//?i selectivity ratio of 2.00. This differs

from the activity of the racemic mixture of 8-fluoroTMQ (62) as shown

in Table 4 on page 71 in which the ft 2 / ft I selectvity ratio was 8.

Moreover, the potency of racemic-8FTMQ (62) was significantly greater than the potencies of either R- or S-8FTMQ in these experiments. Since the racemic mixture of 62 has not been tested in the same assay as the separate enantiomers of 62, it may not be accurate to compare these potencies and ratios. At this time, however, the reasons for this apparent discrepancy are unclear. It is of interest to note that both

S(+)-62 and R(-)-62 are partial agonists on //^-adrenoceptors. 112

Table 8

Activities of Racemic TMQ (50). S(+)-62 and R(-)-62 on Beta-Systems (Guinea Pig Trachea and Atria)

Trachea (0j)J Atria )* Selectivity

Potency Potency Ratio

Drug PEcso6 IA-C R,tio pECsob I A '

TMQ 7.41 ± 0.06 0.98 ± 0.01 1.00 7.52 ± 0.21 1.00 ± 0.00 1.00 1-00

S(+)8F-TMQ 6.56 ± 0.17 0.92 ± 0.04 0.14 6.88 * 0.31 0.82 * 0.06 0.23 061

R(-)8F-TMQ 5.45 * 0.13 0.90 ± 0.04 0.01 5.22 ± 0.33 0.64 ± 0.08 0.005 2.00

'Values are the mean ± SEM of N - 4*6 hpECjo " -log ECJ0 (M) CI.A. - Intrinsic Activity - maximum response of drug relative to the response of 10“5 M ISO

The enantiomers of 8 -fluoroTMQ were also examined for their inhibitory activity against U46619-induced human platelet aggregation.

Both enantiomers were concentration dependent inhibitors of U46619 on human platelets with pIC^Q values of 5.61 and 4.40 for R(-)-62 and

S(+)-62, respectively, and were about 5- and 100-fold less potent than

TMQ (50) [pIC50 = 6.36] (Table 9). In this TXA2 system, R(-)-62 is more potent than S(+)-62, which is in contrast to the stereoselectivity found on /^-adrenergic receptors. This same trend was found for TMQ

(50) on TXA2 systems, in which R(+)-TMQ was more potent than S(-)-TMQ

[131]. To date, the enantiomers of 8 -fluoroTMQ have not been examined for inhibition of U46619-induced contraction on rat thoracic aorta

(TXA2 t receptors).

To date, the trifluoromethylated catecholimidazolines and trifluoromethylated norepinephrine derivatives have not been tested for for biological activity. 113

Table 9

Inhibitory Activities of TMQ (50). S(+)-62 and R(-)-62 on U46619-Induced Human Platelet Aggregation

Platelets' (Aggregation)

Potency Drug PIC5ok Ratio

TMQ 6.36 ± 0.13 1.00

R(-)8F-TMQ 5.61 ± 0.19 0.18

S(+)8F-TMQ 4.40 ± 0.08 0.01

'Values are the mean ± SEM of N - 3-5 bpIC50 - -log IC50 (M)

3.3 SUMMARY

1. The 5-trifluoromethyl derivative of trimetoquinol was synthesized.

2. With respect to fl -adrenergic activity, the 5-position of TMQ is

extremely sensitive to trifluoromethyl substitution which com­

pletely abolishes both and /?2"agonist activities while

imparting weak antagonist activity on (1^ receptors.

3. Compared to TMQ, the 5-trifluoromethyl derivative of TMQ causes a

dramatic decrease in inhibition of U46619-induced human platelet

aggregation•

4. An efficient synthesis of 2-hydroxy-3-(trifluoromethyl)anisole has

been described.

5. Treatment of an amide of an appropriately substituted phenethyl-

amine with Bischler-Napieralski reaction conditions using aceto- 114

nitrile as the solvent, gave an unexpected 3-aminoisoquinoline

pro- duct. This may find applicability as a method of synthesiz­

ing other 3-aminoisoquinoline derivatives.

6 . Current approaches to the synthesis of 8 -trifluoromethylTMQ have

been unsuccessful.

7. The enantiomers of 8 F-TMQ have been separated using a preparative

Chiralcel OD HPLC column.

8 . On /^-adrenergic systems, S(+)8F-TMQ is at least 10-fold more

potent than R(-)8F-TMQ.

9. R(-)8F-TMQ was about 14-fold more potent as an antagonist of

TXA2 -mediated aggregation in human platelets than S(+)8F-TMQ.

10. These results demonstrate that optimal activity for

/^-adrenoceptor agonist and TXA2 antagonist' properties of 8FTMQ

isomers is associated with the S- and R-configuration, respective­

ly.

11. .The synthesis of a series of trifluoromethylated catecholimidazo-

lines has been described.

12. The synthesis of 6 -trifluoromethyl norepinephrine has been

described.

13. The approach to the synthesis of the 2- and 5-trifluoromethylated

norepinephrine derivatives has been described. CHAPTER IV EXPERIMENTAL

Melting points (uncorrected) were determined on a Thomas-Hoover

capillary melting point apparatus. Infrared data were collected on an

Analect RFX-40 FTIR spectrometer. The NMR spectra were obtained on

either an IBM AF-250 FTNMR spectrometer (250MHz) or an IBM AF-270 FTNMR

spectrometer (270MHz) and are reported in parts per million. Mass

spectra were obtained at the College of Pharmacy by use of a Kratos

MS25 RFA double focusing mass spectrometer or at the Ohio State Univer­ sity Chemical Instrumentation Center by use of a VG 70-250S (or Kratos

MS-30) mass spectrometer. Optical rotations were measured on a Perkin-

Elmer 24] polarimeter. CD spectra were taken on a Jasco J-500A spec- tropolarimeter. Elemental analyses were performed by Galbraith Labora­ tories, Inc., Knoxville, TN, and were within +0.4% of the theoretical values for the elements indicated.

Anhydrous tetrahydrofuran was produced by refluxing with and distil­ lation from calcium hydride. Tetrahydrofuran was stored refluxing with sodium using benzophenone as an indicator for dryness. Acetonitrile was dried by refluxing over phosphorus pentoxide for approximately 3 hours followed by distillation. Toluene was dried by refluxing over-

- 115 - night with and distillation from calcium hydride. Anhydrous benzene

was produced by refluxing with and distillation from sodium. The remov­

al of toluene in vacuo was facilitated by the addition of methanol to

form a lower boiling azeotrope.

3,4-Dimethoxy-2-(trifluoromethyl)benzaldehyde (78)

To a solution of 2-iodo-3,4-dimethoxybenzaldehyde (4.12 g, 14.10 mmol) in N-methyl-2-pyrrolidinone (80 ml) in an anhydrous environment under argon was added Cu(I)I (5.37 g, 28.20 mmol) and sodium trifluor- oacetate (7.67 g, 56.40 mmol). The solution was heated under argon.

At 165°C, the solution began to liberate carbon dioxide and was allowed to stir at 175°C for 4 hours. The solution was cooled, diluted with diethyl ether (150 ml) and washed with water (1 x 200 ml). The aqueous layer was extracted once with diethyl ether (200 ml) and the combined ether layers were washed with water (2 x 400 ml), brine (1 x 400 ml), dried (MgSO^) and concentrated to a brown oil. Flash column chromatog­ raphy performed using a gradient solvent system of dichloromethane/ hexane 20:80 to 50:50 (R^ = 0.60 silica on glass using dichloromethane as solvent) afforded a yellow oil, 1.31 g (40%): IR (neat, cm"1) 1687

(C=0); JH NMR (Acetone-D6) r5 10.21-10.20 (br q, 1H, CHO), 7.72 (d, Jm 117

= 8.67, 1H, ArH), 7.42 (d, Jm = 8.64, 1H, ArH), 4.02 (s, 3H, OCH3 ),

3.88 (s, 3H, 0CH3); 13C NMR BB (Acetone-Dg) 5 189.13 (q, JC -C -C -CF3 =

5.4 Hz, CHO), 158.93 (s, Aromatic C bonded to 0CH3) , 149.20 (s, Aromat­

ic C bonded to 0CH3), 129.28 (s, Aromatic C bonded to aldehyde), 126.72

(s, Aromatic C), 125.33 (q, JcF3 = 275.6 Hz, CF3), 124.26 (q, =

30.5 Hz, Aromatic C bonded to CF3), 116.42 (s, Aromatic C), 61.91 (s,

0CH3), 56.96 (s, 0CH3); *^F NMR (CFC13 = 0 ppm external reference)

<5-52.50; MS m/z 234 (M+ , base); Anal. Calcd. for C^gHgF303 : C, 51.29;

H, 3.87. Found: C, 50.94; H, 3.82.

3 ,4-Dimethoxy-2-(trifluoromethyl)benzylalcohol (79)

To a solution of benzaldehyde 78 (4.33 g, 18.49 mmol) In THF:H20

(9:1 v/v) (100 ml) cooled to 0°C was added NaBH^ (1.40 g, 36.98 mmol).

The solution was heated to 40°C for 2.5 hours, cooled and quenched with

H2O and NaH2 P0 ^. The solution was concentrated, taken up Into diethyl ether (100 ml) and H2O (100 ml), the layers separated and the aqueous layer was extracted with diethyl ether (2 x 100 ml). The combined organic layers were dried (MgSO^) and concentrated to a yellow oil.

Gravity column chromatography was performed using a gradient solvent 118

system of ethyl acetate/hexane 20:80 to 50:50 to yield a yellow oil,

3.81 g, (87%): IR (neat, cm"*) 3600-3200 (OH), absence of C=0; *H NMR

(CDCI3 /TMS) <5 7.29 (d, Jm = 8.67 Hz, 1H, ArH), 7.06 (d, Jm = 8.61 Hz,

1H, ArH), 4.76 (d, J = 1.20 Hz, 2H, CH2), 3.89 (s, 3H, OCH3 ), 3.88 (s,

3H, OCH3 ), 1.97 (br s, 1H, OH); MS m/z 236 (M+ , base); Anal. Calcd.

for C 10Hn F303 : C, 50.85; H, 4.69. Found: C, 50.63; H, 4.81.

3.4-Dlmethoxy-2 -(trifluoromethyl)benzylbromide (80)

To a solution of benzylalcohol 79 (463 mg, 1.96 mmol) in dichloro­

methane (10 ml) cooled to 0°C was added phosphorus tribromide (0.20 ml,

2.16 mmol). The solution was allowed to stir for 2 hours at 0°G and quenched with ice water (10 ml), saturated NaHC03 solution (10 ml) and dichloromethane (10 ml) was added. The organic layer was separated, washed with H20 (2 x 20 ml), brine (1 x 20 ml), dried (MgSO^) and con­ centrated to a light yellow oil, 431 mg (73.5%): IR (neat, cm”*) absence of OH; *H NMR (CDCI3/TMS) 8 7.16 (d, JQ = 8.6 Hz, 1H, ArH),

7.02 (d, JQ = 8.6 Hz, 1H, ArH), 4.62 (q, J = 1.4 Hz, 2H, CH2 Br), 3.90

(s, 3H, 0CH3), 3.89 (s, 3H, OCH3 ); MS m/z 219 (M+ - Br, base); Anal.

Calcd. for C 10H 10BrF302 : C, 40.16; H, 3.37. Found: C, 40.55; H, 3.47. 119

3 .4-Dimethoxy-2-(trifluoromethyl)benzvlcyanide (81)

CF CH

CH.

To a solution of benzylbromide 80 (2.21 g, 7.39 mmol) in DMF (50 ml)

was added NaCN (1.81 g, 36.94 mmol) and the solution allowed to stir at

room temperature for 30 minutes. Water (100 ml) was added and the

solution was extracted with dichloromethane (3 x 100 ml). The combined

organic layers were washed with H20 (2 x 300 ml), brine (1 x 100 ml), dried (MgSO^), and concentrated to an orange oil. Gravity column chro­ matography was performed using a gradient solvent system of ethyl acetate/hexane 20:80 to 50:50. The resultant light yellow oil was crystallized from ethyl acetate/hexane to afford white needles, 1.67'g

(92%): m.p. 41-43°C; IR (FCBr, cm'1) 2253 (CN); JH NMR (CDCI3/TMS) S

7.27 (d, JQ = 8.6 Hz, 1H, ArH), 7.07 (d, JQ = 8.6 Hz, 1H, ArH), 3.91

(s, 3H, OCH3 ), 3.89 (s, 3H, OCH3 ), 3.85 (q, J = 1.9 Hz, 2H, CH2CN); MS m/z 245 (M+ , base); Anal. Calcd. for Cj^H^FgNC^: C, 53.88; H, 4.11; N,

5.71. Found: C, 53.76; H, 4.13; N, 5.66. 120

3 .4-Dlhydroxy-2-(trlfluoromethyl)benzylcyanide (82)

To a solution of benzylcyanide 81 (1.66 g, 6.77 mmol) in dichloro­ methane (25 ml) cooled to 0°C under argon was added 1 M boron tribrom­ ide in dichloromethane (20.30 ml). The solution was allowed to stir overnight at room temperature, concentrated under reduced pressure and taken up into ethyl acetate (100 ml) and water (100 ml). The organic layer was separated and washed with water (3 x 100 ml), brine (1 x 100 ml), dried (MgSO^), treated with charcoal and concentrated to a brown powder. Recrystallization from ethyl acetate/hexane afforded crystals,

1.21 g (82%): m.p. 149-151°C; IR (KBr, cm"1) 3400-3100 (br, OH), 2280

(CN); 1li NMR (Acetone-Dg) r> 7.08 (d, JQ = 8.2 Hz, 1H, ArH), 6.93 (d,

JQ = 8.2 Hz, 1H, ArH), 3.94 (q, J = 2.0 Hz, 2H, CH2 CN); MS m/z 217 (M+ , base); Anal. Calcd. for C9HgF3N02 : C, 49.78; H, 2.78; N, 6.45. Found:

C, 49.77; H, 2.72; N, 6.30. 121

3,4-Dibenzyloxy-2-(trifluoromethyl)benzylcyanIde (83)

PhCH

PhCH

To a solution of benzyl cyanide 82 (1.15 g, 5.30 mmol) in acetone

(50 ml) was added potassium carbonate (1.65 g, 11.93 mmol), potassium

iodide (133 mg, 0.80 mmol) and 97% benzyl chloride (1.40 ml, 11.93

mmol). The solution was stirred overnight at reflux, cooled, concen­

trated under reduced pressure and taken up into ethyl acetate (50 ml)

and water (50 ml). The ethyl acetate layer was washed with water (3 x

50 ml), brine (1 x 50 ml), dried (MgSO^) and concentrated. The residue

was passed through a gravity silica column using 20% ethyl acetate/80%

hexane (R^ =0.38 silica on glass using 20% ethyl acetate/80% hexane as

solvent) to yield an orange oil, 1.81 g (86%): IR (neat, cm’1) 2254

(CN); 1H NMR (CDCI3 /TMS) S 7.44-7.30 (m, 11H, ArH), 7.17 (d, JQ = 8.6

Hz, 1H, Aril), 5.17 (s, 2H, ArCH2 0 ) , 5.06 (s, 2H, ArCH20) , 3.88 (br q,

2 H , CH2CN); MS m/z 306 (M+ - Benzyl), 91 (base); Anal. Calcd. for

C23H 18F3N02 : C » 69-52 ; fI> 4 -57; N, 3-52. Found: C, 69.60; H, 4.53; N,

3.39. 122

N-f 2-f 3,4-Dibenzyloxy-2-(trifluoromethyl)phenyllethyl1 -3.4.

5-trimethoxyphenylacetamide (85)

OCHg

To a solution of benzylcyanide 83 (1.61 g, 4.05 mmol) in dry THF (50

ml) cooled to 0°C under argon was added 1 M BHg'THF (16.20 ml). The

solution was refluxed overnight under argon, cooled to 0°C, quenched

with methanol (50 ml) and concentrated. The residue was taken up into

methanol (50 ml) and concentrated (repeated two more times). The

resultant oil was taken up into toluene (50 ml) and to the solution was added 3,4,5-trimethoxyphenylacetic acid (0.92 g, 4.05 mmol). The solu­

tion was refluxed for 72 hours with removal of water via a Dean-Stark

trap, cooled and concentrated under reduced pressure to an oil. The oil was taken up into dichloromethane ((50 ml), washed with water (1 x

50 ml), 10% HC1 solution (2 x 50 ml), saturated NaHCOj solution (2 x 50 ml), brine (1 x 50 ml), dried (MgSO^) and concentrated to a beige sol­ id. Recrystallization from toluene/hexane yielded a white solid, 2.10 g (85%): m.p. 88-90°C; IR (KBr, cm"1) 3302 (NH), 1641 (C=0); XH NMR

(CDCI3 /TMS) <5 7.47-7.29 (m, 10H, ArH), 7.00 (d, JQ = 8.5 Hz, 1H, ArH),

6.72 (d, JQ = 8.5 Hz, 1H, ArH), 6.41 (s, 2H, ArH), 5.48 (br t, 1H, NH),

5.14 (s, 2H, ArCH20), 5.02 (s, 2H, ArCHjO), 3.85 (s, 3H, OCH3 ), 3.82 123

(s, 6 H , 2 x OCH3), 3.47-3.40 (m, 4H, 2 x ArCH2), 2.92-2.87 (m, 2H,

CH2N); MS m/z 609 (M+), 91 (base); Anal. Calcd. for C3 4 H34F3N06 : C,

66.99; H, 5.62; N, 2.30. Found: C, 66.69; H, 5.63; N, 2.35.

6 ,7-Dibenzyloxy-l-(3.4,5-trlmethoxybenzyl)-5-trifluoromethyl-

1,2,3,4-tetrahydroisoquinoline hydrochloride (8 6 )

CF

•HC1 NH PhCH-0 OCH,

N OCH OCH3

To a solution of amide 85 (816 mg, 1.34 mmol) in dry acetonitrile

(50 ml) under argon was added phosphorus oxychloride (0.87 ml, 9.37

mmol). The solution was stirred at reflux for 5 hours, cooled and con­

centrated under reduced pressure. The residue was taken up into metha­

nol (50 ml) and concentrated (repeated two more times). The residue

was taken up into methanol (50 ml), cooled to 0°C and to the solution

was added sodium borohydride (253 mg, 6.70 mmol). The solution was

allowed to stir overnight at room temperature under argon and concen­

trated under reduced pressure. The residue was dissolved into diethyl

ether (100 ml), water (50 ml) and 10% NaOH solution (50 ml). The water

layer was extracted with diethyl ether (2 x 50 ml). The combined organic layers were washed with water (2 x 300 ml), brine (1 x 300 ml), dried (MgS04 ) and concentrated to a yellow oil. The oil was passed through a gravity silica column using 1% methanol/99% chloroform as the 124

solvent. The resultant oil was converted to the hydrochloride salt

using 3 N methanolic HC1. Recrystallization from dichloromethane/

hexane yielded white crystals, 338 mg (40%): m.p. 146-149°C; *H NMR

(CDCI3 /TMS) <5 7.38-7.30 (m, 10H, ArH), 6.48 (s, 1H, ArH), 6.42 (s, 2H,

ArH), 5.04 (d, 1H, HA of HA B , J = 9.9 Hz, ArCH20 ) , 4.96 (d, 1H, Hg of

HA B . J = 9.9 Hz, ArCH2 0), 4.86 (d, 1H, HA of HA B , J = 11.7 Hz, ArCH20 ) ,

4.77 (d, 1H, Hb of Ha b , J = 11.7 Hz, ArCH20 ) , 3.81 (s, 3H, OCH3 ) , 3 '79

(s, 6 H, 2 x OCH3 ), 3.57-3.50 (m, 1H, ArCHN), 3.22-3.10 (m, 6 H,

ArCH2 CH2N and ArCH2); MS m/z 593 (M+ - HC1), 412 (M+ - trimethoxybenzyl

- HC1), 322, 181, 91 (base); Anal. Calcd. for ^ H ^ C ^ N C ^ : C, 64.81;

H, 5.60; N, 2.22. Found: C, 64.69; H, 5.59; N, 2.19.

6 ,7-Dihydroxy-l-(3.4.5-trimethoxybenzyl)-5-trifluoromethyl-1,2.3,

4-tetrahydroisoquinoline hydrochloride (63)

CF HO

HO OCH

OCH OCH3

A solution of tetrahydroisoquinoline 86 (150 mg, 0.24 mmol) in an equivolume mixture of methanol:concentrated HCl (20 ml) was refluxed for 2 hours, cooled and concentrated under reduced pressure. The resi­ due was taken up into methanol (25 ml) and concentrated (repeated two more times). The resultant solid was recrystallized from methanol/ diethyl ether to afford a white powder, 81 mg (75%): m.p. 260-263°C 125 »

with decomposition; IR (KBr, cm"*) 3400-3200 (br, OH); *H NMR (CD3OD)

f5 6.81 (s, 1H, ArH), 6.63 (s, 2H, ArH), 4.74 (dd, J = 5.8 and 8.5 Hz,

1H, ArCHN), 3.83 (s, 6 H, 2 x 0CH3), 3.76 (s, 3H, OCH3 ), 3.49-3.02 (m,

6 H, ArCH2CH2N and ArCH2 ); MS (FAB) 414 (M+ - HC1), 232 (base); Anal.

Calcd. for C 2 0 H23C1F3N05 : C, 53.40; H, 5.15; N, 3.11. Found: 53.44; H,

5.21; N, 3.10.

2 -Hydroxy-3-(trifluoromethyl)anisole (8 8 )

To a solution of 2.15 M n-butyllithium in hexane (63.60 ml) in dry

THF (200 ml) cooled to -78°C under argon, was added dropwise a solution of commercially available (Aldrich) 3-trifluoromethylanisole (20.00 g,

113.50 mmol) in dry THF (20 ml). After complete addition, the mixture was warmed to 0°C and allowed to stir for 4 hours. Cu(I)Br (16.28 g,

113.50 mmol) was then charged into the reaction vessel and the solution allowed to stir for an additional 2 hours at 0°C. The flask was equipped with a pasteur pipette with an attached drying tube. Dry air was drawn into the reaction mixture at 0°C for 45 minutes by applying a vacuum aspirator to the flask. Concentrated HC1 (22.40 ml) was added and the solution stirred at room temperature for 20 hours. 10% HC1 126

solution (75 ml) was added and the solution was extracted with ether (2

x 200 ml). The combined ether layers were washed with saturated NaHC03

solution (1 x 100 ml), dried (MgS04 ) and concentrated under reduced

pressure to a brown oil. Gravity column chromatography was performed

using a gradient solvent system of dichloromethane/hexane 20:80 to

40:60 (1?£ = 0.67 silica on glass using dichloromethane as solvent).

The resultant light yellow solid was recrystallized from ethanol/water

to yield white needles, 12.77 g (58.5%): m.p. 51-53°C; FeClg positive;

IR (KBr, cm"1) 3410 (OH); !H NMR (5 (Acetone-Dg) 9.65 (br s, 1H, OH),

7.21 (d, J = 8 Hz, HI, ArH), 7.06 (dd, Jffl = 1.1 Hz, JQ = 8 Hz, 1H,

ArH), 6.88 (apparent t, J = 8 Hz, ArH), 3.84 (s, 3H, 0CH3^. MS m/z 192

(M+); Anal. Calcd. for CgH704 F3 : C, 50.05; H, 3.67. Found: C, 49.99;

H, 3.63.

N,N-Dimethyl-4-hydroxv-3-methoxy-5-(trifluoromethyl)benzylamine oxalate

(90)

c h 3o N(CH-)

• ( C 0 2H) HO

CF 3

To a solution of 37% formaldehyde (6.34 ml, 78.06 mmol) and 40% dimethylamine (8.78 ml, 78.06 mmol) in ethanol (100 ml) was added phe­ nol (8 8 ) (10.00 g,52.04 mmol). The solution was heated to reflux for 2 hours, cooled and concentrated under reduced pressure. The resultant solid was dissolved into methanol (50 ml) to which oxalic acid dihyd­

rate (6.56 g, 52.04 mmol) in methanol (20 ml) was added. The white

crystalline precipitate was collected to yield 13.23 g (75%): m.p.

201-202°C; IR(KBr, cm"1) 3026 (NH), 3227 (OH); XH NMR (DMS0-D6 ) S 7.36

(s, 1H, ArH), 7.22 (s, 1H, ArH), 4.11 (s, 2H, CH2), 3.85 (s, 3H, 0CH3) ,

2.26 (s, 6 H, CH3x2); MS m/z 249 (M+ - C2H204 ), 205 (M+ - C2 H204 -

NMe2 ), 185, 58 (base); Anal. Calcd. for Cj-jHj^F^NOg: C, 46.02; H, 4.75;

N, 4.13. Found: C, 45.88; H, 4.77; N, 4.10.

4-Hydroxy-3-methoxy-5-(trifluoromethyl)benzylcyanide (91)

To a solution of the free base of benzylamine 90 (7.32 g, 29.37

mmol) in dichloromethane (100 ml) was added CH3 I (14.00 ml) and the

solution allowed to stir overnight at room temperature. The solution was concentrated under reduced pressure and the residue dissolved in

DMSO (100 ml) to which NaCN (2.16 g, 44.05 mmol) was added and the solution allowed to stir at room temperature for 7 hours. The reaction mixture was diluted with H20 (100 ml), acidified to pH 1 (litmus) with

6 N HC1 and extracted with ethyl acetate (3 x 200 ml). The combined organic layers were washed with water (2 x 600 ml), brine (1 x 600 ml), dried (MgS04) and concentrated under reduced pressure to a yellow sol­ 128

id. The solid was recrystallized from ethyl acetate/hexane to yield

yellow needles, 4.90 g (72%): m.p. 110-111°C; IR (KBr, cm"1) 2264 (CN),

3364 (OH); NMR (CDCl-j/TMS) <5 7.06 (d, 1H, J = 1.2 Hz ArH), 6.99 (d,

1H, J = 1.7 Hz, ArH), 6.15 (s, 1H, OH), 3.97 (s, 3H, OCH3 ), 3.72 (s,

2H, CH2); MS m/z 231 (M+ , base); Anal. Calcd. for C 10H8F3N02 : C, 51.96;

H, 3.49; N, 6.059. Found: C, 51.93; H, 3.45; N, 4.10.

3,4-Dihydroxy-5-(trifluoromethyDbenzylcyanide (92)

To a solution of benzylcyanide 9J. (3.63 g, 15.70 mmol) in dichloro­ methane (25 ml) cooled to 0°C under argon was added 1 M boron tribrom­ ide in dichloromethane (31.40 ml). The solution was warmed to room temperature and allowed to stir overnight at which time the reaction mixture was concentrated under reduced pressure and taken up into ethyl acetate (250 ml). The organic solution was washed with water (2 x 250 ml), brine (1 x 250 ml), dried (MgSO^) and concentrated to a white sol­ id which was recrystallized from ethyl acetate/hexane to afford a white solid, 3.27 g (96%): m.p. 151-152°C; IR (KBr, cm"1) 3230 (br, OH), %

2267 (CN); XH NMR (Acetone-D6) & 8.85 (br s, 2H, 2 x ArOH), 7.14 (d,

Jm = 1.5 Hz, 1H, ArH), 7.07 (d, Jffl = 1.5 Hz, 1H, ArH), 3.88 (s, 2H, 129

CH2); MS m/z 217 (M+), 197 (base); Anal. Calcd. for CgH6F 3N02 : C,

49.78; H, 2.78; N, 6.45. Found: C, 49.46; H, 2.89; N, 6.41.

3 ,4-Dibenzyloxy-5-(trifluoromethyl)benzylcyanlde (93)

PhCH20

PhCH20 c f 3

To a solution of benzylcyanide 92 (3.83 g, 17.64 mmol), potassium

carbonate (5.48 g, 39.69 mmol) and potassum iodide (0.44 g, 2.65 mmol)

in acetone (100 ml) was added 97% benzyl chloride (4.71 ml, 39.69

mmol). The solution was stirred at reflux for 5 hours, cooled and con­

centrated under reduced pressure. The residue was dissolved into ethyl

acetate (100 ml) and water (100 ml). The organic layer was separated

and washed with water (2 x 100 ml), brine (1 x 100 ml) and dried

(MgSO^) and concentrated to an orange oil. The oil was passed through

gravity silica column using 20% ethyl acetate/80% hexane as the solvent

to yield a yellow solid which was recrystallized from ethyl acetate/

hexane to afford white cubic crystals, 5.79 g (82%): m.p. 68-70°C; IR

(KBr, cm"1) 2254 (CN); ]H NMR (CDCI3 /TMS) <5 7.44-7.30 (m, 10H, ArH),

7.17 (s, 1H, ArH), 7.14 (s, 1H, ArH), 5.17 (s, 2H, ArCH20 ) , 5.09 (s,

2H, ArCH20), 3.73 (s, 2H, CH2CN); MS m/z 306.0751 (M+ - Bn), 91 (base);

Anal. Calcd. for C23H 18F 3N02 : C, 69.52; H, 4.57; N, 3.52. Found: C,

69.59; H, 4.56; N, 3.56. 130

2-f 3,4-Dibenzyloxy-5-(trifluoromethyl)phenyl1ethylamlne hydrochloride

(94) .

PhCH20

PhCH20

To a solution of benzylcyanide 93 (5.10 g, 12.83 mmol) in dry THF

(100 ml) cooled to 0°C under argon was added 1 M BHg'THF solution (51.3

ml). After refluxing for 18 hours under argon, the solution was cooled

to 0°C and cautiously quenched with methanol (50 ml). The solution was

concentrated under reduced pressure, taken up into methanol (50 ml) and

concentrated again (repeated once again). The resultant colorless oil

was taken up into methanol (20 ml) and with ice bath cooling 6 N metha-

nolic HC1 (2.25 ml) was added. The solution was concentrated, taken up

into methanol (100 ml) and concentrated (repeated once again). Recrys­

tallization from methanol/diethyl ether yielded a white solid, 5.16 g

(90%): m.p. 147- 150°C; IR (KBr, cm-1) 3400 (NH); XH NMR (CD3 OD) <5

7.52-7.26 (m, 11H, ArH), 7.15 (d, Jm = 1.8 Hz, 1H, ArH), 5.23 (s, 2H,

ArCH20), 5.04 (s, 2H, ArCH20), 3.23-3.17 (m, 2H, CH2N), 3.01-2.95 (m,

211, ArCH2); MS m/z 401 (M+ - HC1), 91 (base); Anal. Calcd. for

C23H23C1F3N02 1/2 H20: C > ■ 82; H,5.41; N, 3.13. Found: C, 61.83; H,

5.35; N, 3.16. N-f 2 -f 3.4-Dlbenzyloxy-5-(trifluoromethyl)phenyl1 ethyl1-3,4.

5-trimethoxyphenylacetamide (95)

P h C H ,0

OCHg

Phenylethylamine hydrochloride 94 (4.76 g, 10.65 mmol) dissolved

into dichloromethane (100 ml) was washed with saturated NaHCOg solution

(3 x 20 ml), water (3 x 100 ml), dried (MgSO^) and concentrated to an

oil. The oil was dissolved into toluene (100 ml) and to the solution

was added commercially available (Aldrich) 3,4,5-trimethoxyphenylacetic

acid (2.41 g, 10.65 mmol). The mixture was heated to reflux for 72

hours with removal of water via a Dean-Stark trap, cooled, and concen­

trated to a yellow solid. The solid taken up into dichloromethane (75

ml) was washed with water (1 x 75 ml), 10% HC1 solution (2 x 75 ml),

water (1 x 75 ml), saturated NaHCOg solution (2 x 75 ml), and brine (1

x 75 ml), dried (MgSO^) and concentrated under reduced pressure to a yellow oil which solidified upon standing. Recrystallization from hot

toluene/diethyl ether produced white crystals, 5.15 g (79%): m.p.

114- 115°C; IR (KBr, cm'1) 3280 (NH), 1649 (C=0); XH NMR (CDClg/TMS) <3

7.46-7.30 (m, 10H, ArH), 7.01 (s, 1H, ArH), 6.98 (s, 1H, ArH), 6.40 (s,

2H, 2 x ArH), 5.51 (br t, 1H, NH), 5.11 (s, 2H, ArCH20), 5.06 (s, 2H,

ArCH20), 3.83 (s, 3H, OCH3 ), 3.82 (s, 6 H, 2 x OCH3 ), 3.51-3.43 (m, 4H, 132

ArCH2C and ArCH2CO), 2.78 (t, J = 7.1 Hz, CH2 N ) ; MS m/z 609 (M+), 91

(base); Anal. Calcd. for C 3 4 H3AF3N06 : C, 66.99; H, 5.62; N, 2.30.

Found: C, 6 6 .8 6 ; H, 5.54; N, 2.34.

3-f 2-f 3,4-dibenzyloxy-5-(trifluoromethy1)phenyl1 ethyl1 amino-1-methy1-

6 ,7,8 -trimethoxylsoquinoline hydrochloride (96)

PhCH 2 0 ’ HC1

3 P h C H ,0 o c h 3

o c h 3

o c h 3

To a solution of amide 95 (1.00 g,1.64 mmole) in dry acetonitrile

(30 ml) under an argon atmosphere was added phosphorus oxychloride

(0.77 ml, 8.20 mmole) and heated to reflux. After 5 hours, the solu­

tion was cooled and concentrated under reduced pressure. The residue was taken up into ethyl acetate (50 ml) and washed with 10% NaOH solu­

tion (2 x 50 ml), water (1 x 50 ml), brine (1 x 50 ml), dried (MgS04 ) and concentrated to a yellow oil. The oil was passed through a gravity silica column using 30% ethyl acetate/70% hexane as the solvent. The hydrochloride salt was made with 3 N methanolic HC1 and recrystallized from dichloromethane/diethyl ether to afford yellow crystals, 121 mg

(11%): m.p. 138-140°C; *H NMR (CDCl3 /TMS) <5 7.89 (br t, 1H, NH), 7.49

(d, 1H, Jm = 1.59, ArH), 7.38-7.26 (m, 10H, ArH), 7.08 (d, 1H, Jffl =

1.59, ArH), 6.51 (s, 1H, ArH), 6.26 (s, 1H, ArH), 5.26 (s, 2H, ArCH20 ) ,

5.00 (s, 2H, ArCH20), 4.00 (s, 3H, 0CH3), 3.96 (s, 3H, 0CH3), 3.85 (s, 133

3H, OCH3), 3.49-3.45 (m, 2H, ArCH2), 3.11 (s, 3H, CH3 ), 3.02 (br t, 2H,

CH2N); MS m/z 261 (base), 248, 91; Anal. Calcd. for C 3 4 H36C1F3N205 : C,

63.30; H, 5.62; N, 4.34. Found: C, 63.32; H, 5.45; N, 4.05.

N-Trifluoroacetyl-6 .7-dibenzyloxy-8-lodo-l-(3.4.5-trimethoxybenzyl)-

1,2,3,4-tetrahydroisoquinoline (99)

PhCHjO

CF PhCHjO OCH

OCH och3

To a solution of 6,7-dibenzyloxy-8-iodo-l-(3,4,5-trimethoxybenzyl)-

1,2,3,4-tetrahydroisoquinoline [145] (710 mg, 1.07 mmol) in dichloro­

methane (25 ml) was added pyridine (169 mg, 2.14 mmol). Trifluoroacet-

ic anhydride (271 mg, 1.29 mmol) was added dropwise to the solution

followed by the addition of a catalytic amount of dimethylaminopyridine

(20 mg, 0.16 mmol). The solution was stirred overnight at room temper­ ature and then diluted with dichloromethane (25 ml). The organic layer was washed with water (1 x 50 ml), 10% HC1 solution (1 x 50 ml), water

(1 x 50 ml), saturated NaHC03 solution (1 x 50 ml), water (1 x 50 ml), dried (MgSO^), and concentrated to a brown foam. Gravity column chro­ matography was performed using 10% diethyl ether/90% dichloromethane as the solvent to yield a white foam. Crystallization from ethyl acetate/ hexane produced a white solid, 579 mg (72%); m.p. 131-134°C; IR (KBr, cm'1) 3446 (NH), 1678 (C=0), 1591 (amide N); NMR (CDC13/TMS) 8 134

7.53-7.33 (m, 10H, ArH), 6.82 and 6.74 (1H In total, ArH), 6.38 and

6.26 (2H In total, ArH), 5.95-5.90 (m, 1H, ArCHN), 5.13-5.12 (m, 2H,

ArCH2 0), 5.07-4.97 (m, 2H, ArCH2 0), 3.82 and 3.81 (3H in total, 0CH3),

3.80 and 3.74 (6H in total, OCH3 x 2), 3.80-3.70 (m, obscured under

methoxy peaks), 3.51-3.44 (m, 1H, CH2), 3.20-2.72 (m, 3H, CH2),

2.43-2.36 (m, 1H, CH2); *^F NMR (CFCI3 = 0 ppm as internal reference)

<5 -68.59, -70.03 (.23 to 1 ratio); MS FAB 747 (M+ ), 566, 438, 348,

181, 91 (base); Anal. Calcd. for C^HggF^INOg: C, 56.24; H, 4.45; N,

1.87. Found: C, 55.98; H, 4.41; N, 1.81.

2 ,3-Dibenzyloxy-5,6.13.13a-tetrahydro-l-fodo-9.10.11-trimethoxy-

8 H-dibenzofa.glquinolizlne hydrochloride (1 0 2 )

■ HC1

OCH

OCH OCH3

To a solution of HMPA (20 ml) and copper powder (769 mg, 12.11 mmol) in a steel tube was added trifluoromethyl iodide (2.59 g, 13.22 mmol).

The solution was stirred and heated to 120°C for 2.5 hours, and cooled.

n In an Atmosbag (Aldrich) purged with and under argon, the solution was filtered through celite (to remove excess copper) into a 50 ml round bottom flask equipped with stir bar and 6 ,7-dibenzyloxy-

8 -iodo-1-(3,4,5-trimethoxybenzy1)-1,2,3,4-tetrahydroisoquinoline [145]

(400 mg, 0.61mmol). The resultant green solution was stirred under argon at room temperature overnight. The mixture was heated to 70°C and monitored by TLC (1% methanol/99% chloroform and several drops of ammonium hydroxide). After 1 week, TLC showed the absence of starting material and the reaction mixture was cooled, diluted with ice water

(25 ml) and ethyl acetate (25 ml), and the layers separated. The water layer was extracted with ethyl acetate (25 ml). The combined organic layers were washed with water (2 x 50 ml), brine (1 x 50 ml), dried

(MgSO^) and concentrated. The residue was passed through a silica gravity column using 55% ethyl acetate/45% hexane. The resultant oil was converted to the hydrochloride salt using 3 N methanolic HC1 and treated with charcoal in methanol. Recrystallization from methanol/ diethyl ether yielded white crystals, 61 mg (14%): m.p. 202-203°C with decomposition; JH NMR (CDClg/TMS) S 7.49-7.30 (m, 10H, ArH), 6.85 (s,

1H, ArH), 6.43 (s, 1H, ArH), 5.11 (s, 2H, ArCH20), 5.06 (d, HA of HAR,

1H, J = 10.1 Hz, ArCH20), 4.96 (d, H0 of HA B , 1H, J = 10.1 Hz, ArCH20),

4.77-4.70 (m, 1H, ArCHN), 4.64 (d, HA of HA B , 1H, J = 16.6 Hz,

C-8 -methylene), 4.45. (d, HB of HAB, 1H, J = 16.6 Hz, C-8 -methylene),

3.94 (s, 3H, OCH3 ), 3.86 (s, 3H, OCH3 ), 3.85 (s, 3H, OCH3 ), 3.94-3.85

(m, obscured under methoxy peaks, 1H, CH2), 3.75-3.65 (m, 1H, CH2),

3.41-3.38 (br m, 2H, CH2), 2.99-2.69 (m, 2H, CH2); MS m/z 663 (M+ ), 91

(base); Anal. Calcd. for C 3 4 H35C1IN05 1H20: C, 56.88; H, 5.19; N,

1.95. Found: C, 56.56; H, 5.00; N, 1.97. 136

2,3-Dlbenzyloxy-12-chloro-5,6,13,13a-tetrahydro-9.10,11-trlmethoxy-8H-

dlbenzofa.glquinolizlne (103)

• HC1 PhCH-0 OCH

OCH och3

A mixture of 6 ,7-dlbenzyloxy-8-iodo-l-(trimethoxybenzy1)-1 ,2 ,3,4-

tetrahydroisoquinollne hydrochloride [145] (60 mg, 0.09 mmol), 37% for­

maldehyde (4 ml), ethanol (2.5 ml) and water (4 ml) was refluxed for 3

hours, cooled and basified with NH^OH. The solution was extracted with

ethyl acetate (20 ml) and the ethyl acetate layer was washed with water

(1 x 20 ml), brine (1 x20 ml), dried (MgSO^) and concentrated to an oil which solidified on standing. The solid was treated with 3 N methanol-

ic HC1 and recrystallized from methanol/diethyl ether gave white crys­ tals, 50 mg (76%): m.p. 195-196°C with decomposition; *H NMR

(CDCI3 /TMS) (5 7.49-7.31 (m, 10H, ArH), 6.85 (s, 111, ArH), 5.12-5.11

(m, 2H, ArCH20)', 5.07 (d, HA of HAfi, 1H, J = 10.2 Hz, ArCH20 ) , 5.00 (d,

HB of HAB» 1H> J = 1 0 '2 Hz> ArCH20), 4.76-4.69 (m, 1H, ArCHN), 4.63 (d,

Ha of Ha b , 1H, J = 17.1 Hz, C-8 -methylene), 4.50 (d, Hg of HA B , 1H, J =

17.1 Hz, C-8 -methylene), 3.94 (s, 3H, OCH3 ), 3.925 (s, 3H, OCH3 ), 3.922

(s, 3H, OCII3 ), 3.87-3.79 (m. obscured under methoxy peaks, 2H, CH2),

3.44-3.33 (m, 2H, CH2), 3.01-2.94 (m, 1H, CH2), 2.63-2.51 (m, 1H, CH2);

MS m/z 699 (M+ , 37C1), 697 (M+ , 3 5 C1), 468, 230 (3 7 C1), 228 (35C1), 91 137

(base); Anal. Calcd. for C3 4 H34Cl2 IN05 1.5H2 0: C, 53.63; H, 4.90; N,

1.84. Found: C, 53.68; H, 4.69; N, 2.14.

Preparative Chiralcel OP- HPLC Separation of the Enantlomers of

6 ,7-Dibenzyloxy-8-fluorotrimetoquinol (104) PhCH20 PhCH-0 • HC1 NH PhCHjO OCH OCH

OCH OCH OCH o c h 3 3

Enantiomer 1 of 104 Enantiomer 2 of 104

The free base of 6 ,7-dibenzyloxy-8-fluoro-1-

(3,4,5-trimethoxybenzyl)- 1,2,3,4-tetrahydroisoquinoline (104) [135]

was resolved on a preparative Chiralcel 0D^ (Diacel) HPLC column (2 cm

x 50 cm). The mobile phase was 70/30 hexane/isopropanol and the flow

rate was 6 ml/min. The separation was performed using a Beckman System

Gold using a UV detector: k'j = 2.51; k '2 = 3.53; a = 1.41; Rg =

0.69; Enantiomer 1 : Retention time = 88.79 minutes; Enantiomer 2:

Retention time = 114.71 minutes. The hydrochloride salts were made using 3 N methanolic HC1 and crystallized from diethyl ether:

Enantiomer 1HC1: m.p. 207.5-209.5°C with decomposition; NMR (CD30D)

S 7.47-7.26 (m, 10H, ArH), 6. 8 6 (s, 1H, ArH), 6.53 (s, 2H, ArH), 5.16

(s, 3H, ArCH20), 5.02 (s, 2H, ArCH2 0), 4.97-4.92 (m, 2H, ArCHNH), 3.79

(s, 6 H , 0CH3 x 2), 3.73 (s, 3H, 0CH3), 3.53-3.44 (m, 1H, CH2),

3.35-3.26 (m, obscured by CD30D peak, 2H, CH2) , 3.13-3.05 (m, 3H, CH2); 138

[

ellipticity, in MeOH) t^l2 37 = +15,030.

Enantiomer 2HC1: m.p. 208-209°C; NMR (CD3OD) same as enantiomer 1

described above; [a]25D = -14.7° (c = 4.6 mg/ml in MeOH); CD (reported

in molecular ellipticity, in MeOH) J234 = "17>403.

S(+)-3,4-Dihydroxy-8-fluoro-l-(3,4,5-trimethoxybenzyl)-

1,2,3,4-tetrahydroisoquinoline (S(+)-62)

HO

NH HO OCH

OCH OCH3

To a Parr bottle containing 10% Pd/C (18 mg) purged with argon was added Enantiomer 1 of 104 (180mg, 0.31 mmol) in methanol (50 ml). The solution was hydrogenated at 40 psi. overnight, filtered through celite and the filtrated was concentrated to a dark orange oil. Trituration from diethyl ether gave a beige solid, 44 mg (35%): d.p. 115-120°C;

FeCl3 positive; JH NMR (CD3OD) <5 6.56 (s, 2H, ArH), 6.51 (s, 1H,

ArH), 3.81 (s, 6 H, OCH3 x 2), 3.74 (s, 3H, OCH3 ), 3.49-2.91 (m, 7H, 3 x

CH2, CH); [a]25d = +4.4° (c = 2.5 mg/ml in MeOH); CD (reported in molecular ellipticity, in MeOH) 1^3221 = "6 >906> f ^ 3 235 = +6 .°89>

[01253 = _182> [0)278 = +1»545. 139

R(-)-3,4-Dihydroxy-8-fluoro-l-(3,4,5-trlmethoxybenzyl)-

1,2,3,4-tetrahydroisoquinoline (R(-)-62)

HO

HO OCHg

OCHg OCHg

Procedure same as for S(+)-62 using Enantiomer 2 of 104 (100 mg,

0.17 mmol). Trituration from diethyl ether gave a beige solid, 60 mg

(8 8 %): d.p. 116-120°C; FeCl3 positive; *H NMR (CD3OD) same as S(+)-62

described above; [ a j ^ p = -4.2° (c = 2.5 mg/ml in MeOH); CD (reported

in molecular ellipticity, in MeOH) [^ ]223 = +1>091, 235 =

[0]255 = -273, [«1280 = -818.

Ethyl-2-f 3,4-dibenzyloxy-5-(trifluoromethyl)phenyl]iminoacetate hydro­

chloride (105)

PhCH.O

To a solution of benzylcyanide 93 (2.00 g, 5.03 mmol) in benzene (5 ml) cooled to 0°C was added absolute ethanol (0.44 ml, 7.55 mmol) and excess hydrochloride gas. The solution was stirred at room temperature

for 1 hour, then refrigerated. After 48 hours, the solution was poured

into diethyl ether (100 ml), with stirring a white precipitate formed.

The suspension was refrigerated for 1 hour, the white precipitate fil­

tered off and dried under vacuum to yield 2.21 g (92%):m.p. 123-125°C;

IR (KBr, cm _1) 1652 (C=N); JH NMR (CDCI3 /TMS) <5 7.61 (d, J = 1.4 Hz,

1H, ArH), 7.49-7.29 (m, 10H, ArH), 7.19 (d, J = 1.4 Hz, 1H, ArH), 5.22

(s, 2H, ArCH2 0), 5.08 (s, 2H, ArCH2 0), 4.62 (q, J = 6.9 Hz, 2H,

OCH2CH3), 4.02 (s, 2H, CH2 ), 1.70 (br s, 1H, NH), 1.44 (t, J = 6.9 Hz,

3 H , 0CH2CH3); MS m/z 443 (M+ - HC1), 349, 91 (base); Anal. Calcd. for

C25H25C1F3N03 : C ’ 6 2 -57 ; H > 5.25; N. 2.92. Found: C, 62.25; H, 5.22;

N, 2.97.

2-[3.4-Dibenzyloxy-5-(trifluoromethyl)benzyl1imidazoline hydrochloride

(106)HC1

CF

NH PhCH-0

• HCl

To a solution of iminoacetate 105 (1.10g, 2.29 mmol) in chloroform

(25 ml) was added ethylene diamine (0.34 ml, 5.04 mmol) in chloroform

(10 ml). Upon addition of ethylene diamine, a white precipitate imme­ diately formed and the suspension was allowed to stir at room tempera­ ture under argon overnight. The solution was poured into saturated 141

NaHCOj solution (50 ml), the layers separated and the water layer

extracted with chloroform (2 x 50 ml). The combined organic layers

were washed with water (1 x 100 ml), brine (1 x 100 ml), dried (MgSO^)

and concentrated to a colorless oil. The free base proved to be

extremely difficult to crystallize. The hydrochloride salt was made

using 3 N methanolic HC1 and recrystallized from diethyl ether/hexane

to yield white crystals, 379 mg (33%); m.p. 230-231°C with decomposi­

tion; IR (KBr, cm"1) 3200-3100 (br, NH); 1H NMR (CD3OD) <5 7.51-7.26

(m, 11H, ArH), 7.22 (d, J = 1.6 Hz, 1H, ArH), 5.24 (s, 2H, ArCH2 0),

5.06 (s, 2H, ArCH20), 3.92 (s, 2H, CH2), 3.91 (s, 4H, NCH2 CH2N ) ; MS m/z

440 (M+ - HC1), 91 (base); Anal. Calcd. for C ^ H ^ C l F g N ^ ^ H ^ : C,

60.67; H, 5.30; N, 5.66. Found: C, 60.49; H, 5.13; N, 6.02.

2-f 3,4-Dibenzyloxy-5-(trifluoromethyl)benzy11 imidazoline oxalate

(106)(C02H)q

CF

To a solution of iminoacetate 105 (2.21 g, 4.60 mmol) in chloroform

(50 ml) was added ethylene diamine (0.68 ml, 10.13 mmol) in chloroform

(20 ml). The solution was allowed to stir under argon at room tempera­ ture. After 20 hours, the solution was poured into saturated NaHCO^ 142

solution (100 ml), the layers separated and the water layer extracted

once with chloroform (75 ml). The combined chloroform layers were

washed with water (1 x 150 ml), brine (1 x 150 ml), dried (MgSO^) and

concentrated to an oil. The oil was taken up into dichloromethane (5

ml) and to the solution was added oxalic acid dihydrate (580 mg, 4.60

mmol) dissolved in methanol (5 ml). Upon dropwise addition of diethyl

ether, a white solid crystallized which was collected and dried over­

night under vacuum to afford 1.96 g (80%): m.p. 223-225°C; IR (KBr,

cm"1) 3400-3100 (br, NH): XH NMR (DMSO-Dg) <5 7.65-7.30 (m, 12H, ArH),

5.24 (s, 3H, ArCH20), 5.02 (s, 3H, ArCH20), 3.93 (s, 2H, CH2), 3.81 (s,

4H, NCH2CH2N); MS m/z 440 (M+ - C2 H204 ), 349, 91 (base); Anal. Calcd.

for C2 7 H2 5 F3N206 : C, 61.13; H, 4.75; N, 5.28. Found: C, 60.74; H, 4.72;

N, 5.25.

2 -[3,4-Dihydroxy-5-(trifluoromethyl)benzyl1 Imidazoline half oxalate

(71)1/2(C02H)2

CF

HO

NH HO * H(C02H ) 2

A solution of imidazoline 106'(C02H)2 (800 mg, 1.51 mmol) and 10%

Pd/C (95 mg) in ethanol (50 ml) was hydrogenated at 40 psi overnight.

The solution was filtered through celite and the filtrate was concen- 143

trated under reduced pressure. Recrystallization from methanol/diethyl

ether gave white crystals, 233 mg (50%): m.p. 202°C; IR (KBr, cm-1)

3400-3000 (br, OH and NH); h NMR (CD3OD) 67.04 (br s, 1H, ArH), 6.92

(s, 1H, ArH), 3.87 (s, 4H, NCH2CH2N ) , 3.75 (s, 2H, CH2 ); MS m/z 260 (M+

- CH02), 259, 191 (base); Anal. Calcd. for 1/2

(C02H )2 1/4H20: C, 46.53; H, 4.07; N, 9.04. Found: C, 46.63; H, 4.24;

N, 8.72.

4 15-Dibenzyloxy-2-(trifluoromethvDbenzaldehyde (108)

CF

0

To a solution of 2-iodo-4,5-dibenzyloxybenzaldehyde (2.50 g, 5.63

mmol) in HMPA (50 ml) in a steel tube with ice-bath cooling was added copper powder (7.15 g, 112.54 mmol) and trifluoromethyl iodide (14.81 g, 75.59 mmol). The solution, in the sealed steel tube, was heated to

110°C via an oil bath. After 48 hours, the solution was cooled, quenched with ice water (100 ml) and diluted with ethyl acetate:diethyl ether (1:1 v/v) (100 ml). The resultant emulsion was filtered through celite, the layers separated and the water layer extracted with ethyl acetate:diethyl ether (1:1 v/v) (100 ml). The combined organic layers were washed with water (2 x 200 ml), brine (1 x 100 ml), dried (MgSO^) 144

and concentrated to a brown oil. The oil was passed through a silica

gravity column using 20% ethyl acetate/80% hexane as the solvent (I?£ =

0.55 silica on glass using 20% ethyl acetate/80% hexane as the sol­

vent). Recrystallization from hot hexane afforded white needles, 1.22

g (56%): m.p. 95-96°C; IR (KBr, cm"1) 1692 (C=0); h NMR (CDCI3 /TMS) 3

10.23 (q, J = 2.1 Hz, 1H, CHO), 7.70 (s, 1H, ArH), 7.48-7.32 (m, 11H,

ArH), 5.26 (s, 2H, ArCH20), 5.25 (s, 2H, ArCHjO); 19F NMR (CFCI3 = 0

ppm external reference) <5 -54.25; MS m/z 386 (M+), 295 (M+ - benzyl),

91 (base); Anal. Calcd. for C2 2 H£7F3 0 3 : C, 68.39, H, 4.44. Found: C,

68.55, H, 4.46.

4 ,5-Dibenzyloxy-2-(trifluoromethyl)benzylalcohol (109)

PhCH-0

To a solution of benzaldehyde 108 (1.62 g, 4.19 mmol) in THF:H20

(9:1 v/v) (50 ml) cooled to 0°C was added sodium borohydride (317 mg,

8.38 mmol). The solution was warmed to 45°C for 2.5 hours, cooled to

0°C and quenched with water and NaH2 P0^. The solution was concentrated and taken up into diethyl ether (50 ml) and water (50 ml). The water layer was extracted with diethyl ether ( 2 x 50 ml). The combined eth­ er layers were washed with water (1 x 100 ml), dried (MgSO^) and con- 145

centrated to a white solid. Recrystallization from dichloromethane/

hexane afforded white needles, 1.55 g (95%): m.p. 107-108°C; IR (KBr,

cm-1) 3400-3150 (br, OH), absence of C=0; 1H NMR (CDCI3 /TMS) S

7.47-7.30 (m, 11H, ArH), 7.20 (s, 1H, ArH), 5.23 (s, 2H, ArCH20 ) , 5.16

(s, 2H, ArCH2 0), 4.76 (d, J = 4.9 Hz, 2H, CH2), 1.74 (br t, J = 5.8 Hz,

1H, D2 0 exchangeable, OH); MS m/z 388 (M+ ), 91 (base); Anal. Calcd. for

C22H19F3°3: C » 68-04 ; H » 4 -93 - Found: C, 68.18; H, 4.98.

4 ,5-Dibenzyloxy-2-(trifluoromethyl)benzylbromide (110)

CF

Br

To a solution of benzyl alcohol 109 (1.46 g, 3.76 mmol) in dichloro­ methane (25 ml) cooled to 0°C was added phosphorus tribromide (1.12 g,

4.14 mmol). The solution was stirred at 0°C for 2 hours, diluted with water (25 ml) and the separated water layer was extracted with dichlo­ romethane (2 x 25 ml). The combined organic layers were washed with brine (1 x 75 ml), dried (MgSO^) and concentrated to a white solid.

The solid was taken up into hot hexane and filtered. The filtrate was crystallized from hot hexane to yield white needles, 1.17 g (69%): m.p. 101-103°C; IR (KBr, cm"1) absence of OH; XH NMR (CDCI3 /TMS) S

7.46-7.32 (m, 10H, ArH), 7.16 (s, 1H, ArH), 7.08 (s, 1H, ArH), 5.21 (s, 146

2 H , ArCH2 0), 5.16 (s, 2H, ArCH2 0), 4.55 (s, 2H, ArCH2). MS m/z 452

(M+ , 8 1 Br), 450 (M+ 79Br), 91 (base); Anal. Calcd. for C2 2 H 18BrF302 : C,

58.55; H, 4.02. Found: C, 58.86; H, 4.24.

4,5-Dibenzyloxy-2-(trifluoromethyl)benzylcyanide (111)

PhCH20

PhCH20

To a solution of benzyl bromide 110 (826 mg, 1.83 mmol) in DMF (10

ml) was added sodium cyanide (200 mg, 4.09 mmol) and the solution was

stirred at room temperature for 30 minutes. Water (40 ml) was poured

into the solution and the solution was extracted with dichloromethane

(3 x 50 ml). The combined organic layers were washed with water (2 x

100 ml), brine (1 x 100 ml), dried (MgSO^), and concentrated to a col­

orless oil which solidified upon drying in vacuo (R^ = 0.42 silica on

glass using 20% ethyl acetate/80% hexane as solvent). Recrystallization

from ethyl acetate/hexane gave white needles, 558 mg (77%): m.p.

98-99°C; IR (KBr, cm"1) 2253 (CN); XH NMR (CDCI3 /TMS) <5 7.48-7.34 (m,

10H, ArH), 7.21 (s, 1H, ArH), 7.18 (s, 1H, ArH), 5.24 (s, 2H, ArCH20) ,

5.17 (s, 2H, ArCH20), 3.84 (s, 211, CH2); MS m/z 397 (M+) , 91 (base);

Anal. Calcd. for C23 H 18F 3N02: C, 69.52; H, 4.57; N, 3.53. Found: C,

69.58; H, 4.33; N, 3.17. 147

Ethyl-2-f 3,4-dlbenzyloxy-2-(trlfluoromethyl)phenyl1Imlnoacetate hydro­

chloride (1 1 2 )

CF

PhCH-0 OCH-CH

NH • HC1 PhCH-0

To a solution of benzylnitrile 83 (1.34 g, 3.37 mmol) in benzene (6

ml) cooled to 0°C was added absolute ethanol (0.29 ml, 5.06 mmol) and

excess HC1 gas and stirred for 1 hour at room temperature, then refrig­

erated. After 76 hours, diethyl ether (75 ml) was added and the solu­

tion was refrigerated one additional hour. The resultant white precip­

itate was filtered and dried under vacuum to yield 1.28 g (79%): m.p.

132-133°C; IR (KBr, cm"1) 1652 (C=N); NMR (CDCI3 /TMS) <5 7.47-7.29

(m, 11H, ArH), 7.16 (s, 1H, ArH), 5.16 (s, 2H, ArCH20 ) , 5.08 (s, 2H,

ArCH20 ) , 4.61 (q, J = 7.0 Hz, 2H, OCH2CH3), 4.32 (apparent d, J = 2.0

Hz, 2H, CH2 ), 1.70 (br s, 2H, NH2) , 1.36 (t, J = 7.0 Hz, 3H, OCH2CH3);

MS m/z 443 (M+ - HC1), 91 (base); Anal. Calcd. for C2 5 H25C1F3N03 : C,

62.57; H, 5.25; N, 2.92. Found: C, 62.77; H, 5.36; N, 2.92. 148

2-f3 ,4-Dibenzyloxy-2-(trIfluoromethyl)benzy11Imidazoline oxalate (113)

CF •(C02H) PhCH,Q

HN PhCH-0

To a solution of imidate 112 (429 mg, 0.89 mmol) in chloroform (5

ml) was added ethylene diamine (0.13 ml, 1.97 mmol) in chloroform (2

ml) with ice-bath cooling, The solution was stirred under argon at room

temperature overnight. The solution was poured into saturated NaHCOg

solution (50 ml), the layers were separated and the water layer was

extracted with chloroform (1 x 75 ml). The combined organic layers

were washed with water (2 x 50 ml), brine (1 x 50 ml), dried (MgS04)

and concentrated to a yellow solid. The solid was dissolved in a mini­

mum amount of dichloromethane and to it was added oxalic acid dihydrate

(112 mg, 0.89 mmol) in methanol (5 ml). Crystallization upon dropwise addition of diethyl ether gave white crystals, 343 mg (71%): m.p.

205°C with decomposition; IR (KBr, cm'1) 3200-3100 (br, NH); NMR

(CDgOD) 8 7.51-7.26 (m, 11H, ArH), 7.22 (d, J = 8 . 6 Hz, 1H, ArH), 5.24

(s, 2H, ArCH2 0), 5.05 (s, 2H, ArCH20 ) , 4.04 (br. q, 2H, CI?2), 3.90 (s,

4 H , NCH2CH2N); MS m/z 440 (M+ - C2 H204 ) , 349 (M+ - C 2H204 - benzyl), 91

(base); Anal. Calcd. for C2 7 H25F3N2 06 l/2 H20: C, 60.11; H, 4.86; N,

5.19. Found: C, 60.27; H, 4.92; N, 5.12. 149

2-f 3.4-Dlhydroxy-2-(trlfluoromethyl)benzy11 Imidazoline oxalate (70)

To a Parr bottle containing 10% Pd/C (20 mg) purged with argon was

added imidazoline 113 (200 mg, 0.37 mmol) in methanol (25 ml). The

solution was hydrogenated at 40 psi overnight, filtered through celite

and the filtrate was concentrated under reduced pressure. The residue

was triturated from diethyl ether to afford a white powder, 121 mg.

(89%): m.p. 139-142°C with decomposition; FeClg positive; IR (KBr,

cm"1) 3400-3100 (br, OH), 1610 (C=N); NMR (CD3OD) S 6.96 (d, J =

8.1 Hz, 1H, ArH), 6.71 (d, J = 8.1 Hz, 1H, ArH), 3.93 (br. s, 2H, CH2),

3.87 (s, 4H, NCH2CH2N); MS m/z 260 (M+ - C2 H204) , 218 (base); Anal.

Calcd. for C 13H 13F3N2 06 H20: C, 42.40; H, 4.11; N, 7.61. Found: C,

42.28; H, 4.05; N, 7.91. 150

Ethyl-2-f 4.5-dibenzvloxy-2-(trlfluoromethyl)phenyl1lminoacetate hydro­

chloride (114)

PhCH.O

NH • HC1 PhCH-0 CF

To a solution of benzylcyanide 111 (765 mg, 1.92 mmol) in benzene (5 ml) cooled to 0°C was added absolute ethanol (0.17 ml, 2.89 mmol) and excess hydrochloride gas. The solution was stirred at room temperature for 1 hour and refrigerated. After 48 hours, diethyl ether (50 ml) was added to the solution and it was allowed to stand refrigerated for 1 hour. The resultant white precipitate was collected and dried under vacuum to yield white solid, 647 mg (70%): m.p. 158-159°C; IR (KBr, cm'1) 1647 (C=N); XH NMR (CDCI3 /TMS) <) 7.49-7.28 (m, 11H, ArH), 7.19

(s, 1H, ArH), 5.26 (s, 2H, ArCH20) , 5.16 (s, 2H, ArCH20 ) , 4.53 (q, J =

6.9 Hz, 2H, OCH2CH3), 4.15 (s, 2H, CH2), 1.63 (br s, 2H, NH2), 1.30 (t,

J = 6.9 Hz, 3H, OCH2CH3); MS m/z 443 (M+ - HC1), 91 (base); Anal.

Calcd. for C2 5 H2 5 C1F3N03 : C, 62.57; H, 5.25; N, 2.92. Found: C,

62.87; H, 5.38; N, 2.92. 151

2-f4.5-Dlbenzyloxy-2-(trifluoromethyl)benzy11 Imidazoline oxalate (115)

•(co2h )2 P h C H - O

HN PhCHo0 CF

To a solution of imidate 114 (647 mg, 1.35 mmol) In chloroform (10 ml) was added ethylene diamine (0.20 ml, 2.97 mmol) in chloroform (5 ml). The solution was stirred under argon at room temperature over­ night. The solution was poured into saturated NaHCO^ solution (50 ml), the layers were separated and the water layer was extracted with chlo­ roform (1 x 75 ml). The combined organic layers were washed with water

(1 x 50 ml), brine (1 x 50 ml), dried (MgSO^) and concentrated to a white solid. The solid was dissolved in a minimum amount of dichloro- methane and to it was added oxalic acid dihydrate (170 mg, 1.35 mmol) in methanol (5 ml). Crystallization upon dropwise addition of diethyl ether gave a white solid, 561 mg (78%): m.p. 192-193°C; IR (KBr, cm'1)

3200-3000 (br, NH); XH NMR (CD3OD) <5 7.48-7.29 (m, 11H, ArH), 7.17 (s,

1H, ArH), 5.24 (s, 2H, ArCH2 0 ) , 5.20 (s, 2H, ArCHgO), 4.00 (s, 2H,

CH2 ), 3.86 (s, 4H, NCH2CH2N); MS m/z 440 (M+ - C2 H204 ), 349 (M+ -

C2fl2°4 " benzyl), 91 (base); Anal. Calcd. for C2 7 H2 5 F3N2 0gl/4 H20: C,

60.62; H, 4.80; N, 5.24. Found: C, 60.60; H, 4.77; N, 5.20. 152

2-f 4,5-Dlhydroxy-2-(trifluoromethyl)benzyl1 imidazoline half oxalate

(72) 1/2 (C02H)2

To a Parr bottle containing 10% Pd/C (50 mg) purged with argon was

added imidazoline 115 (525 mg, 0.99 mmol) in methanol (50 ml). The

solution was hydrogenated at 40 psi overnight, filtered through celite

and the filtrate was concentrated under reduced pressure. The residue

was crystallized from methanol/diethyl ether to afford a white powder,

132 mg (43%): d.p. 265°C; FeCl3 positive; IR (KBr, cm'1) 3400-3100 (br,

OH), 1684 (C=N); NMR (CD3OD) <5 7.11 (s, 1H, ArH), 6.89 (br s, 1H,

ArH), 3.95-3.85 (m, 6H, Cll2 and NCH2CH2N ) ; MS m/z 260 (M+ - CH02), 218,

191 (base); Anal. Calcd. for C 11H 11F 3N202 'l/2 (C02H )2 1/2 H20: C,

45.87; H, 4.17; N, 8.91. Found: C, 45.48; H, 3.97; N, 8.61. 153

2-Amino-1-f4 .5-dibenzyloxy-2-(trifluoromethyl)pheny!1ethanol hydrochlo­

ride (116)

PhCH-0 CF

HO • HC1

To benzaldehyde 108 (723 mg, 1.87 mmol) and zinc iodide (29 mg, 0.09

mmol) in a dry flask was added 98% trimethylsilyl cyanide (0.28 ml,

2.06 mmol). The solution was stirred under argon for 4 hours, diluted

with water (10 ml) and diethyl ether (20 ml), and the water layer was

extracted with diethyl ether (2 x 20 ml). The combined ether layers

were washed with water (3 x 30 ml), dried (MgSO^) and concentrated to a

yellow solid. The solid was taken up into dry THF (25 ml) and to the

solution was added 1 M BH3 THF (5.6 ml). The solution was stirred at

reflux overnight, cooled, concentrated, taken up into methanol (25 ml)

and concentrated (this was repeated). The hydrochloride salt was made using 3 N methanolic HC1 and recrystallization from methanol/diethyl ether gave white crystals, 665 mg (78%): m.p. 211-213°C; IR (KBr, cm”*)

3400-3200 (br, OH), 3200-2800 (br, NH); JH NMR (CD3OD) <5 7.50-7.29 (m,

12H, ArH), 5.25 (s, 2H, ArCH20), 5.21-5.18 (m, obscured under solvent peak, 1H, CH), 5.18 (s, 2H, ArCH20 ) , 3.01 (dd, J = 12.8 Hz and 3.2 Hz,

1H, CH2), 2.88 (dd, J = 12.8 Hz and 9.9 Hz, 1H, CH2) ; MS m/z 387 (M+ -

HC1 - CH2N), 91 (base); Anal. Calcd. for C2 3 H2 3ClF3N0 3 l/2 H20: C,

59.68; H, 5.23; N, 3.03. Found: C, 59.59; H, 5.26; N, 3.00 154

2-14,5-Dlhydroxy-2-(trifluoromethyl)pheny11-2-methoxyethylamine hydro­

chloride (117)

HO CF

HO

och3 • HC1

A solution of ethanolamine 116 (200 mg, 0.44 mmol) in an equivolume mixture of methanol:concentrated HC1 (20 ml) was refluxed for 4 hours, cooled and concentrated under reduced pressure. The residue was taken up into methanol (25 ml) and concentrated (this was repeated twice) to a yellow oil. Trituration from diethyl ether provided a hygroscopic beige solid, 90 mg (75%): m.p. 96-98°C; IR (KBr, cm-1) 3600-3000 (br,

OH and NH); *H NMR (CD3OD) 7.08 (s, 1H, ArH), 7.05 (s, 1H, ArH),

4.76-4.73 (m, obscured by solvent peak, 1H, CH), 3.25 (s, 3H, OCH3 ),

3.02-2.98 (m, 2H, CH2 N); MS m/z 251 (M+ - HC1), 221 (M+ - HC1 - OCH3 , base); Anal. Calcd. for C 10H13ClF3N0 3 l/2 HjO: C, 40.49; H, 4.76; N,

4.72. Found: C, 40.34; H, 5.00; N, 4.47. 155

2-Amino-1~f 4.5-dihydroxy-2-(trifluoromethyl)pheny11ethanol hydrochlo­

ride (76)

HO CF

HO

HO •HC1

Ethanolamine 116 (212 mg, 0.47 mmol) and 10% Pd/C (21 mg) in metha­

nol (25 ml) in a Parr bottle was hydrogenated overnight at 40 psi. The

solution was filtered through celite and the filtrate was concentrated.

The resultant residue was recrystallized from methanol/dichloromethane

to yield crystals, 80 mg (62%): m.p. 183-185°C; IR (KBr, cm-1)

3400-3000 (br, OH and NH); XH NMR (CD3OD) <5 7.23 (s, 1H, ArH), 7.04

(s, 1H, ArH), 5.14 (d, J = 8.9 Hz, 1H, CH2), 3.04-2.85 (m, 2H, CH2);

FAB MS 238 (MH+ ), 200 (MH+ - CH2NH, base); Anal. Calcd. for

C9 H 11C1F3N03 H20: C, 37.06; H, 4.49; N, 4.80. Found: C, 37.25; H,

4.33; N, 5.06. 156

4-Hvdroxy-3-methoxy-5-(trlfluoromethyl)benzaldehyde (118)

CF

HO

CH,0

0

To a solution of phenol 88 (6.69 g, 34.82 mmol) in trifluoroacetic acid (80 ml) was added hexamethylenetetramine (9.76 g, 69.64 mmol).

The solution was refluxed for 2.5 hours, cooled and concentrated under reduced pressure. The residue was dissolved into 1 N HC1 solution (100 ml) and extracted with dichloromethane (3 x 100 ml). The combined organic layers were washed with water (2 x 200 ml), brine (1 x 200 ml), dried (MgSO^) and concentrated to a yellow solid. Recrystallization from dichloromethane/petroleum ether gave white crystals, 3.27 g (42%): m.p. 147-149°C; Lit. m.p. 151-152°C; IR (KBr, cm-1) 3400-3000 (br, OH),

1679 (C=0); !H NMR (CDC1-/TMS) S 9.86 (s, 1H, CHO), 7.68 (d, J = 1.6 J in

Hz, 1H, ArH), 7.57 (d, Jffl = 1.6 Hz, 1H, ArH), 6.74 (br s, 1H, OH), 4.03

(s, 3H, 0CH3); MS m/z 220 (M+ , base). 157

3,4-Dlbenzyloxy-5-(trifluoromethyl)benzaldehyde (119)

CF

PhC H jO

H PhCH20

To a solution of phenol 88 (500 mg, 2.27 mmol) in dichloromethane

(15 ml) under argon cooled to 0°C was added 1 M boron tribromide in dichloromethane (13.6 ml). The solution was stirred for 24 hours at room temperature, quenched with methanol (15 ml), concentrated, diluted with methanol (20 ml) and concentrated again. The residue was dis­ solved in methanol (20 ml) and 2 ml wet volume of Dowex 50x8-100 (pre­ viously washed with methanol) and allowed to stir overnight under argon. The Dowex resin was filtered off and the filtrate concentrated to a solid. The solid was taken up into acetone (25 ml) and to it was added 97% benzyl chloride (0.60 ml, 5.11 mmol), potassium carbonate

(706 mg, 5.11 mmol) and potassium iodide (56 mg, 0.34 mmol). The solu­ tion was refluxed for 24 hours, cooled, concentrated, diluted with water (20 ml) and extracted with ethyl acetate (3 x 20 ml). To the combined ethyl acetate extracts was added 1 N HC1 solution ( 60 ml) and the solution was stirred for 3 hours. The layers were separated and the organic layer washed with water (2 x 60 ml), dried (MgSO^) and con­ centrated to a brown oil which solidified on standing. Gravity column chromatography was performed using 10% ethyl acetate/90% hexane as the 158

solvent (R^ = 0.58 silica on glass using 25% ethyl acetate/75% hexane).

Recrystallization from hot hexane afforded white crystals, 91 mg (10%):

m.p. 68-69°C; IR (KBr, cm'1) 1701 (C=0); NMR (CDCI3 /TMS) 6 9.93 (s,

1H, CHO), 7.74 (s, 2H, ArH), 7.49-7.30 (m, 10H, ArH), 5.23 (s, 2H,

ArCH20), 5.22 (s, 2H, ArCH20); MS m/z 386 (M+ ), 295 (M+ - benzyl), 91

(base); Anal. Calcd. for C^H^F-jOg: C, 68.39; H, 4.44. Found: C,

68.36; H, 4.52.

3 ,4-Dihydroxy-2-(trifluoromethyl)benzaldehyde (121)

CF

HO

HO

To a solution of dimethoxy benzaldehyde 78 (300 mg, 1.28 mmol) in dichloromethane (10 ml) in a dry flask under argon cooled to 0°C was

added 1 M boron tribromide in dichloromethane (3.80 ml). The solution was stirred overnight at room temperature, cooled to 0°C and to it was cautiously added water (10 ml). The solution was stirred for 24 hours at room temperature, diluted with ethyl acetate (50 ml) and the layers separated. The organic layer was washed with water (2 x 50 ml), brine

(1 x 50 ml), dried (MgSO^) and concentrated to a solid. Recrystalliza­ tion from ethyl acetate/hexane produced a biege solid, 178 mg (67%): m.p. 164-165°C with decomposition; FeCl^ positive; IR (KBr, cm'1) 159

3300-3100 (br, OH), 1665 (C=0); JH NMR (Acetone-Dg) S 10.19 (octet,

JH-F = 2 7 Hz and JH-H = 0 6 Hz’ 1H» CH0)» 9,32 (br s ’ 0H)» 7 '44 (dd» J

= 8.3 Hz and 0.6 Hz, 1H, ArH), 7.19 (d, J = 8.3 Hz, 1H, ArH); MS m/z

206 (M+), 55 (base); Anal. Calcd. for C8 H5F303 : C, 46.62; H, 2.44.

Found: C, 46.76; H, 2.49.

3,4-Dihydroxy-5-(trifluoromethyl)benzaldehyde (1 2 2 )

CF

HO

HO

0

To a solution of phenol 88 (250 mg, 1.14 mmol) in dichloromethane

(10 ml) in a dry flask under argon cooled to 0°C was added 1 M boron tribromide in dichloromethane (3.40 ml). The solution was stirred overnight at room temperature, cooled to 0°C and to it was cautiously added water (10 ml). The solution was stirred for 24 hours at room temperature, diluted with ethyl acetate and the layers separated. The organic layer was washed with water (2 x 25 ml), brine (1 x 25 ml), dried (MgSO^) and concentrated to a solid. Recrystallization from ethyl acetate/hexane produced a biege solid, 176 mg (75%): m.p.

188-189°C; Lit. m.p. 188-192°C; IR (KBr, cm"1) 3500-3100 (br, OH), 1643

(C=0); lH NMR (Acetone-D6) <5 9.86 (s, 1H, CH0), 7.68 (m, 1H, ArH),

7.58 (d, Jffl = 1.7 Hz, 1H, ArH); MS m/z 206 (M+ ), 101 (base). Part 2 SYNTHESIS OF HYDRAZINIUM ANALOGS OF DOPAMINE AGONISTS AND ANTAGONISTS

- 160 - CHAPTER V INTRODUCTION

5.1 LOCATION. DISTRIBUTION AND FUNCTION OF DOPAMINE

Dopamine is an endogenous catecholamine, along with norepinephrine and epinephrine. The biosynthesis, release and metabolism of dopamine is similar to the other catecholamines and is described in Chapter I.

Although dopamine is found in both central and peripheral tissues, the concentrations of dopamine are much greater in the CNS than in the periphery [163]. Dopaminergic neurons are located in three principal areas of the brain:

1. The nigrostriatal system. This area contains approximately 80% of

the total brain content of dopamine [163]. Here the dopaminergic

cell bodies are located in the substantia nigra and the axons ter­

minate in the putamen and caudate nucleus [164].

2. The mesolimbic system. The dopaminergic cell bodies in this sys­

tem are located in the ventral tegmental area and terminals are in

the limbic forebrain which includes the and the

olfactory tubule [164],

3. The tuberoinfundibular system. The dopaminergic cell bodies are

located in the arcuate nucleus of the hypothalamus and the termi­

nals in the external layers of the median eminence [164].

- 161 - 162

The nigrostriatal system Is Involved In the regulation of the extrapyramidal nervous system (EPS) in the control of fine motor movements. The EPS involves the control of normal posture, muscle tone, and coordinated motion [163]. The mesolimbic system plays a role in emotions and memory [163]. Both Parkinson's disease and have abnormal dopamine functions [165,166], The former is a result of dopamine neuron loss in the midbrain, while the latter is suggested to be a result of hyperdopnminergic activity [165,166].

The tuberoinfundibular system regulates the secretion of from the pituitary gland. This in turn induces lactation in mammals

[163,164]. Dopamine provides an inhibitory effect on this system.

In the periphery, dopamine concentrations are highest in neuronal, tissues, such as heart, kidney, lungs, liver and spleen [167]. The actions of dopamine in the periphery are dose-dependent. At low dose infusions (2 to 5 fi g per kilogram per minute), dopamine has a vasodilatory effect in renal and mesenteric vessels with minimal vasodilation in cerebral and coronary vessels [168]. This vasodilation action is mediated through specific dopaminergic receptors as well as by dopamine stimulation of /^-receptors. The vasodilation is antagonized by dopamine receptor antagonists, confirming a dopamine receptor mediated vasodilation [169]. In doses of 5 to 10 fig per kilogram per minute, dopamine increases myocardial contractility and via /^-receptor stimulation [168]. In large doses of

>20 fig per kilogram per minute, dopamine causes vasoconstriction and increased peripheral vascular resistance through activation of 163

oc-adrenoceptors In arteries and veins in all vascular beds [168].

Dopamine Is used therapeutically In the treatment of shock.

5.2 DOPAMINE RECEPTOR SUBTYPES

Dopamine receptors were first classified into two types, D^ and D£,

by Kebabian and Caine in 1979 [170]. According to this hypothesis,

activation of D^ receptors mediates dopaminergic effects by stimulation of dopamine sensitive adenylate cyclase [170]. Conversely, activation of D2 receptors results in no enhancement or inhibition of cAMP formation via adenylate cyclase inhibition [170]. Activation of D£ receptors has also been found to activate K+ channels and inhibit

OJ. intracellular Ca mobilization [171]. D£ receptor stimulation may indirectly inhibit phosphatidylinositol turnover as well [171].

Dopamine has /lM potency on Dj receptors while possessing nM potency on

D2 receptors [170]. Furthermore, on D^ receptors, (125) is a partial agonist with /iM potency, while on D2 receptors, apomorphine is a full agonist at nM levels [170].

n - ch

HO

HO

125 164

Both Dj and D2 receptors exist in two interconvertible receptor

states which display high and low affinities for dopaminergic agonists

[172,173]. The high affinity state is thought to be a ternary complex composed of the dopaminergic agonist, the receptor and a guanine nucleotide binding protein [172,173]. Binding of GTP or the stable GTP analog 5 1-guanylylimidodiphosphate causes the dissociation of the ternary complex. This results in the conversion of the D2 receptor

from the high affinity agonist state to the low affinity agonist state

[172]. A similar situation is thought to exist for Dj receptors. The functional role of the two different affinity states of the dopamine receptors remains unclear [174].

In the periphery, Goldberg and Kohli [175] have proposed the classification of DA^ and DA2 for dopamine receptors, based on studies on the cardiovascular and renovascular systems. Stimulation of the DAj receptor, located postjunctionally, causes smooth muscle relaxation

[167]. DA2 receptors are located prejunctionally in the sympathetic nerve endings and are responsible for the inhibition of norepinephrine release [167]. Postjunctional DA2 receptors also exist. The D2 and

DA2 receptors have distinct similarities and may be shown to be identical [176,177]. Although pharmacological differences between and DAj receptors have been reported [175], these differences may not be as great as initially reported and they may also be identical

[176,177].

The synthesis of selective agonists and antagonists for the Dj and

D2 receptors provided further evidence for the existence of two 165

dopamine receptors subtypes [178]. The most widely used selective Dj

agonist is SKF-38393 (126) (Figure 34) [179]. The receptor agonist

activity of SKF-38393 (126) resides in the 1R enantiomer [178].

Another potent selective agonist is (127) which is being

evaluated in clinical trials [180]. The 7-chloro-N-methylated analog

of SKF-38393 (126), SCH-23390 (128), is a potent and selective D:

receptor antagonist [178].

Dj Agonists:

Cl

HO H-CHj

HO

HO 126 127

Dx Antagonist:

C

h -c h 3

HO

128

Figure 34: Selective receptor agonists and antagonists 166

Selective D2 receptor agonist and antagonists are derived from

several diverse chemical classes (Figure 35). The 2-aminotetralin

derivative PPHT (N-0434, 129), the partial LY171555 (130) and

the oxaergoline derivative (+)PHN0 (131) are all selective D2 receptors

agonists [178]. The benzamide sulpuride (132) was one of the first

selective D2 receptor antagonists. An analog which is more potent than

as a D2 antagonist includes YM-09151-2 (133) [178].

Several investigators have suggested the presence of subtypes of both Dj (DAj) and D2 (DA2) dopamine receptors [181,182]. This is based on variations in pharmacological responses which are difficult to explain with only two types of dopamine receptors. Although the presence of these hypothesized subtypes has not been conclusively established, cloning of the Dj and D2 receptors may prove their existence. The D j /D2 dopamine receptor classification has been challenged by the recent cloning and characterization of the Dg dopamine receptor [183]. In the past two years and in recent months, the intensity of work on the cloning of dopamine receptors has increased dramatically.

In 1988, a rat D2 dopamine receptor cDNA was isolated using the hamster /?2-adrenergic receptor gene as a hybridization probe [184].

The rat dopamine receptor cDNA has been cloned and expressed as a 415 amino acid protein [184]. The deduced amino acid sequence of the rat

D2 receptor shows that it is a G-protein coupled receptor with seven stretches of hydrophobic residues which could represent seven transmembrane spanning regions [184]. Furthermore, the large third 167

D 2 Agonists:

OH

KN.

N

130

129

HO,

132

D 2 Antagonists:

OCH. OCH.

NH NH

CH.NH

SO-NH

133

Figure 35: Selective D2 agonists and antagonists cytoplasmic loop and the relatively short carboxy terminus of the rat

D2 receptor suggests that it is coupled to (similar to the a 2 _adrenoceptor as discussed in Part I of this document) [184]. 168

A human pituitary D 2 receptor has also been recently cloned [185].

The amino acid sequence was found to be 96% identical with the cloned

rat D2 receptor [184,185]. Similarly, the human D2 receptor contains

seven putative transmembrane regions [185]. The major difference

between the rat and the human D2 receptor is that the human D2 receptor

contains an additional 29 amino acid sequence in the putative third

cytoplasmic loop [185]. Grandy et al. [185] have suggested two

possible hypotheses for this difference:

1. The human gene contains the extra 87 bases which encode for the

additional 29 amino acid sequence and the rat gene does not.

2. Both human and rat genes possess two distinct genes which encode

for two distinct D2 dopamine receptors.

In defense of the first hypothesis, Southern blot analysis by Grandy et

al. [185] suggested that only one human D2 dopamine receptor gene

exists.

It is of interest to note that the coding sequence of the human D2

receptor gene is interrupted by introns [185]. Except for , this

is different than most G-protein coupled receptors which are

intronless. As a result of this intron-exon sequence, using the rat D 2

dopamine receptor gene, two groups [186,187] independently have shown

that the gene produces two D2 receptor isoforms via alternative

messenger RNA splicing. One isoform corresponds to the 415 amino acid

D2 receptor [184] and the other isoform contains an additional 29 amino

acid sequence [185]. Monsma et al. [187] have suggested the nomenclature D2L and D2 g for the "longer" and "shorter" D 2 receptor 169

isoforms, respectively. The significance of the two isoforms is

unclear.

By transfecting a mouse fibroblast cell line with rat genomic DNA,

Todd et al. [188] have established a cell line which expresses D2

receptors from a gene different than the rat D2 receptor gene initially described by Bunzow et al. [184]. And although it has the pharmacological profile of a D2 receptor, this novel D2 receptor was 2+ shown to be coupled to external Ca flux resulting in an increase in intracellular Ca^+ [188],

Using a probe derived from the rat D2 receptor sequence [184],

Sokoloff et al. [183] isolated a novel clone from a rat brain cDNA library termed "Dg" [183]. The Dg receptor gene, like the D2 receptor gene, contains introns although alternative mRNA splicing to form different isoforms does not appear to occur [183]. The Dg receptor is a 446 amino acid protein. Although it contains seven putative transmembrane regions characteristic of G-protein coupled receptors, Dg receptors expressed in Chinese hamster ovary (CHO) cells do not affect cAMP formation [183].

Pharmacological differences exist among the D2 and Dg receptors.

Dopamine is almost 20 times more potent at Dg receptors than at D2 receptors [183]. Although agonists such as apomorphine (125) were equipotent on both Dg amd D2 receptors, putative selective agonists such as (130) were 100 times more potent at Dg than at D2 receptors [183]. Butyrophenone and type nonselective Dj/D2 antagonists were 10 to 20-fold more active at D2 170

than at Dg receptors, whereas benzamide antagonists such as sulpiride

(132) were only 2 to 3 times more potent at T>2 than at Dg receptors

[183]. The Dg receptor Is mainly expressed In the limbic system and It

Is found as both an autoreceptor and a postsynaptlc receptor [183].

Interestingly, the pituitary, which is the prototypic location of D 2

receptors, appears to be completely devoid of Dg receptors [183].

Host recently, the human Dj dopamine receptor has been cloned and

expressed simultaneously by three independent groups [189,190,191].

Unlike the D2 and Dg receptors, the coding region of the Dj receptor

gene is intronless and encodes a 446 amino acid protein [189,190,191].

The Dj receptor also has seven hydrophobic stretches which may

represent seven transmembrane regions typical of G-protein coupled receptors [189,190,191]. Furthermore, the cloned Dj receptor is positively coupled to adenylate cyclase [189,190,191]. Sunahara et al.

[191] have shown that the intronless human Dj receptor gene is located on chromosome 5. Data from both Northern and Southern blot analyses suggest the existence of more than one Dj subtype [189,190,191]. CHAPTER VI STATEMENT OF PROBLEMS AND OBJECTIVES

A number of hypothetical models of the dopamine receptor binding

site have been proposed on the basis of structure activity

relationships of dopamine agonists and antagonists. These models have

been extensively reviewed in recent years [163,192-195]. At

physiological pH, dopamine, and related amine containing agonists and

antagonists, exist in an equilibrium between the uncharged amine form

and the charged ammonium ion form. One area of intense research

concerns whether the amine moiety of dopamine agonists and antagonists

act at Dj and'D2 receptors in the charged or uncharged state.

In the past, several investigators have suggested that the amine

moiety of dopamine agonists and antagonists act in the basic or

uncharged state [196,197]. Others have proposed that the orientation

of the unshared electron pair on nitrogen is critical for optimum

receptor interaction [198,199]. The synthesis and biological

evaluation of permanently charged and uncharged dopamine agonists and antagonists has provided strong evidence that dopamine and related agonists and antagonists interact with dopaminergic receptors in the charged state (Figure 36) [200].

- 171 - Se(CH3)2 I*

137 136 + N(CH3)3 I-

138 HO,

HO 139

Figure 36: Permanently charged and uncharged analogs of dopamine ago­ nists

The permanently charged 'dimethylsulfonium analog of dopamine, 134

[201], was shown to have direct dopaminergic activity on D2 receptors

[202,203]. This proved that the presence of a nitrogen atom is not

required for stimulation of dopamine receptors. Conversely, the

permanently uncharged methylsulfide analog of dopamine, 135, was

inactive at D2 receptors [204]. The permanently charged

dimethylselenonium and trimethylammonium analogs of dopamine (136 and

138, respectively), as well as the quaternary salt of apomorphine

(139), were all found to exert D2 receptor agonist activity [202-205].

However, the permanently uncharged methylselenide derivative 137 could not activate D2 receptors [204]. In comparison to dopamine (DA) and 173

dimethyl dopamine (DMDA), the rank order potency of the above analogs

on D2 receptors follows the trend DMDA > DA > -N(Me)g+I , -S(Me)2+I ,

-Se(Me)2+I > » -SMe, -SeMe [202,203]. These results indicate that a

postively charged functionality at the amine position of dopamine and

related agonist compounds is necessary for dopaminergic receptor

activation.

A recent investigation of N-substituted norapomorphines further

supports this hypothesis [206]. N-Ethyl and N-propyl norapomorphine

derivatives (140 and 141, respectively) were found to have potent D2

receptor agonist activity (Figure 37) [206]. Conversely, the

N-trifluoroethyl (142) and N-pentafluoropropyl (143) derivatives were

found to be completely devoid of dopaminergic receptor agonist activity

[206]. It was hypothesized that the highly electronegative fluorine

atoms decrease the electron density on the nitrogen atom rendering it

less basic. Thus the electron poor nitrogen atom is less able to be

cationic at physiological pH and unable to bind to dopamine receptors

in the proposed requisite charged form [206].

The biological evaluation of permanently charged and uncharged dopamine antagonists showed similar results (Figure 38) [207,208]. The dimethylsulfonium (145) and the trimethylammonium (146) derivatives of chloropromazine (144), a potent , both had D2 antagonist activity [207]. Similarly, permanently charged pyrrolidinium (147) and tetrahydrothiophenium (148) analogues of sulpuride (132) were also found to have D 2 receptor antagonist activity

[208]. Conversely, the permanently uncharged sulpuride analog 149 was 174

HO. HO,

HO' HO 140 141

HO HO,

HO HO 142 143

Figure 37: N-Substituted norapomorphines

inactive as a D 2 receptor antagonist [208]. Thus, the rank order potency trend for the above dopamine antagonists is -NI?2H+ > "NRg+ =

-SR2+ » > SR [208].

There are several possible reasons why the permanently charged analogs of dopamine agonists and antagonists are less active and have lower affinity for D2 receptors than the parent compounds which possess primary, secondary and tertiary amines. One explanation may be that the presence of the permanent positive charge decreases the lipophilicity of the compound to an extent that impairs its ability to partition into the lipophilic membrane near the receptor [200]. Thus, the obtainable effective concentration of the permanently charged 175

Figure 38: Permanently charged and uncharged dopamine antagonists analogs at the ligand binding site on the receptor is significantly

less than that of the parent amine compound [200].

A second explanation may involve a difference in the ligand binding interaction of the permanently charged analogs versus the parent amine

(Figure 39) [200], In the charged ammonium form, the parent amines possess a proton available for hydrogen bonding as well as a positive charge. The cationic ammonium ligand may then interact with an anionic site (such as a carboxylate or phosphate anion) on the receptor via a reinforced ionic bond [200]. Conversely, the permanently charged analogs do not possess a proton at the charged site. Thus, the interaction of the permanently charged species with an anion on the 176

receptor could only be through ionic bonding [200], Reinforced ionic bonds have a bond strength of 10 Kca1/mole whereas the bond strength of an ionic bond is 5 Kcal/mole [209]. Thus, the higher affinity and activity of the parent amines versus the permanently charged analogs may be attributed to the greater bond strength of the reinforced ionic bonding which inhibits the dissociation from the receptor [200].

R

(Drug) i+ w — R (Drug H+ R (Drug

H Solnforced Ionic Bond

R*c«ptor«

Ionic Bond

Figure 39: Interaction of dopaminergic amines and permanently charged analogs with the D£ receptor. (Modified from [200])

In support of the second hypothesis, Williamson and Strange [210] have recently provided evidence for the role of a carboxyl group in ligand binding to the D2 receptor. Treatment of D 2 dopamine receptors in bovine caudate nucleus membrane with 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB, a free sulfhydryl modifying agent), 1,2-cyclohexanedione

(an arginine residue modifier) or N-acetylimidazole (modifies tyrosyl residues and may affect free amines) had no effect on [^H]

(150) binding [210]. 177

NH

150

However, treatment of bovine D2 receptors . with

N,N'-dicyclohexylcarbodiimide (DCC, modifies predominantly carboxyl O groups) caused a dramatic decrease in [ H]spiperone (150) binding

[210]. Furthermore, a variety of D£ dopamine receptor antagonists

protected against DCC receptor modification suggesting that the

carboxyl group is at the ligand binding site [210]. Effects of pH on O [ H]spiperone binding to D2 receptors show that binding decreases from

pH 7 to 3 with a 50% decrease in binding at pH 5.2 [210]. The data

suggests that a single ionizing group, which has a pKa in the range

expected for a carboxyl group, is critical for ligand binding

[209,210].

It is currently believed that the ligand binding site of the

G-protein coupled receptors (which includes the D 2 receptor) is within

the hydrophobic core formed by the seven transmembrane helices. The

sequence of the rat D 2 receptor [184], as well as the sequence of the human T>2 receptor [185,187], all include the presence of two residues in the second and third helix (Asp®® and Asp^^, respectively) and a residue in the second helix (Glu®®).

The carboxyl groups of these residues could in theory be modified by 178

DCC which may result in the inhibition of ligand binding [210]. It is

of interest to note that an aspartyl residue corresponding to Asp**^ is

conserved in all G-protein linked receptors which bond cationic amines

[57].

With this in mind, the objective of this investigation was to examine the hypothesis which proposes that dopamine agonist and antagonist amines interact with an anionic group (perhaps Asp*^) on the D2 receptor via a reinforced ionic bond. The target compounds for this study were the chloropromazine hydrazinium analogs 151 and 152 and the dopamine hydrazinium analogs 153 and 154 (Figure 40). These hydrazinium analogs have the characteristic of being permanently charged while also possessing a group which can act as a hydrogen bond donor (Figure 41). Thus, the idea was that the target compounds may be able to enhance the activity of the permanently charged compounds by virtue of providing an additional hydrogen bond at the appropriate site on the D2 receptor. To examine the optimum position of the proposed hydrogen bond, the targets represent compounds which contain hydrogen bonding groups on either side of the permanently charged functionality. 179

Figure 40: Target hydrazinium dopamine agonist and antagonist analogs

CH, CH, / (Drug) (Drug) •N- ■N CH, I I \ H CH, CH,

Receptors Receptoi

Figure 41: Interaction of hydrazinium functionalities with a carboxy- late ion CHAPTER VII

RESULTS AND DISCUSSION

7.1 CHEMISTRY

The chloropromazine hydrazinium analog 151 had been previously prepared, as reported in the patent literature by Rudner and Prapas

[211], with a variety of counterions. Their method of synthesis involved the treatment of the appropriate tertiary amine with either gaseous chloramine or hydroxylamine O-sulfonic acid [211]. In both of these methods, a large excess of the tertiary amine must be used. For this investigation, hydrazinium 151 was prepared using the improved

N-amination method reported by Tamura et al. [212,213], which involves the treatment of tertiary amines with O-mesitylenesulfonylhydroxylamine

(MSH, 155).

CH

H , N — 0 — S CH

CH 3 155

MSH was prepared by the procedure of Tamura et al. [213] which is an improved method of the original synthesis by Carpino [214]. This involved the treatment of ethyl N-hydroxyacetimide and triethylamine with mesitylene chloride, followed by hydrolysis of the isolated

- 180 - 181 hydroxamate with 70% perchloric acid. MSH was isolated as a white amorphous solid, the product confirmed by *H NMR spectroscopy, and was used unpurified. MSH rapidly decomposes at room temperature to a viscous brown gum and thus must be stored at 0°C. The synthesis of chloropromazine hydrazinium 151 involved the treatment of a solution of the free base of chloropromazine (144) in dichloromethane cooled to 0° with 1.1 eq. of MSH for 5 minutes (Scheme XXI). The solution was concentrated and recrystallization from dichloromethane/ether gave the desired hydrazinium 151 as the mesitylenesulfonate salt.

Scheme XXI

Synthesis of chloropromazine hydrazinium analog 151

MSH

c h 2c i 2 cl

CH.

CH, 144

Retrosynthetic analysis of chloropromazine hydrazinium 152 (Scheme

XXII) led to the N,N-dimethylhydrazone derivative, the synthesis of which had been reported in the literature by Corral et al. [215].

Following this procedure, chlorophenothiazine 156 was deprotonated with sodium amide in xylenes, followed by alkylation with

2-chloroacetaldeliyde diethyl acetal to afford the desired substituted chlorophenothiazine 157 (Scheme XXIII). Corral et al. [215] purified 182 chlorophenothiazine 157 via distillation at 188°C (0.2 mm) to afford the desired compound as a crystalline solid. Due to equipment

limitations in our laboratory, diethyl acetal 157 was purified via flash silica column chromatography using 5% ethyl acetate/95% hexane as the solvent. A light yellow oil was obtained which was resistant to crystallization; however, NMR and mass spectra were consistent with pure diethyl acetal 157■ Interestingly, the above reaction conditions were critical as treatment of 2-chlorophenothiazine (156) with sodium hydride in THF/DMSO (or toluene/DMF) followed by addition of bromoacetaldehyde dimethyl acetal resulted in starting material only. .

Scheme XXII

Retrosynthesis of 152

Refluxing diethyl acetal 157 in dioxanetl^O (5:1 v/v) with 1.5 eq. oxalic acid for 7 hours, Corral et al. [215] obtained acetaldehyde 158 in 60% yield recrystallized from benzene/hexane. This yield was 183

Scheme XXIII

Synthesis of N,N-dimethylhydrazone 159

1) NaNH xylenes

Dloxane:H20 5:2 v/v H„NNMe, ► (COjH) 2 • H 20 abs. EtOH

reflux

159

improved by refluxing diethyl acetal 157 in dioxaneil^O (5:2 v/v) with

1.5 eq oxalic acid overnight. Recrystallization from diethyl ether gave the desired acetaldehyde 158 in 74% yield as white needles.

Acetaldehyde 158 was somewhat light sensitive as the white needles turned a purple-gray color when exposed to light over several weeks. A quantitative yield of dimethylhydrazone 159 was obtained by refluxing acetaldehyde 158 in absolute ethanol with N,N-dimethylhydrazine for 3 hours.

The formation of the trimethylhydrazonium derivative 160 was very sensitive to solvent effects (Scheme XXIV). Treating dimethylhydrazone

159 with an excess of iodomethane in absolute ethanol at room temperature for 48 hours resulted in a reddish-brown solution which was 184

a mixture of compounds containing a minimal amount of product as determined by the NMR spectrum. Heating an ethanolic solution of dimethylhydrazone 159 and an excess of iodomethane to reflux for 4.5 hours again resulted in a mixture. After treating with charcoal; however, the desired hydrazonium salt 160 was isolated in only 7% yield. The low yield was perhaps the result of decomposition during

isolation attempts. Diethyl ether was next used as the solvent. The rationale was to form the hydrazonium salt which would be insoluble in ether and would precipitate from the solution. The desired product could then be collected with minimal manipulations. After 72 hours; however, only starting material was present. Refluxing hydrazone 159 in an excess of iodomethane neat [216] for 6 hours again gave 7% yield of the desired hydrazonium 160. A successful synthesis of hydrazonium

160 was finally achieved by stirring a solution of hydrazone 159 and an excess of iodomethane in acetone overnight at room temperature.

Recrystallization from acetone/ether gave hydrazinium 160 as a white powder in 69% yield. In CDCl^/TMS, the methinyl proton of hydrazone

159 appeared as a triplet at 6.51 ppm (Figure 42). Interestingly, quaternization to hydrazonium 160, results in a downfield shift of the methinyl proton triplet to 8.54 ppm as shown in Figure 43. 185

Scheme XXIV

Synthesis of N,NtN-trlmethylhyrirazonlum iodide 160

CHj I tbs. EtOH ------► N. R. rn. temp. 48 hours

159 CHj

CHj I

abs. EtOH N 4.5 hours reflux

160 (Y: 7%) CHj I

Et20 ft N. R. (Starting Material) ra. temp. 72 hours

CH, I M Cl reflux 6 hours

CH, 160 (Y: 7%)

CH, I Acetone Cl rn. tenp. 48 hours

160 (Y: 69%) iue42: oto f h NR pcrm f 5 I CDClg/TMS In L59 of spectrum NMR the of Portion : 2 4 Figure

17053 'C t V v *’ 186 187

• 7 V K -

160

Figure A 3 : Portion of the NMR spectrum of 160 in CD^OD

The literature concerning reduction of hydrazonium salts is sparse.

One problem is that in addition to reduction of the C=N bond, cleavage of the N-N bond may also occur [217]. Reduction of hydrazonium 160 was

initially attempted using the method of Siddiqui et al. [218] in which hydrazones are catalytically reduced to substituted hydrazines using palladium on carbon (Pd/C) in solvents such as dioxane or benzene.

Hydrogenation of a solution of hydrazonium 160 and 5% Pd/C (20% w/w) in dioxane at 50 psi for 6 hours, resulted in a complex mixture. 188

Hydrazones have also been reduced using dlborane In diglyme followed

by saturating the solution with hydrogen chloride gas [219]. One

problem with this method is the formation of Inseparable mixtures of

mono- and dichloride salts [219]. In the case of using this method for

the reduction of hydrazonium 160, the added problem would be the

formation of a mixture of iodide and chloride salts. Although ion

exchange chromatography would result in a homogeneous counterion

content, in the process of eluting the compound off the column, the problem of a mixture of mono- and di- salts would remain.

The use of lithium aluminum hydride for the reduction of

N,N-dimethylhydrazone 159 was reported by Corral et al. [215], This method was chosen for the reduction of hydrazonium 160, although it had been reported that quaternary hydrazonium salts, when reduced with aluminum hydrides, may undergo subsequent ring closure to the aziridine

[220]. Thus, following the procedure of Corral et al., hydrazonium 160 was refluxed with 0.6 eq. of lithium aluminum hydride in ether for one hour (Scheme XXV). The NMR spectrum showed the starting material as the major component and only a very small amount of the desired product. The NMR spectrum of the yellow oil obtained using a 3-fold excess of lithium aluminum hydride showed the complete absence of starting material. However, along with the desired product, several impurities in the aromatic region were present. Titrating the amount of lithium aluminum hydride from 1 eq., which resulted in an approximately equal mixture of starting material and desired product, to 2 eq. resulted in complete reduction to the desired product. 189

Recrystallization from methanol/diethyl ether gave the target compound hydrazinium 152 as white crystals in 48% yield. Interestingly, whereas the methinyl proton of hydrazonium 160 appeared as a triplet at 8.54 ppm in the NMR spectrum (Figure 43 on page 187), the corresponding methylene protons adjacent to the hydrazinium functionality in 152 appeared as a triplet extremely upfield as shown in Figure 44.

Scheme XXV

Synthesis of Hydrazinium Derivative 152

0.6 eq. L1A1H4 Starting Material Et20 (very small amount of reflux, 1 hour desired product)

160

3 eq. L1A1H4 ^ + impurity

Et jO n h .n +,c h 3 reflux, 1 hour 7 N c h CH j 3 152

1 eq. LiAlH4 Mixture of starting material and desired product Et20

reflux, 1 hour

2 eq. LiAlH4 “N Cl

EtjO NH +.CH, 3 reflux, 1 hour 7 CH CHj 3

152 (Y: 48%) CH, f XI CH,I> 152

jjulwL X

■ i‘-r" > T I I I I | I I I I | I I I I |- l'~ l T *~ I | 1 T I 1 | 7 .0 E. 6.0 5.5 >.0 4 .5

Figure 44: *H NMR spectrum of 152 in CDgOD 191

The synthesis of dopamine hydrazinium analogs 153 and 154 has not been achieved thus far. The accomplishments towards the synthesis of

these compounds to date are as follows. The approach to the synthesis of analog 153 Involved N-amination of the dlmethoxy protected dimethyldopamine (162), which was prepared by the method of Glnos et al. [221]. It is of interest to note that attempts to prepare

3.4-dimethoxy protected dimethyldopamine by the reductive methylation method of Borgman et al. [222] was not successful. This method involved hydrogenation of a solution of commercially available

3 .4-dimethoxy-/?-phenethylamine (161), 10% Pd/C (30% w/w) and formaldehyde in methanol (Scheme XXVI). Although Borgman et al. [222] reported a 75% yield of the desired dlmethoxy dimethyldopamine (162), this investigator repeatedly isolated only tetrahydroisoquinoline derivative (163). Attempts to N-aminate dimethoxy dimethyldopamine

(162) with MSH or hydroxylamine-O-sulfonic acid and sodium methoxide were repeatedly unsuccessful. The reason for this is unclear and is still being investigated.

The approach to the synthesis of dopamine hydrazinium analog 154 involved the same methodology used to obtain the corresponding chloropromazine hydrazinium analog 152. Thus, commercially available

3.4-dibenzyloxybenzaldehyde (164) was refluxed with

N,N-dimethylhydrazine and a catalytic amount of p-toluenesu]fonic acid in absolute ethanol to afford the N,N-dimethylhydrazone 165 (Scheme

XXVII). The hydrazonium 166 was easily prepared by treating dimethylhydrazone 165 with an excess of iodomethane in absolute ethanol at room temperature. 192

Scheme XXVI

Attempted Synthesis of Dlmethoxy dimethyldopamine (162) via Reductive Methylatlon

MeOH

163

Scheme XXVII

Synthesis of Dlmethylhydrazonlum 166

H-NNMe

cat. pTsOH PhCH-0 abs. EtOH

abs. EtOH PhCH.O

Reduction of the hydrazonium and deprotection of the dibenzyloxy groups was attempted in one step using catalytic hydrogenation (Scheme 193

XXVIII). Although Siddiqui et al. [218] reported that the solvents benzene, dioxane, cyclohexane, and toluene favored the formation of

substituted hydrazines, methanol was initially used as the solvent due to solubility reasons. When hydrazonium 166 was hydrogenated in methanol with 10% Pd/C (10% w/w) for 5 hours, no reaction had occurred.

Under the assumption that the catalyst was being poisoned, the concentration of 10% Pd/C was increased to 30% w/w and the solution was hydrogenated for 7 hours. The NMR spectrum showed at least four products. Benzyloxy peaks as well as methinyl peaks appeared in the spectrum, although the solution appeared to be FeClg postive. In the event that the mixture consisted of partially deprotected and reduced and nonreduced hydrazonium species, a solution of hydrazonium 166, 10%

Pd/C (30% w/w) in methanol was hydrogenated for 24 hours. The NMR spectrum showed the absence of benzyloxy groups and the absence of the hydrazonium methinyl peak; however, the spectrum was not consistent with the desired product. Attempts to purify the oil to identify the product resulted in a brown residue. Because methanol was used as the solvent, the product may have been a result of reductive cleavage of the N-N bond [218]. Changing the solvent to benzene (which formed a suspension) and again hydrogenating with 10% Pd/C (30% w/w) overnight resulted in recovery of starting material only. It appeared that this would not be a synthetically useful route.

The alternate approach was to reduce the hydrazonium to the hydrazinium in one step and then removal of the dibenzyloxy groups in a separate step. Hydrazonium 166 was reduced by refluxing with 3 eq. of 194

Scheme XXVIII

Attempts to the Synthesis of HydrazI n i u m 154 via Hydrogenation

CHj H 2, SO psi PhCH-0 i r a' 10% Pd/C (10% w/w) 1 Starting material MeOH, 5 hours PhCH-0 XT'- 166 H 2, 50 psl

10% Pd/C (30% w/w) 1 Mixture of ~ 4 different MeOH, 7 hours compounds

H 2, 50 psl

10% Pd/C (30% w/w) ■ ► ??/Not desired product MeOH, 24 hours

H 2, 50 psl

10% Pd/C (30% w/w) ------► Starting material Benzene, overnight

lithium aluminum hydride in ether for 1 hour to give hydrazinium 167

(Scheme XXIX). The product was confirmed by NMR and FAB mass spectra. Elemental analysis of hydrazinium 167 did not agree with the calculated values. The elemental analysis suggested that hydrazinium

167 contained between one and two molar equivalents of iodide counterion; therefore, analysis for iodine content is currently being performed.

Hydrazinium 167 was carried on to the last deprotection step. As described above, removal of the dibenzyloxy groups via catalytic hydrogenation was not a useful method for this compound due to possible 195

Scheme XXIX

Synthesis of Hydrazinium 167 and Attempted Deprotection

CHj CH l.i^.CH, 1 iv-CH, P h C H , 0i _ ^ ^ J&T -N 3 _ PhCH-0 w - vv u+^ y y v \ H) • * “ *■■► Y Y ^

PhCH20 * reflux, 1 hour PhCH20 lii 167

CHj

PhCH,0. ^ ^ J r ™ 3 4 eq. TMSI NH \ ”■■■'■ ► ??/Not desired product l | CH3 CHjCN

PhCH-02 I' 50*C- 2 hours 167

Pd/C - ► Starting material cyclohexene, EtOH reflux, 48 hours

poisoning of the catalyst and possible reduction of the N-N bond.

Deprotection using refluxing HC1 In methanol was also not useful due to

formation of a mixture of counterions as well as the formation of a

mixture of mono- and di-salts. The use of trimethylsily1 iodide (TMSI) has been reported to be an efficient and mild method of removal of benzyl ethers [223]. Treatment of hydrazinium 167 with 4 eq. of TMSI

in acetonitrile at 50°C for 2 hours resulted in a brownish-orange colored solid. The NMR spectrum showed the absence of benzyloxy groups; however, the solid was not FeCl3 positive. The catechol was not reprotected as trimethylsily1 ethers as there were no 196 trimethylsilyl peaks In the NMR spectum. The aromatic region was also not consistent with the desired product. Catalytic transfer hydrogenation [224] was used next in the attempt to remove the dibenzyloxy groups. A solution of hydrazinium 167, 10% Pd/C (1:1 w/w), and 25 eq. of cyclohexene in absolute ethanol was heated to reflux

[225]. Typically, deprotection by this method occurs rapidly in about

1 hour. After 48 hours, still only starting material was observed in the NMR spectrum. Conditions for catalytic transfer hydrogenation which may be successful for this system is the use of

1,4-cyclohexadiene as the hydrogen donor, and/or palladium black as a more active catalyst [2 2 1 ].

Another approach to the synthesis of hydrazinium analog 154 was simultaneously pursued. Catalytic hydrogenation using platinum oxide has been reported to reduce hydrazones to the corresponding hydrazines in good yields [226]. However, platinum oxide is not a useful catalyst for the removal of benzyloxy protecting groups. Therefore, the free catechol hydrazonium iodide 170 was prepared as shown in Scheme XXX.

Hydrogenation of alcoholic solutions containing hydrazinium 170, platinum oxide (10% w/w to 20% w/w) for 3 hours to overnight at 50 psi was unsuccessful. 197

Scheme XXX

Synthesis of Hydrazonium 170 and Attempts to Reduce with Platinum Oxide

H H.NNMe

abs. EtOH Et.O

?H3 h2 ,n+'CH3 HO Pt02 (10% w/w) CH, Starting material EtOH, 3 hours HO XT'

170 «2 Pt02 (30% w/w) 1 Starting material MeOH, overnight

7.2 BIOLOGY

The biological evaluation of chloropromazine hydrazinium analogs 151 and 152 is currently in progress. Hydrazonium derivatives 160 and 170 will also be tested for biological activity.

7.3 SUMMARY

1. The synthesis of chloropromazine hydrazinium analog 151 using

O-mesitylenesulfonylhydroxylamine (MSH) was described.

2. The synthesis of hydrazinium derivative 152 was described.

3. Approaches to the synthesis of dopamine hydrazinium analogs 153

and 154 have been described. CHAPTER VIII EXPERIMENTAL

General information concerning instrumentation, solvent preparation and elemental analysis may be found in the introduction of Chapter 4.

Additionally, dry xylenes were obtained by refluxing overnight with and distillation from calcium hydride.

1-[3-(2-Chlorophenothiazin-10-yl)propyl 1 -1,1-dimethylhydrazinium mesi- tylenesulfonate (151)

2 3 'OMes

To a solution of free base Chlorpromazine (Sigma) (730 mg, 2.29 mmol) in dichloromethane ( 5 ml) cooled to °C was added

O-mesitylenesulfonylhydroxylamine [213,214] (543 mg, 2.52 mmol). The solution was stirred at room temperature for 5 minutes and concentrat­ ed. Recrystallization from dichloromethane/diethyl ethe,r gave a white solid, 812 mg (66%): m.p. 161-165°C; IR (KBr, cm"1) 3270 (NH); h NMR

- 198 - 199

(CD3OD) S 7.27-6.95 (m, 711, ArH of phenothlazine), 6.86 (s, 2H, ArH of

mesitylenesulfonate), 4.04 (t, J = 6.4 Hz, 2H, CH2 ), 3.58-3.52 (m, 2H,

CH2 ), 3.18 (s, 6 H, NMe2 ), 2.62 (s, 6H, CH^ x 2 of mesitylenesulfonate),

2.31-2.25 (m, 2H, CH2 ), 2.22 (s, 3H, CH3 of mesitylenesulfonate); FAB

MS 336 (M+ - mesitylenesulfonate, 3 ^C1), 334 (M+ - mesitylenesulfo­

nate, 3 5 C1, base); Anal. Calcd. for C26^32^^3®3^2: C, 58.47; H, 6.04;

N, 7.87. Found: C, 58.33; H, 6.08; N, 7.69.

2-Chlorophenothiazln-10-ylacetaldehyde (158)

Cl

V0

To a suspension of 2-chlorophenothiazine (Aldrich) (lO.OOg, 42.80 mmol) in dry xylenes (100 ml) was added 95% sodium amide (2.11 g, 51.34 mmol). The solution was stirred at reflux under argon for 5 hours, at which time 2-chloroacetaldehyde diethyl acetal (6.60 g, 42.80 mmol) was added dropwise. The solution was stirred at reflux overnight, cooled and the salts filtered off. The filtrate was concentrated to a brown oil. The oil was taken up into ethyl acetate (100 ml) and washed with water (2 x 100 ml), brine (1 x 100 ml), dried (MgSO^), treated with charcoal and concentrated to a brown oil. Flash silica column chroma­ tography was performed using 5% ethyl acetate/95% hexane as the sol­ vent. (Rf = 0.47 silica on glass using 10% ethyl acetate/90% hexane). 200

Fractions containing a mixture of of the desired diethyl acetal 157 and starting material (R^ = 0.38) were concentrated, taken up into hexane and the resultant white precipitate (starting material) was filtered off. The filtrate and the fractions containing only product were con­ centrated to a light yellow oil (diethyl acetal, 157), 8.95 g (60%):

NMR (CDCI3 /TMS) b 7.18-7.09 (m, 2H, ArH), 7.02-6.86 (m, 5H, ArH), 4.82

(t, J = 4.7 Hz, 111, CH), 4.01 (d, J = 4.7 Hz, 211, CH?), 3.68 (dq, J m

= 9.2 Hz and Jyic = 7.1 Hz, 2H, 0CH2 CH3), 3.55 (dq, Jgem = 9.2 Hz and

Jy^c = 7.1 Hz, 2H, OCH2CH3 ), (the ethoxy methylenes appear as AB doub­ let of doublets, as a result of geminal coupling, which are further split to quartets by the vicinal methyl groups), 1.19 (t, J = 7.1 Hz,

6H, OCH2CH3 x 2); MS m/z 351 (M+ , 37C1), 349 (M+ , 35 C1), 103 (base,

CH(0F.t)2). To the above diethyl acetal oil (157) (7.90 g, 22.59 mmol) in dioxane (100 ml) and water (20 ml) was added oxalic acid dihydrate

(4.27 g, 33.87 mmol). The solution was stirred at reflux overnight, cooled and neutralized with solid sodium bicarbonate. The precipitate was filtered off and the solution concentrated. The residue was taken up into dichloromethane (125 ml), washed with water (2 x 125 ml), brine

(1 x 125 ml), dried (MgSO^), treated with charcoal and concentrated to a beige solid. Recrystallization from diethyl ether gave white nee­ dles, 4.60 g (74%): m.p. 112-115°C; (Lit. m.p. 112-114°C (benzene/ hexane); IR (KBr, cm"1) 1717 (C=0); 1H NMR (CDC13 /TMS) b 9.81 (t, J =

1.3 Hz, 1H, CH0), 7.19-6.93 (m, 5H, ArH), 6.59-6.55 (m, 2H, ArH), 4.51

(d, J = 1.3 Hz, 2H, CH2); MS m/z 277 (M+ , 37C1), 275 (M+ , 35C1), 248,

246 (base), 233. 201

2-[2-(2-Chloro-10-phenothlazln-10-yl)ethylldenel-1,1-dlmethylhydrazine

( 151 ),

Cl

N — CH

To a solution of acetaldehyde 158 (1.00 g, 3.63 mmol) in absolute ethanol (40 ml) was added 98% N,N-dimethylhydrazine (0.34 ml, 4.35 mmol). The solution was stirred at reflux under argon for 3 hours, cooled and concentrated under reduced pressure to a yellow oil. The oil was taken up into ethyl acetate (40 ml), washed with water (2 x 40 ml), brine (1 x 40 ml), dried (MgSO^), concentrated under reduced pres­ sured and dried under vacuum to a yellow oil, 1.15g (100%): IR (neat)

1591 (C=N); NMR (CDC13/TMS) <5 7.15-7.07 (m, 2H, ArH), 6.99-6.83

(m, 5H, ArH), 6.51 (t, J = 4.9 Hz, 1H, CH=N), 4.56 (d, J = 4.9 Hz, 2H,

CH2), 2.80 (s, 6H, CH3 x 2); MS m/z 317 (M+ , 35C1), 233, 85 (base). 202

2 -f 2-(2-Chlorophenothiazln-10-yl)ethylidene1-1,1,1-trimethylhydrazlnlum

Iodide (160)

To a solution of N,N-dimothy.1 hydrazone 159 (1.00 g, 3.15 mmol) in acetone (20 ml) was added f^H^I (2.00 ml). The solution was stirred at room temperature overnight and concentrated under reduced pressure to a yellow solid. Recrystallization from acetone/diethyl ether gave a white solid, 999 mg (69%): m.p. 139-140°C with decomposition (begins to turn brown at 132°C); IR (KBr, cm"1) 1592 (C=N); NMR (CD3OD) 3 8.54

(t, J = 3.1 Hz, 1H, CH=N), 7.22-6.90 (m, 7H, ArH), 4.98 (d, J = 3.1 Hz,

2H, CH2), 3.44 (s, 9H, CH3 x 3); FAB MS 334 (M+ , 37C1), 332 (M+ ; 35C1),

248, 246 (base), 275; Anal. Calcd. for C 17H igClIN3S'l/4H20: C, 43.98;

H, 4.23; N, 9.05. Found: C, 43.98; H, 4.28; N, 8.92. 203

2-f2-(2-Chlorophenothlazin-10-yl)ethyl1-1,1,1-trlmethylhydrazlnium iod­

ide (152)

A solution of hydrazonium 160 (300 mg, 0.65 mmol) and 95+% lithium aluminum hydride (52 mg, 1.30 mmol) in diethyl ether (30 ml) was refluxed under argon for 1 hour. The solution was cooled, quenched with ethyl acetate and a minimum amount of water and filtered. The filtrate was concentrated to a yellow foam and dried under vacuum over­ night. Recrystallization from methanol/diethyl ether gave white crys­ tals, 145 mg (48%): m.p. 189°C with decomposition; IR (KBr, cm *) 3139

(NH); 1H NMR (CD3OD) S 7.28-6.95 (m, 7H, ArH), 4.10 (t, J = 6.0 Hz,

2H, CH2), 3.36 (t, J = 6.0 Hz, 2H, CH2), 3.27 (s, 9H, CH3 x 3); FAB MS

336 (M+ - I", 37C1), 334 (M+ - I", 35C1), 248 (M+ - l“ -

CH2NHNMe3 , 3 7 C1), 246 (base, M+ - I" - CH2NHNMe3 , 35C1),; C ^ H ^ C l I N g S :

C, 44.22; H, 4.58; N, 9.10. Found: C, 44.11; H, 4.66; N, 9.13. 204

2-f3,4-Bis(benzyloxy)benzylidenel-1,1-dlmethylhydrazlne (165)

PhCH20 N CH,

PhCH20

To a solution of 3 ,4-dibenzyloxybenzaldehyde (Aldrich) (10.00 g,

31.41 mmol) in absolute ethanol (200 ml) was added p-toluenesulfonic acid monohydrate (300 mg, 1.58 mmol) and 98% N,N-dimethylhydrazine (2.9 ml, 37.69 mmol). The solution was stirred at reflux for 4 hours, cooled and concentrated under reduced pressure to a light yellow solid.

Recrystallization from ethanol/water gave a white solid, 10.09 g (89%): m.p. 60-62°C; IR (KBr, cm _1) 1596 (C=N); XH NMR (CDCI3 /TMS) <5

7.50-7.28 (m, 11H, ArH), 7.19 (s, 1H, ArCHN), 6.99 (dd, J •- 8.2 and 1.8

Hz, 1H, ArH), 6.88 (d, J = 8.2 Hz, 111, ArH), 5.18 (s, 2H, ArCH20), 5.15

(s, 2H, ArCH2 0), 2.92 (s, 6 H, CH3 x 2); MS m/z 360 (M+), 269 (M+ - ben­ zyl), 91 (base); Anal. Calcd. for C2 3 H2 4 N202 : C, 76.64; H, 6.71; N,

7.77. Found: C, 76.41; H, 6.73; N, 7.84. 205

2-[3,4-Bis(benzyloxy)benzylidenel-1,1,1-trimethylhydrazinium_____ iodide

(166)

PhCH20

PhCH20

To a solution of dimethylhydrazone 165 (208 mg; 0.58 mmol) in abso­ lute ethanol (20 ml) was added CH3 I (0.4 ml). The solution was stirred at room temperature for 72 hours. The resultant turbid solution was filtered to afford a white solid (washed with diethyl ether), 249 mg

(85%): m.p. 180-182°C; IR (KBr, cm-1) 1617 (C=N); ]H NMR (CDCI3 /TMS) S

9.85 (s, 1H, CH=N), 7.70 (dd, J = 8.4 and 1.8 Hz, 1H, ArH), 7.60 (d, J

= 1.8 Hz, 1H, ArH), 7.51-7.31 (m, 10H, ArH), 6.97 (d, J = 8.4 Hz, 1H,

ArH), 5.23 (s, 2H, ArCH20 ) , 5.20 (s, 21!, ArCH20), 3.85 (s, 9H, CH3 x

3); MS m/z 360 (M+ - CH3 I), 269 (M+ - CH3 I - benzyl), 91 (base); FAB MS

375 (M+ - I, base); Anal. Calcd. for C24 H2 7 IN202 : C, 57.38; H, 5.42; N,

5.58. Found: C, 57.57; H, 5.56; N, 5.58. 206

2-f 3,4-Bis(benzyloxy)benzyl1-1,1,1-trimethylhydrazinlum iodide (167)

CH

CH

To a solution of hydrazonium 166 (l.OOg, 1.99 mmol) in diethyl ether

(50 ml) under argon was charged 95+% lithium aluminum hydride (238 mg,

5.97 mmol). The solution was stirred at reflux for 1 hour, cooled and

quenched with ethyl acetate and filtered. The solids were washed with

ethyl acetate followed by methanol. The filtrate was concentrated and

recrystallized from methanol/diethyl ether to afford white crystals,

842 mg (84%): d.p. 205-208°C; IR (KBr, cm"1) 3152 (NH); LH NMR

(CDC1,/TMS) 5 7.49-7.24 (m, 10H, ArH), 7.17 (d, J = 1.6 Hz, 1H, ArH), J HI

6.97 (t, J = 6.3 Hz, 1H, NH), 6.91 (dd, JQ = 8.2 Hz and Jm = 1.7 Hz,

111, ArH), 6.86 (d, JQ = 8.2 Hz, 1H, ArH), 5.24 (s, 2H, ArCH20), 5.12

(s, 2H, ArCH20), 4.04 (d, J = 6.3 Hz, 2H, CH2), 3.43 (s, 9H, CH3 x 3);

FAB MS 377 (M+ - I"), 318 (M+ - i" - NMe3), 91 (base); Anal. Calcd. for

C24H29IN2°2: C » 57-15; H > 5 -80; N > 5 -S5- Found: C, 46.19; H, 5.55; N,

4.14 (suggests more than one molar equivalent of iodide, iodine elemen­ tal analysis pending). 207

2-(3,4-Dlhydroxybenzylidene)-l,1-dlmethylhydrazlne (169)

CH, I 3 HO

To a solution of 3 ,4-dihydroxybenzaldehyde (Aldrich) (1.00 g, 7.24 mmol) in absolute ethanol (20 ml) was added p-toluenesulfonic acid monohydrate (20 mg, 0.11 mmo]) and 98% N,N-dimethylhydrazine (0.66 ml,

8.69 mmol). The solution was stirred at reflux under argon for 4 hours, cooled and concentrated to a dark green-brown solid. The solid was taken up into methanol (50 ml), treated with charcoal and concen­ trated to a yellow solid. Recrystallization from hot methanol gave yellow needles (washed with ethyl acetate), 808 mg (62%): m.p.

159-160°C with decomposition; FeClg positive; IR (KBr, cm”^) 3480 (OH),

1624 (C=N); NMR (CD3OD) <5 7.32 (s, 111, CH=N), 7.08 (d, J = 2.0Hz,

1H, ArH), 6.85 (dd, J = 8.2 and 2.0 Hz, 1H, ArH), 6.71 (d, J = 8.2 Hz,

HI, ArH), 2.83 (s, 6 H, CH3 x 2); MS m/z 180 (M+ , base), 165, 150; Anal.

Calcd. for C9H 12N202 : C, 59.99; H, 6.71; N, 15.54. Found: C, 59.66; H,

6 .8 8 ; N, 15.47. 208

2-(3,4-Dlhydroxybenzylidene)-l,1,1-trimethylhydrazinium Iodide (170)

To a solution of dimethylhydrazone 169 (200 mg, 1.11 mmol) in abso­ lute ethanol (20 ml) was added CH3 I (0.4 ml). The solution was stirred at room temperature for 72 hours, diethyl ether was added to induce precipitation. The solution was stirred an additional 48 hours and filtered to afford a yellow solid, 306 mg (85%): m.p. 174-175°C with decomposition; FeClg positive; IR (KBr, cm-*) 3479 (OH), 1652 (C=N); *H

NMR (CD3OD) S 8.83 (s, 1H, CH=N), 7.34 (d, J = 2.0 Hz, -111, ArH), 7.24

(dd, J = 8.2 and 2.0 Hz, 1H, ArH), 6.87 (d, J = 8.2 Hz, 1H, ArH), 3.52

(s,- 9H, CH3 x 3); MS m/z 180 (M&S+. - CH3 I), 135, 58 (base, CH2=NMe2) ;

FAB MS 195 (M+ - I, base); Anal. Calcd. for C 10H 1 5 IN202 : C, 37.28; H,

4.69; N, 8.70. Found: C, 37.53; H, 4.64; 8.49. BIBLIOGRAPHY

1. Lefkowitz, R. J. In "Cecil Textbook of Medicine", 17th ed., Wyngaarden, J. B., Smith, L. H., Eds.; W. B. Saunders Company: Philadelphia, 1985; p 90-95.

2. Guyton, A. C. "Textbook of Medical Physiology", 7th ed.; W. B. Saunders, Company: Philadelphia, 1986; p686-697 and p78-79.

3. Lefkowitz, R. J.; Hoffman, B. B.; Taylor, R. In "Goodman and Gilman's Pharmacological Basis of Therapeutics", 8 th ed.; Gilman, A. G., Rail, T. W., Nies, A. S., Taylor, P., Eds.; Pergamon Press: New York, 1990; p84-121.

4. Cooper, J.R.; Bloom, F. E.; Roth, R. H. "The Biochemical Basis of Neuropharmacology", 5th ed.; Oxford University Press, Inc.: New York, 1986; p203-214.

5. Blaschko, H. J. Physiol. (London) 1939, 96, 50P.

6 . Nagatsu, T.; Levitt, M.; Udenfiend, S. Biochem. Biophys. Res. Commun. 1964, 14, 543.

7. Nogrady, T. "Medicinal Chemistry: A Biochemical Aproach"; Oxford University Press, Inc.: New York, 1988; pl62-185.

8 . Langer, S. Z. Pharmacol. Rev. 1980, 32, 337.

9. Kopin, I, J. Pharmacol .Re v . 1985, 37^, 333.

10. Mimnaugh, M. N.; Gearien, J. E. In "Principles of Medicinal Chemistry", 3rd. ed., Foye, W. 0. (Ed.): Lea and Febiger: Philadelphia, 1989; p343.

11. Ahlquist, R. P. Am. J. Physiol. 1948, 153, 586.

12. Lands, A. M.; Arnold, A.; McAuliff, J. P.; Luduena, F. P.; Brown, T. G. Nature (London) 1967, 214, 597.

13. Minneman, K. P.; Hegstrand, L. R.; Molinoff, P. B.; Mol. Pharmacol. 1979, 16, 34.

- 209 - 210

14. Arch, J. R. S.; Ainsworth, A. T . ; Cawthorne, M. A.; Piercy, V.; Sennltt, M. V.; Thody, V. E.; Wilson, C.; Wilson, S. Nature (London) 1984, 309, 163.

15. Emorine, L. J . ; Marullo, S.; Brlend-Sutren, M . ; Patey, G,; Tate, K.; Delavler-KLutchko, C; Strosberg, A. D. Science 1989, 245, 1118.

16. Alder-Graschinsky, E.; Langer, S. Z. Br. J. Pharmacol. 1975, 53, 43.

17. Miso, Y,; Kubo, T. Med. Res. Rev. 1986, 6 , 197.

18. Langer, S. Z. Blochem. Pharmacol. 1974, 23, 1793.

19. Berthelsen, S.; Pettinger, W. Life Sci■ 1977, 21, 595.

20. Minneman, K. P. Pharmacol■ Rev. 1988, 40, 87.

21. Ruffolo, Jr., R. R. Neurotransmissions 1990, 6 , 1.

22. McGrath, J. C. Blochem. Pharmacol. 1982, 31, 467.

23. Bevan, J. A. J. Pharmacol. Exp. Ther. 1981, 216, 83.

24. McGrath, J.; Wilson, V. Trends Pharmacol. Sci. 1988, 9, 162.

25. McGrath, J. C.; O'Brien, J. W. Br. J. Pharmacol■ 1987, 91, 355.

26. Flavahan, N. A.; Vanhoutte, P. M. Trends Pharmacol. Sci. 1986, 7, 347.

27. Morrow, L. A.; Creese, I. Mol. Pharmacol. 1986, 29, 321.

28. Johnson, R. D.; Minneman, K. Mol. Pharmacol. 1987, 31, 239.

29. Han, C.; Abel, P. W.; Minneman, K. P. Mol. Pharmacol. 1987, 32, 505.

30. Han, C.; Abel, P.; Minneman, K. Nature (London) 1987, 329, 333.

31. Cotecchia, S.; Schwinn, D . ; Randall, R.; Lefkowitz, R. J.; Caron, M. G.; Kobilka, B. K. Proc. Natl. Acad. Sci. USA 1988, 85, 7159.

32. Terman, B. I.; Riek, R. P.; Grodski, A.; Hess, H.-J.; Grahman, R. M. Mol. Pharmacol. 1990, 37, 526.

33. Schwinn, D. A.; Lomasney, J. W.; Lorenz, W.; Szklut, P.; Fremeau, Jr., R. T.; Yang-Feng, T. L . ; Caron, M. G.; Lefkowitz, R. J.; Cotecchia, S. J. Biol. Chem. 1990, 265, 8183. 211

Bevan, J. A.; Bevan, R. D. ; Kite, K . ; Oriowo, M. A. Trends Pharmacol. Sci. 1988, 9, 87.

Bevan, J. A.; Bevan, R. D.; Shreeve, S. M. FASEB J. 1989, 3, 1696.

Timmermans, P. B. M. W. M.; van Zwieten, P. A. J. Med. Chem. 1982, 25, 1389.

37 Ruffolo, Jr.; R. R.; Sulpizio, A. C.; Nichols, A. J.; DeMarinis, R. M.; Hieble, J. P. Naunyn-Schmiedeberg1s Arch. Pharmacol. 1987, 336, 415.

38 Hieble, J. P.; Sulpizio, A. C.; Nichols, A. J . ; Willette, R. N.; Ruffolo, Jr., R. R. J. Pharmacol. Exp. Ther■ 1988, 247, 645.

39 Bylund, D. B. Pharmacol. Blochem. Behav. 1985, 22, 835.

40 Nahorskl, S. R.; Barnett, D. B.; Cheund, Y. D. Clin. Sci. 1985, 6 8 , 39s.

41 Bylund, D. B.; Ray-Prenger, C.; Murphy, T. J. J. Pharmacol. Exp. Ther. 1988, 245, 600.

42 Kobilka, B. K.; Matsul, H.; Kobilka, T. S.; Yang-Feng, T. L.; Francke, U.; Caron, M. G . ; Lefkowitz, R. J.; Regan, J. W. Science 1987, 238, 650.

43 Murphy, T. J.: Bylund, D. B. J. Pharmacol. Exp. Ther. 1988, 244, 571.

44 Bylund, D. B. Trends Pharmacol. Sci. 1988, 9, 356.

45 Regan, J. W.; Kobilka, T. S.; Yang-Feng, T. L.; Caron, M. G.; Lefkowitz, R. J.; Kobilka, B. K. Proc. Natl. Acad. Sci. USA 1988, 85, 6301.

46 Michel, M. C.; Insel, P. A. Trends Pharmacol. Sci. 1989, 10, 342.

47 Lehmann, J.; Koenig-Berard, E.; Vitou, P. Life Sci. 1989, 45, 1609.

48 Brown, C. M.; MacKinnon, A. C.; McGrath, J. C.; Spedding, M . ; Kilpatrick, A. T . ; Br. J. Pharmacol. 1990, 99, 803.

49 Timmermans, P. B. M. W. M. In "Recent Advances in Receptor Chemistry"; Melchiorre, C.; Giannella, M. (Eds.); Elsevier: New York, 1988, p29-41. 212

50. Taylor, P.; Insel, P. A. In "Principles of Drug Action: The Basis of Pharmacology, 3rd ed."; Pratt, W. B.; Taylor, P. (Eds.); Churchill Livingstone Inc.: New York, 1990, pl44-198.

51. Jones, D. T.; Masters, S. B.; Bourne, H. R.; Reed, R. R. J. Biol. Chem. 1990, 265, 2671.

52. Neer, E. J . ; Clapham, D. E. Nature (London) 1988, 333, 129.

53. Yatani, A.; Mattera, R.; Codina, J.; Graf, R.; Okabe, K.; Padrell, E.; Iyengar, R.; Brown, A. M.; Birnbaumer, L. Nature (London) 1988, 336, 680.

54. Graziano, M. P.; Gilipan, A. G. Trends Pharmacol. Sci. 1987, 8 , 478.

55. Isom, L. L.; Limbird, L. E. In "The Alpha-2 Adrenergic Receptors" Limbird, L. E. (Ed.); Humana Press: New Jersey, 1988, p323-363.

56. O'Dowd, B. F.; Lefkowitz, R. J . ; Caron, M. G. Ann. Rev. Neurosci. 1989, 12, 67.

57. Strader, C. D . ; Sigal, I. S.; Dixon, R. A. F. FASEB J. 1989, 3, 1825.

58. Dohlman, H. G.; Caron, M. G.; DeBlasi, A.; Frielle, T.; Lefkowitz, R. J. Biochemistry 1990, 29, 2335.

59. Dixon, R. A. F. ; Sigal, I. S.; Candelore, M. R.; Register, R. B.; Scattergood, W.; Rands, E.; Strader, C. D. EMBO J. 1987, 6 , 3269.

60. Dixon, R. A. F . ; Sigal, I. S.; Rands, E.; Register, R. B.; Candelore, M. R.; Blake, A. D . ; Strader, C. D. Nature (London) 1987, 326, 73.

61. Strader, C. D . ; Sigal, I. S.; Candelore, M. R.; Hill, W. S.; Dixon, R. A. F. J. Biol. Chem. 1988, 263, 10267.

62. Strader, C. D.; Candelore, M. R.; Hill, W. S.; Sigal, I. S.; Dixon, R. A. F. J. Biol. Chem. 1989, 264, 13572.

63. Dixon, R. A. F. ; Sigal, I. S.; Strader, C. D. Cold Spring Harbor Symp. Quant. Biol. 1988, 53, 487.

64. IJzerman, A. P.; van Vlijmen, H. W. Th. J. Comput. Aided Mol. Des. 1988, 2, 43.

65. Kobilka, B. K . ; Kobilka, T. S.; Daniel, K.; Regan, J. W.; Caron, M. G.; Lefkowitz, R. J. Science 1988, 240, 1310. 213

66. Strader, C. D.; Dixon, R. A. F.; Cheung, A. H.; Candelore, M. R.; Blake, A. D.; Sigal, I. S. J. Biol. Chem. 1987, 262, 16439.

67. Cotecchia, S.; Exum, S.; Caron, M. G.; Lefkowitz, R. J. Proc■ Natl. Acad. Sci. USA 1990, 87, 2896.

68. Lefkowitz, R. J.; Hausdorff, W. P.; Caron, M. G. Trends Pharmacol. Sci. 1990, 1J., 190.

69. Benovic, J. L.; Regan, J. W.; Matsui, H.; Mayor, Jr., F.; Cotecchia, S.; Leeb-Lundberg, L. M. F.; Caron, M. G.; Lefkowitz, R. J. J. Biol. Chem. 1987, 262, 17251.

70. Caron, M. G . ; Kobilka, B. K.; Frielle, T . ; Benovic, J. L . ; Regan, J. W.; Collins, S.; McGregor, C.; Lefkowitz, R. J. In "Recent Advances In Receptor Chemistry" Melchiorre, C.; Giannella, M. (Eds.); Elsevier: New York, 1988, p63-76.

71. Ruffolo, Jr., R. R. In "Recent Advances In Receptor Chemistry" Melchiorre, C.; Gianella, M. (Eds.); Elsevier: New York, 1988, p77-84.

72. Ruffolo, Jr., R. R. In "Adrenoceptors and Catecholamine Action: Part B" Kunos, G. (Ed.); Wiley-Interscience: New York, 1983, pl-50.

73. Easson, L. H.; Stedman, E. Blochem . J. 1933, 27,1257.

74. Ruffolo, Jr., R. R.; Waddell, J. E. Life Sci. 1982, 31, 2999.

75. Tlmmermanns, P. B. M. W. M.; Chiu, A. T . ; Thoolen, M. J. M. C. In "Comprehensive Medicinal Chemistry Vol. 3" Hansch, C . ; Sammes, P. G.; Taylor, J. B.; Emmett, J. C. (Eds.); Pergamon Press: New York, 1990, pl33-185.

76. Ruffolo, Jr., R. R.; Rice, P. J.; Patil, P. N.; Hamada, A.; Miller, D. D. Eur. J. Pharmacol. 1983, 8 6 , 471.

77. Ruffolo, Jr., R. R.; DeMarinis, R. M.; Wise, M . ; Hieble, J. P. In "The Alpha-2 Adrenergic Receptors" Limbird, L. E. (Ed.); Humana Press: New Jersey, 1988, pll5-186.

78. Pullman, B.; Coubeils, J-L.; Courriere, Ph.; Gervois, J-P. J. Med. Chem. 1972, 15, 17.

79. Tlmmermanns, P. B. M. W. M.; van Zwieten, P. A. Eur. J. Med. Chem.- Chim. Ther. 1980, 15, 323.

80. Karjalalnen, A. J.; Virtanen, R. E; Sevolalnen, F.. J. U.K. Patent Application GB 2206880. 214

81. Scheinin, H.; Virtanen, R.; Macdonald, E.; Lammintausta, R.; Scheinin, M. Prog. Neuro-Psychpharmacol. and Biol■ Psychiat. 1989, 13, 635.

82. Amemiya, Y.; Venkataraman, B. V.; Patil, P. N. ; Shams, G.; Romstedt, K.; Feller, D. R.; Hsu, F.-L.; Miller, D. D. J. Med. Chem. in preparation for submission.

83. Miller, D. D.; Hamada, A.; Clark, M. T.; Adejare, A.; Patil, P. N.; Shams, G.; Romstedt, K. J.; Kim, S. U.; Intrasuksri, U.; McKenzie, J. L.; Feller, D. R. J. Med. Chem. 1990, 33, 1138.

84. Main, B. G. In ’’Comprehensive Medicinal Chemistry Vol. 3" Hansch, C.; Sammes, P. G.; Taylor, J. B.; Emmett, J. C. (Eds.); Pergamon Press: New York, 1990, pl95-198.

85. Bergman, J.; Persson, H.; Wetterlin, K. Experientia 1969, 25, 899.

8 6 . O'Donnell, S. R. Eur. J. Pharmacol■ 1970, 12, 35.

87. Yamato, E.; Hirakura, M.; Sugasawa, S. Tetrahedron Suppl. 1966, 8 (Part 1), 129.

8 8 . Bilezikian, J. P. In "Adrenergic Receptors in Man" Insel, P. A. (Ed.); Marcel-Dekker, Inc.: New York, 1987, p37-67.

89. Ruffolo, Jr., R. R.; Nichols, A. J.; Hieble, J. P. In "The Alpha-2 Adrenergic Receptors" Limbird, L. F.. (Ed.); Humana Press: New Jersey, 1988, pl87-280.

90. Heusch, G. Circulation 1990, 81, 1 .

91. Goldie, R. G. Am. Rev. Respir. Pis. 1990, 141, S151.

92. Campbell, W. B. In "Goodman and Gilman's The Pharmacological Basis of Therapeutics, 8 th ed." Gilman, A. G.; Rail, T. W.; Nies, A. S.; Taylor, P. (Eds.); Pergamon Press: New York, 1990, p600-617.

93. Gerrard, J. M. In "Prostaglandins and Leukotrienes" Marcel Dekker, Inc.: New York, 1985, p47-59.

94. Lasslo, A. Fed. Proc. 1984, 43, 1382.

95. Handin, R. In "Harrison's Principles of Internal Medicine, 11th ed., Vol. l" Braunwald, E.; Isselbacher, K. J.; Petersdorf, R. G.; Wilson, J. D.; Martin, J. B.; Fauci, A. S. (Eds.); McGraw-Hill: New York, 1987, p266-272.

96. Plow, E. F.; Ginsberg, M. H.; Marguerie, G. A. In "Biochemistry of Platelets" Phillips, D. R.; Shuman, M. A. (Eds.); Academic Press: Orlando, FL, 1986, p225-256. 215

97. Packham, M. A.; Mustard, J. F. In "Blood Platelet Function and Medicinal Chemistry" Lasslo, A. (Ed.); Elsevier Biomedical: New York, 1984, p61-128.

98. Taylor, M. D. In "Comprehensive Medicinal Chemistry, Vol. 2" Hansch, C.; Sammes, P. G.; Taylor, J. B. (Eds.); Pergamon Press: New York, 1990, p481-500.

99. Majerus, P. W.; Broze, Jr.; G. J.; Miletich, J. P.; Tollefsen, D. M. In "Goodman and Gilman's The Pharmacological Basis of Therapeutics, 8 th ed." Gilman, A. G.; Rail, R. W.; Nies, A. S.; Tayor, P. (Eds.); Pergamon Press: New York, 1990, pl311-1313.

100. Burch, R. M.; Mais, D. E.; Saussy, Jr., D. L . ; Halushka, P. V. Proc. Natl. Acad. Sci. USA 1985, 82, 7434.

101. Feller, D. R., unpublished work.

102. Berridge, M. J. Blochem. J. 1984, 220, 345.

103. Huang, E. M.; Detwiler, T. C. In "Biochemistry of Platelets" Phillips, D. R.; Shuman, M. A. (Eds.); Academic Press: Orlando, FL, 1986, pl-9.

104. Page, C. P. Trends Pharmacol. Sci. 1988, 9, 6 6 .

105. deGaetano, G.; Cerletti, C.; Dejana, E.; Vermylen, J. Drugs, 1986, 31, 517.

106. Fuster, V.; Chesebro, J. Mayo Clin. Proc. 1981, 56. 102.

107. Moncada, S; Higgs, E. A. Clin. Haematol. 1986, T5, 273.

108. Milner, P. C.; Martin, J. F. Brit. Med. J. 1985, 290, 1767.

109. Tanabe, M . ; Terashita, Z.; Fujiwara, S.; Shimamoto, N . ; Goto, N.; Nishikawa, K.; Hirata, M. Cardiovasc■ Res. 1982, _16, 99.

110. Smith, E. F.; Defer, A. M . ; Aharony, D . ; Smith, J. B.; Magolda, R. L.; Claremon, D.; Nicolaou, K. C. Prostaglandin 1981, 21, 443.

111. Schror, K.; Smith, E. F.; Bickerton, M.; Smith, J. B.; Nicolaou, K. C.; Magolda, R.; Lefer, A. M. Am. J. Physiol. 1980, 238, H87.

112. Grover, G. J.; Schumacher, W. A. Basic Res. Cardiol. 1989, 84, 103.

113. Ruffolo, Jr., R. R.; Nichols, A. J. Adv. Drug Res■ 1988, 17, 307. 216

114. Konieczkowski, M.; Dunn, M.; Hass Is, A. Fed. Proc. 1982, 41, 1543.

115. Mals, D. E.; Saussey, Jr., D. L.; Chaikhounl, A.; Kochel, P. J.; Knapp, D. R.; Hamanaka, N.; Halushka, P. V. J. Pharmacol. Exp. Ther. 1985, 233, 418.

116. Bundy, G. L. Tet. Lett. 1975, 24, 1957.

117. LeBreton, G. C.; Venton, D. L . ; Enke, S. E.; Halushka, P. V. Proc. Natl. Acad. Sci. USA 1979, 76, 4097.

118. Nicolaou, K. C.; Magolda, R. L . ; Smith, J. B.; Aharony, D.; Smith, E. F . ; Lefer, A. M. Proc. Natl. Acad. Sci. USA 1979, 76, 2566.

119. Harris, D. N.; Michel, I. M.; Goldenberg, H. J.; Allen, G. T.; Hall, S. E.; Floyd, D. M.; Ogletree, M. L. Thromb. Haemostas. 1989, 62, 311, Abst. 1004.

120. Humphrey, P. P. A.; Hallet, P.; Hornby, E. J.; Wallis, C. J.; Collington, E. W.; Lumley, P. Circulation 1990, 81, Suppl. 1, 1-42.

121. Brewster, A. G.; Brown, G. R.; Jessup, R.; Smithers, M. J. Abstract of Papers. 197th ACS National Meeting, Dallas, TX, 1989, Abst. MEDI 11.

122. Shtacher, G.; Crowley, H. J.; Dalton, C. Biochem. Pharmacol. 1976, 25, 1045.

123. Stegmeier, K.; Pill, J.; Muller-Beckman, B.; Schmitt, F. H.; Witte, E. C.; Wolff, H. P.; Patscheke, H. Thromb. Res. 1984, 35, 379.

124. Girard, Y.; Yoakim-Rancourt, C.; Hamel, P.; Gillard, J. W.; Guindon, Y.; Letts, G.; Evans, J . ; Level He , C . ; Ethier, D.; Lord, A.; Jones, T.; Masson, P.; Ford-Hutchinson, A. W.; Rokach, J. Prostaglandins in Clin. Res.: Cardiovascular System, Prog. Clin. Biol. Res. Series 1989, 301, 585.

125. Tomiyama, T.; Wakabayashi, S.; Kosakai, K.; MRsayuki, Y. J. Med. Chem. 1990, 33, 2323.

126. Iwasawa, Y.; Kiyomoto, A. Jpn. J. Pharmacol. 1967, 17, 143.

127. Mukhopadhyay, A.; Navran, S. S.; Amin, H. M.; Abdel-Aziz, S. A.; Chang, J.; Sober, D. J.; Miller, D. D.; Feller, D. R. J. Pharmacol. Exp. Ther. 1985, 232, 1.

128. Feller, D. R.; Venkatraman, R.; Miller, D. D. Biochem. Pharmacol. 1975, 24, 1357. 217

129. Dalton, C . ; Crowley, H. J.; Czyzewski, L. B. Biochem. Pharmacol. 1976, 25, 2209.

130. Mayo, J. R.; Navran, S. S.; Akbar, H.; Miller, D. D . ; Feller, D. R. Biochem. Pharmacol. 1981, 30, 2237.

131. Ahn, C . ; Romstedt, K. J.; Wallace, L. J.; Miller, D. D.; Feller, D. R. Biochem. Pharmacol. 1988, 37, 3023.

132. Armstrong, R. A.; Jones, R. L . ; Wilson, N. H. Br. J. Pharmacol. 1983, 79, 953.

133. MacIntyre, D. E.; Pollock, W. K. Br. J. Pharmacol. 1983, 78, 600P.

134. Ahn, C. ; Wallace, L. J.; Miller, D. D.; Feller, D. R. Thromb. Res. 1988, 50, 387.

135. Clark, M. T.; Adejare, A.; Shams, G . ; Feller, D. R.; Miller, D. D. J. Med. Chem. 1987, 30, 8 6 .

136. Miller, D. D . ; Clark, M. T.; Adejare, A.; Neidert, K.; Hamada, A.; Shams, G.; Romstedt, K. J.; Feller, D. R. In "Progress in Catecholamine Research, Part A: Basic Aspects and Peripheral Mechanisms" Dahlstrom, A.; Belmaker, R. H.; Sandler, M. (Eds.); Alan R. Liss, Inc.: New York, 1988, p403-407.

137. Bowden, K. In "Comprehensive Medicinal Chemistry, Vol. 4" Hansch, C.; Sammes, P. G . ; Taylor, J. B.; Ramsden, C. A. (Eds.); Pergamon Press: New York, 1990, p234-235.

138. Smith, J.; Williams, H. In "introduction to the Principles of Drug Design"; Wright-PSG: Boston, 1983, p56-68.

139. Ruffolo, Jr., R. R.; Patil, P. N.; Miller, D. D. Naunyn-Schmiedeberg's Arch. Pharmacol. 1983, 323, 221.

140. Kirk, K. L.; Cantacuzene, D.; McCulloh, D. H.; Creveling, C. R. Science 1979, 204, 1217.

141. Lambda-Kanwa1, V. K.; Hamada, A.; Adejare, A.; Clark, M. T . ; Miller, D. D.; Patil, P. N. J. Pharmacol. Exp. Ther.1988, 245, 793.

142. DeJonge, A.; Van Meel, J. C. A.; Timmermans, P. B. M. W. M.; Van Zwieten, P. A. Life Sci. 1981, 28, 2009.

143. Goldberg, L. I.; Kohli, J. D.; Cantacuzene, D . ; Kirk, K. L. ; Creveling, C. R. J. Pharmacol. Exp. Ther■ 1980, 213, 509. 218

144. Leo, A. J. In "Comprehensive Medicinal Chemistry, Vol. 4" Hansch, C.; Sammes, P. G.; Taylor, J. B.; Ramsden, C. A. (Eds.); Pergamon Press: New York, 1990, p298.

145. Tantishaiyakul, V. Ph.D. Dissertation, The Ohio State University, 1989.

146. Matsui, K; Tobita, E . ; Ando, M.; Kondo, K. Chem■ Lett. 1981, 1719.

147. Mann, B. E. In "NMR and the Periodic Table" Harris, R. K.; Mann, B. E. (Eds.); Academic Press: New York, 1978, p9B-100.

148. Ladd, D. L . ; Weinstock, J. J. Org. Chem. 1981, 46, 203.

149. Backstrom, R.; Honkanen, E.; Pippuri, A.; Kairisalo, P.; Pystynen, J.; Heinola, K.; Nissinen, E.; Linden, I.-B.; Mannisto, P. T.; Kaakkola, S.; Pohto, P. J. Med. Chem. 1989, 32, 841.

150. Lambert, G. J.; Duffley, R. P.; Dalzell, H. C.; Razden, R. K. J. Org. Chem. 1982, 47, 3350.

151. Markovich, K. M.; Hamada, A.; Miller, D. D. J. Heterocyclic Chem. 1990, 27, 1665.

152. Kobayashi, Y.; Yamamoto, K.; Asai, T . ; Nakano, M.; Kumadaki, I. J. C. S. Perkin I 1980, 2755.

153. McLoughlin, V. C. R.; Thrower, J. Tetrahedron 1969, 25, 5921.

154. Kametani, T.; Fukumoto, K . ; Terui, T.; Yamaki, K.; Taguchi, E. J. Chem. Soc. (C) 1971, 2709.

155. Kametani, T.; Honda, T.; Ihara, M. J. Chem. Soc. (C) 1971, 3318.

156. Pirkle, W. H.; Pochapsky, T. C. Chem. Rev. 1989, 89, 347.

157. Wainer, I. W. In "A Practical Guide to the Selection and Use of HPLC Chiral Stationary Phases" J. T. Baker: New Jersey, 1988, p4-6.

158. Okamoto, Y.; Kawashima, M.; Hatada, K. J. Chromatogr. 1986, 363, 173.

159. Adejare, A.; Miller, D. J. Chem. Research (S) 1986, 14.

160. Kirk, K. L.; Cantacuzene, D.; Nimitkitpaisen, Y.; McCulloh, D.; Padgett, W. L.; Daly, J. W.; Creveling, C. R. J. Med. Chem. 1979, 22, 1493. 219

161. Evans, D. A.; Carroll, G. L.; Truesdale, L. K. J. Org. Chem. 1974, 39, 914.

162. Adejare, A. Ph.D. Dissertation, The Ohio State University, 1985.

163. Kaiser, C.; Jain, T. Med. Res. Rev. 1985, 5, 145.

164. Sachar, E. J. In "Principles of Neural Science", Kandel, E. R.; S.chwartz, J. H. , Eds.; Elsevier: New York, 1981; p607-609.

165. Snyder, S. H.; Banerjee, S. P.; Yamamura, H. I.; Greenberg, D. Science 1974, 184, 1243.

166. Seeman, P.; Niznik, H. B. FASEB J. 1990 4, 2737.

167. Ince, F. In "Comprehensive Medicinal Chemistry Vol. 3", Hansch, C.; Sammes, P. G.; Taylor, J. B.; Emmett, J. C., Eds.; Pergamon Press: New York, 1990; p291-321.

168. Abboud, F. M. In "Cecil Textbook of Medicine, 17th ed.", Wyngaargen, J. B., Smith, L. H . , Eds.; W. B. Saunders Company: Philadelphia, 1985; p222.

169. Goldberg, L. I.; Hsieh, Y. Y.; Resnekov, L. Prog. Cardiovasc. Pis. 1977, 19, 327.

170. Kebabian, J. W.; Caine, D. B. Nature 1979, 277, 93.

171. Vallar, L. ; Meldolesi, J. Trends Pharmacol. Sci. 1989, 10, 74.

172. Sibley, D. R.; DeLean, A.; Creese, I. J. Biol. Chem. 1982, 257, 6351.

173. Hamblin, M.; Leff, S.; Creese, I. Biochem. Pharmacol. 1984, 33, 877.

174. Miller, D. D . ; Uretsky, N. J. Pharm. Internet■ 1986, 7, 95.

175. Goldberg, L. I.; Kohli, J. D. Trends. Pharmacol. Sci. 1.983, 4, 64.

176. Kebabian, J. Vf. In "Dopamine Receptor Agonists", Poste, G. ; Crooke, S. T . , Eds.; Plenum Press: New York, 1984; p3-22.

177. Stoof, J. C.; Kebabian, J. W. Life Sci. 1984, 35, 2281.

178. Kebabian, J. W. Neurotransmissions 1986, 2, 1.

179. Setler, P. E.; Sarau, H. M.; Zirkle, C. L.; Saunders, H. L. Eur■ J. Pharmacol. 1978, .50, 419. 220

180. Hieble, J. P. Ann. Rep. Med. Chem. 1987, 22, 107.

181. Hilditch, A.; Drew, G. M. Trends Pharmacol. Sci. 1985, 6 , 396.

182. Anderson, P. H.; Gingrich, J. A.; Bates, M. D. ; Dearry, A.; Falardeau, P.; Senogles, S. E.; Caron, M. G. Trends. Pharmacol. Sci. 1990, 11, 231.

183. Sokoloff, P.; Giros, B.; Martes, M. P.; Bouthenet, M. L.; Schwartz, J. C. Nature 1990, 347, 146.

184. Bunzow, J. R.; VanTol, H. H. M . ; Grandy, B. K.; Albert, P.; Salon, J.; MacDonald, C.; Machida, C. A.; Neve, K. A.; Civelli, D. Nature 1988, 336, 783.

185. Grandy, D. K.; Marchionni, M. A.; Makam, H. ; Stofko, R. E.; Alfano, M.; Frothingham, L. Fischer, J. B.; Burke-Howie, K. J.; Bunzow, J. R.; Server, A. C.; Civelli, 0. Proc. Natl. Acad. Sci. USA 1989, 8 6 , 9762.

186. Giros, B.; Sokoloff, P.; Martres, M.-P.; Riou, J.-F.; Emorine, L. J.; Schwartz, J.-C. Nature 1989, 342, 923.

187. Monsma, Jr., F. J.; McVittie, L. D. ; Gerfen, C. R.; Mahan, L. C.; Sibley, D. R. Nature 1989, 842, 926.

188. Todd, R. D . ; Khurana, T. S.; Sajovic, P.; Stone, K. R.; O'Malley, K. L. Proc. Natl. Acad. Sci. USA 1989, 8 6 , 10134.

189. Dearry, A.; Gingrich, J. A.; Falardeau, P.; Fremeau, Jr., R. T . ; Bates, M. D. ; Caron, M. G. Nature 1990, 347, 72.

190. Zhou, Q.-Y.; Grandy, D. K.; Thambi, L . ; Kushner, J. A.; Van Tol, H. H. M.; Cone, R.; Pribnow, D . ; Salon, J.; Bunzow, J. R.; Civelli, 0. Nature 1990, 847, 76.

191. Sunahara, R. K.; Niznik, H. B.; Weiner, D. M.; Stormann, T. M.; Brann, M. R.; Kennedy, J. L . ; Gelernter, J. E.; Rozmahel, R.; Yang, Y.; Israel, Y.; Seeman, P.; O'Dowd, B. F. Nature 1990, 347, 80.

192. Katerinopoulos, H. E.; Schuster, D. I. Drugs of the Future 1987, 12, 2-23.

193. Horn, A. S. In "Comprehensive Medicinal Chemistry Vol. 3", Hansch, C.; Sammes, P. G.; Taylor, J. B.; Emmett, J. C., Eds.; Pergamon Press: New York, 1990; p276-280.

194. Kaiser, C. In "Dopamine Receptor Agonists", Poste, G.; Crooke, S. T. , Eds.; Plenum Press: New York, 1984; pll9-121. 221

195. Kaiser, C.; Jain, T. In "Third SCI-RSC Medicinal Chemistry Symposium", Lambert, R. W., Ed.; Royal Society of Chemistry: London, 1986; pl33-168.

196. Cassady, J. M.; LI, G. S.; Spltzner, E. B.; Floss, H. G.; Clemens, J. A. J. Med. Chem. 1974, 17, 300.

197. Armstrong, J.; Barlow, R. B. Br. J. Pharmacol■ 1976, 57, 501.

198. Nichols, D. E.; In "Dopamine Receptors", Kaiser, C.; Kebabian, J. W., Eds.; American Chemical Society: Washington D. C., 1983; p213-215.

199. Remy, D. C.; Martin, G. E. Ann. Reg. Med. Chem. 1980, _15, 12.

200. Miller, D. D.; Harrold, M.; Wallace, R. A.; Wallace, L. J.; Uretsky, N. J. Trends Pharmacol. Sci. 1988, 9, 282.

201. Anderson, K.; Kitruvilla, A.; Uretsky, N. J.; Miller, D. D. J. Med. Chem. 1981, 24, 683.

202. Wallace, R. A.; Farooqui, T . ; Wallace, L.; Ares, J.; Chang, Y. A.; Miller, D. D.; Uretsky, N. J. Biochem. Pharmacol■ 1987, 36, 3903.

203. Wallace, R. A.; Farooqui, T . ; Wallace, L.; Ares, J.; Chang, Y. A.; Miller, D. D . ; Uretsky, N. J. Biochem. Pharmacol. 1988, 37, 2077.

204. Chang, Y. A.; Ares, J.; Anderson, K.; Sabol, B.; Wallace, R. A.; Farooqui, T . ; Uretsky, N.; Miller, D. D. J. Med. Chem. 1987, 30, 214.

205. Seeman, P.; Guan, H.-C. Mol. Pharmacol. 1987, 32, 760.

206. Gao, Y.; Ram, V. J.; Campbell, A.; Kula, N. S.; Baldessarini, R. J.; Neumeyer, J. L. J. Med. Chem. 1990, 33, 39.

207. Harrold, M. W.; Chang, Y.; Wallace, R. A.; Farooqui, T . ; Wallace, L. J.; Uretsky, N. J.; Miller, D. D. J. Med. Chem. 1987, 30, 1631.

208. Harrold, M. W.; Wallace, R. A.; Farooqui, T.; Wallace, L. J.; Uretsky, N . ; Miller, D. D. J. Med. Chem. 1989, 32, 874.

209. Zimmerman, J. J.; Feldman, S. In "Principles of Medicinal Chemistry, 3rd ed.", Foye, W. 0., F.d.; Lea & Febiger: Philadelphia, 1989, pl3-14.

210. Williamson, R. A.; Strange, P. G. J. Neurochem■ 1990, 55, 1357. 222

211. Rudner, B.; Prapas, A. G. U.S. Patent 3,147,255 Sept. 1, 1964.

212. Tamura, Y.; Minamikawa, J.; Kita, Y.; Kim, J. H.; Ikeda, M. Tetrahedron 1973, 29, 1063.

213. Tamura, Y.; Minamikawa, J.; Miki, Y.; Matsugashita, S.; Ikeda, M. Tet. Lett. 1972, 40, 4133.

214. Carpino, L. A. J. Am. Chem. Soc. 1960, 82, 3133.

215. Corral, C.; Lissavetzky, J.; Madronero, R. Eur. J. Med. Chem. 1978, 13, 389.

216. Moore, J. S.; Stupp, S. I. J. Org. Chem. 1990, 55, 3374.

217. Smith, P. A. S. In "Derivatives of Hydrazine and Other Hydronitrogens Having N-N Bonds"; Benjamin/Cummings: Massachusetts, 1983; p43-79.

218. Siddiqui, A. A.; Khan, N. H.; All, M . ; Kidwai, A. R. Syn. Comm. 1977, 7, 71.

219. Blair, J. A.; Gardner, R. J. J. Chem. Soc. (C) 1970, 1714.

220. Girault, Y. ; Decouzon, M.; Azzaro, M. Tet. Lett■ 1976, 1175.

221. Ginos, J. Z .; Cotzias, G. C.; Tolosa, E.; Tang, L. C.; LoMonte, A. J. Med. Chem. 1975, 18, 1194.

222. Borgman, R. J.; Me Phillips, J. J . ; Stitzel, R. E.; Goodman, I. J. J. Med. Chem. 1973, 16, 630.

223. Lott, R. S.; Chauban, V. S.; Stammer, C. H. J. C. S. Chem. Soc. 1970, 495.

224. Felix, A. M.; Heimer, E. P.; Lambros, T. J.; Tzougraki, C.; Meienhofer, J. J. Org. Chem. 1978, 43, 4194.

225. Witiak, D. T.; Tehim, A. K. J. Org. Chem. 1990, 55, 1112.

226. Freifelder, M. In "Catalytic Hydrogenation in Organic Synthesis"; John Wiley & Sons: New York, 1978; p70-72.