INFORMATION TO USERS

This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted.

The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction.

1. The sign or "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If if was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity.

2. When an image on the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will find a good image of the page in the adjacent frame.

3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is customary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary, sectioning is continued again — beginning below the first row and continuing on until complete.

4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced.

5. PLEASE NOTE: Some pages may have indistinct print. Filmed as received.

Xerox University Microfilms 300 North Zeeb Road Ann Arbor, Michigan 48106 77-2493

SALAZAR, Maria Margarita, 1943- SIGNIFICANCE OF THE INTERACTION OF DRUGS WITH MELANIN.

The Ohio State University, Ph.D., 1976 Pharmacology

Xerox University Microfilms, Ann Arbor, Michigan 48106 SIGNIFICANCE OF THE INTERACTION OF DRUGS WITH MELANIN

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By Maria Margarita Salazar, Farmaceutico (B.S. in Pharm.)

*****

The Ohio State University 1976

Reading Committee: Approved By:

Popat N. Patil Dennis R. Feller John W. Nelson

Division of Pharmacology College of Pharmacy I dedicate this dissertation to Dr. Antonio J. Muskus A.,- in recog­ nition for his tireless effort for the professional and personal de­ velopment of individuals; and for the further development of the Venezuelan university. ACKNOWLEDGEMENTS

I wish to express my sincere appreciation to: Dr. Popat N. Patil, my academic adviser, whose guidance, scientific and personal values, moti­ vated the completion of my work for this degree. Dr. Allan M. Burkman, Dr. Dennis R. Feller, and Dr. John W. Nelson, for their advice, constant encouragement and understanding. The Universidad Central de Venezuela for the financial support throughout these years. Dr. Theodore Sokoloski, for his advice and con­ tribution to the analysis of the data. Dr. Kenji Shimada for his help in the initial experimental part of this study.

Dr. Cellini Cozarelli, whose interest in this project made the studies on the human brain possible.

Mrs. Helen Kelly, for her invaluable help and understanding. My family, for leading and supporting me in the pursuit of higher academic achievement.

This investigation was supported in part (materials and animals) by funds from the United States Public Health Service National Institute of Health, National Eye Institute, Grant EY-01308-01Ai. VITA

October 17, 1943 Born, S. Jose Areocuar, Edo. Sucre, Venezuela

1964 ...... Farmaceutico (B.S. in Pharm.), Universidad Central de Venezuela, Caracas, Venezuela

1969-1971, Instructor, School of Veterinary Medicine, Universidad Central de Venezuela, Maracay, Venezuela

1971 - Present Assistant Professor, School of Veterinary Medicine, Universidad Central de Venezuela, Maracay, Venezuela >

1972-1976 Fellow of the Consejo de Desarrollo Cientifico y Humanistico de la Universidad Central de Venezuela, Caracas, Venezuela

PUBLICATIONS Salazar, M . , Shimada, K. and Patil, P. N.: Duration of atropine mydriasis in nonpigmented and pigmented rabbit eye. The Pharmacologist, 1£: 311, 1974. Salazar, M . , Rahwan, R. and Patil, P. N.: Accumulation of 14C -imipramine by the melanin- and the neuromelanin- containing tissues. Abstracts from the VI International Congress of Pharmacology, Helsinki, Finland, 1975.

Salazar, M . , Shimada, K. and Patil, P. N. : Iris pigmentation and atropine mydriasis. J. Pharmacol. Exp. Ther., 197: 79-88, 1976. iv 14 Salazar, M . , Rahwan, R. and Patil, P. N.: Binding of C- imipramine by pigmented and nonpigmented tissues. Eur. J. Pharmacol., 1976, in press. Salazar, M. and Patil, P. N.: An explanation for the long duration of mydriatic effect of atropine in eye. Invest. Ophthal., 1976, in press.

-FIELDS OF STUDY Major Field: Pharmacology

Minor Fields: Biochemistry, Physiology

Professor, Popat N. Patil

v TABLE OF CONTENTS

Page DEDICATION...... ii

ACKNOWLEDGEMENTS...... iii

VITA...... iv LIST OF TABLES ...... viii LIST OF ILLUSTRATIONS...... X

CHAPTER I. INTRODUCTION

Melanins...... 1 The Melanocyte System ...... 5

Melanosomes...... 19 Chemical Composition of Melanosomes ...... 24

Neuromelanin...... 29 Melanin Synthesis ...... 37

Synthetic Melanins...... 46

Free Radical Properties of Melanin...... 50

Binding of Drugs to M e l a n i n ...... 60 Pathophysiological Significance of Natural Melanins ...... 75

Statement of the Problem...... 81 II. METHODS AND MATERIALS

Experimental Animals...... 84 Bovine Iris Melanin Granules...... 84

vi Page

Human Brain...... 84 Experimental Procedures...... 85 Procedure for Calculating Binding Constants. . 106

Drugs...... 108

III. RESULTS 3 In vitro Uptake of H-Atropine by Tissues. . . Ill 3 Accumulation of H-Atropine by Human Irides and Pigment Epithelium ...... 115 3 In vitro Loss of H-Atropine from Tissues after Repeated Washings ...... 115 3 Ocular Penetration of H-Atropine...... 119

Antimuscarinic Effects of Atropine in the Pigmented and Nonpigmented Rabbit Iris and Fundus S t r i p s ...... 119

Recovery from Atropine Blockade...... 124

In vivo Mydriatic Effects of Atropine...... 129 Binding of "^C-Imipramine by Isolated Rabbit Iris ...... -...... 133

In vitro Loss of 14 C-Imipramine from Isolated Rabbit Iris...... 140 Antimuscarinic Effects of Imipramine in the Pigmented and Nonpigmented Rabbit Iris ...... 143 Binding of 14 C-Imipramme to Human Brain Homogenates ...... 143 Localization of 14 C-Imipramine in Isolated Rabbit Iris by Discontinuous Sucrose Density Gradient Centrifugation . 143 3 Localization of H-Chlorpromazine m Isolated Rabbit Retina by Discontinuous Sucrose Density Gradient Centrifugation . 148

vii Page

Localization of 14 C-Imipramine, 3 H-Haloperidol and H-Chlorpromazine in Human Brain by Discontinuous Sucrose Density Gradient Centrifugation......

Binding of 14 C-Imipramme to Bovine Iris Melanin Granules and Synthetic L-Dopa Melanin ...... 3 Binding of H-Chlorpromazme to Bovine Iris Melanin Granules and Synthetic , M e l a n i n s ...... 161 3 . . . Binding of H-Chlorpromazme to Bovine Ins Melanin Granules in the Presence of some Phenothiazines and other Related Drugs...... 175 3 Binding of H-Chlorpromazine to Synthetic L-Dopa Melanin in the Presence of Some Phenothiazines andother Related Drugs . . 179 3 . . . Binding of H-Chlorpromazme to Bovine Ins Melanin Granules Extracted with Organic Solvents ...... 186

IV. DISCUSSION Differential MydriaticEffects of Atropine. . . 196

Binding of 14 C-Imipramme to Pigmented and Nonpigmented Tissues ...... 205

Localization of Labelled Drugs in Human Brain and Rabbit Retina by Discontinuous Sucrose Density Gradient Centrifugation...... 208

Binding of 3 H-Chlorpromazme to Bovine Iris Melanin Granules and Synthetic Melanins. . 211

V. SUMMARY...... 219

viii Page

APPENDICES 3 A. In vitro Loss and Uptake of H- Atropine from Tissues ...... 226 14 B. Binding of C-Imipramine to Isolated Irides from 6 -Hydroxydopamine Treated Albino and Nonalbino Rabbits ...... 234 BIBLIOGRAPHY...... 237

ix LIST OF TABLES

Table Page 1. Comparison of the Properties of Tyrosine Hydroxylase and Tyrosinase...... 6

2. Formation of Melanosomes and Biosynthesis of Melanin ...... 20 3. Ultrastructural, Histochemical, and Spectroscopic Characteristics of Neuromelanin, Cutaneous Melanin and Lipofuscin Granules ...... 34 4. Preparation of Synthetic Melanins ...... 101 3 5. Accumulation of H-Atropine (30 nCi/ml, 10~5M Incubated for 35 min) by Human Iris, Pigment Epithelium and Rabbit I r i s ...... 116

6 . Sensitivity of Isolated Iris Sphincter to Carbachol from Atropine Treated (96 hours before) and Untreated Rabbits .... 132 7. Binding of Radiolabelled Drugs to Human Brain. Localization by Discontinuous Sucrose Density Gradient Centrifugation .... 162

8 . Summary of the Binding Constants of ^H-Chlorpromazine to Bovine Iris Melanin Granules in the Absence and Presence of Other Drugs ...... 170 9. Summary of the Linear Regression Parameters for the Langmuir Treatment of the Binding of ^H-Chlorpromazine to Bovine Iris Melanin Granules in the Absence and Presence of Other Drugs...... 176

x Table Page 10. Summary of the Binding Constants of ^H-Chlorpromazine to Synthetic L-Dopa Melanin in the Absence and Presence of Other Drugs...... 184 11. Summary of the Linear Regression Parameters for the Langmuir Treatment of the Binding of 3 H-Chlorpromazine to Synthetic L-Dopa Melanin in the Absence and Presence of Other Drugs ...... 185

12. Summary of Binding Constants of 3 H-Chlorpromazine to Isolated Bovine Iris Melanin Granules Extracted with Organic Solvents ...... 191 13. Summary of Linear Regression Parameters for the Langmuir Treatment of the Binding of 3 H-Chlorpromazine to Isolated Bovine Iris Melanin Granules Extracted with Organic Solvents ...... 192

14. In vitro, Accumulation of 3 H-Atropine (30 nCi/ml, 10- 5 M, Incubated for 35 min) by Irides from the Rabbits Treated Topically (24 hr before) with 0.1 ml of 4% A t r o p i n e ...... 232

15. Comparative Blocking Effects of Atropine Against a Single Dose EDgQ of Carbachol on the Pigmented and Nonpigmented Iris Obtained from Atropinesterase Positive Rabbits...... 233

16. Effects of 6 -Hydroxydopamine Pretreatment on the Binding of 14c-Imipramine to Isolated Pigmented and Nonpigmented Rabbit Iris ...... 236

xi LIST OF ILLUSTRATIONS

Figure Page

1. Human Melanocyte System...... 9 2. Mechanisms of Hormonal Control of Pigmentation ...... 17 3. Schematic Representation of a Fully Developed Hair Melanosome ...... 25

4. The Raper-Mason Pathway of Melanin Synthesis...... 39

5. The Site of Action of Catechol-O- methyltransferase (COMT) and Hydroxyindole-O-methyltransferase in the Conversion of Tyrosine to Melanin...... 42

6 . Pathway of Phaeomelanin Synthesis...... 44

7. Electron Paramagnetic Spin Resonance Spectra of Natural and Synthetic M e l a n i n s ...... 53

8 . Electron Paramagnetic Spin Resonance Spectra of Human Hair Melanin...... 55 9. Charge-Transfer Interaction Model Reaction for Compounds Having Free Radical Properties ...... 57 10. Tissue Bath Assembly Used for Recording Tension Changes of Isolated Rabbit I r i s ...... 89 3 11. In vitro Accumulation of H-Atropine (30 nCi/ml) by the Pigmented and Nonpigmented Rabbit Iris ...... 112

xii Figure 12. The Rate of Disappearance of ^H-Atropine from Irides and Stomach Fundus Strips Obtained from Albino and Nonalbino Rabbits...... 13. In vivo Rate of Uptake of Radioactivity after the Topical Application of 0.1 ml of 2% Atropine Sulfate Solution, by the Aqueous Humor and Irides of Albino and Nonalbino Rabbits...... 14. Dose-Response Curves of Carbachol Obtained in Isolated Pigmented and Nonpigmented Rabbit Iris and Fundus Strips Determined in the Presence and Absence (Control) of Atropine ......

15. Arunlakshana and Schild (1959) Plot of Log (DR-1) of the Agonist Carbachol Against Atropine Obtained from the Tissues of Albino and Nonalbino Rabbits. . . .

16. In vitro Rate of Recovery from Atropine Blockade by the Pigmented and Nonpigmented Iris Obtained from Atropinesterase- positive Rabbits ...... 17. Onset and Duration of Atropine Mydriasis in Albino and Nonalbino Rabbits......

18. Duration of Atropine Mydriatic Effect After Instillation of 0.1 ml of 2% 3h -Atropine (0.52 yCi total) in Atropinesterase-negative Rabbits ......

19. Thin Layer Chromatography of Iris Extract from Pigmented Atropinesterase- negative Rabbits ...... 14 20. In vitro Accumulation of C-Imipramine by the Pigmented and Nonpigmented Rabbit Iris...... 21. Rate of Disappearance of 14 C-Iraipramme from Isolated Pigmented and Nonpigmented Rabbit I r i s ......

xiii Figure Page 22. Dose-Response Curve of Carbachol Obtained in Isolated Pigmented and Nonpigmented Rabbit Iris Determined in the Presence and Absence (Control) of Imipramine...... 144 23. ^n vitro Accumulation of 14 C-Imipramine by Homogenates of Human Substantia Nigra and Cerebral Cortex...... 146 24. Localization of 14 C-Imipramine in Isolated Pigmented and Nonpigmented Rabbit Iris by Discontinuous Sucrose Density Gradient Centrifugation ...... 149 25. Localization of 3 H-Chlorpromazme . xn Isolated Pigmented and Nonpigmented Rabbit Retina by Discontinuous Sucrose Density Gradient Centrifugation ...... 152

26. Localization of 14 C-Imipramine xn Human Brain by Discontinuous Sucrose Density Gradient Centrifugation ...... 155 3 27. Localization of H-Haloperidol in Human Brain by Discontinuous Sucrose Density Gradient Centrifugation ...... 157 28. Localization of 3 H-Chlorpromazxne xn Human Brain by Discontinuous Sucrose Density Gradient Centrifugation ...... 159 29. Determination of Equilibrium Time for 14c-Imipramine in Bovine Iris Melanin G r a n u l e s ...... 163 30. Binding of 14 C-Imipramxne to Bovxne Iris Melanin Granules and Synthetic L-Dopa Melanin ...... 165 31. Determination of Equilibrium Time for ^H-Chlorpromazine in Bovine Iris Melanin Granules ...... 167 3 32. Binding of H-Chlorpromazine to Synthetic Melanins...... 171

33. Summary of Regression Lines Representing the Binding of ^H-chlorpromazine to Synthetic Melanins ...... 173

xiv Figure Page 34. Summary of Regression Lines Representing the Binding of ^H-Chlorpromazine to Bovine Iris Melanin Granules in the Absence and Presence of Some Phenothiazines and Related Drugs...... 177 35. Binding of ^H-Chlorpromazine to Bovine Iris Melanin Granules in the Absence and Presence of Some Phenothiazines and Related Drugs...... 180 36. Binding of 3 H-Chlorpromazme to Bovine Iris Melanin Granules in the Absence and Presence of Dopamine ...... 182

37. Summary of Regression Lines Representing the Binding of 3 H-Chlorpromazine to Synthetic L-Dopa Melanin in the Absence and Presence of some Phenothiazines and Related Drugs .... 187 3 38. Binding of H-Chlorpromazine to Synthetic L-Dopa Melanin in the Absence and Presence of some Phenothiazines and Related Drugs .... 189 3 39. Binding of H-Chlorpromazine to Bovine Iris Melanin Granules Unextracted and Extracted with Organic Solvents...... 194

40. Schematic Representation which Explains the Effects of Atropine in the Pigmented and the Nonpigmented Eye...... 202 41. Pathway of Melanin Synthesis Showing Different Precursors and Probable Intermediates Formed Corresponding to the Dopachrome of the L-Dopa Melanin Synthesis According to the Raper-Mason P a t h w a y ...... 213 3 42. In vitro Accumulation of H-Atropine by the Pigmented and Nonpigmented Iris of Atropinesterase-positive and Enzyme- unclassified Rabbits ...... 227

xv Figure Page 3 43. Rate of Disappearance of H-Atropine from Iris and Stomach Fundus Strips Obtained from Atropinesterase-positive and Enzyme-unclassified Albino and Nonalbino Rabbits ...... 231

xvi CHAPTER I

INTRODUCTION

Melanins Melanins (yela?=black) are generally defined as

black or brown pigments of high molecular weight formed

by the enzymic oxidation of phenols. Melanins are widely

distributed in animals and plants and usually occur as melanoproteins, the pigment being strongly bound to the protein. Melanin pigments possess definite chemical char­

acteristics, including considerable physical and chemical

stability. Melanins also possess free radical properties (Harley-Mason, 1965; Mason, 1953; Seiji, 1967; Swartz,

1972; Thomson, 1974). Since numerous brown and black pigments such as uveal melanin, neuromelanin, rheomelanins and lipofuscin occur in tissues it is necessary to distinguish between melanins and other pigments. For example, melanin pig­ ments resulting from oxidation of catechols in cells of the adrenal medulla could be different from the melanin

found in skin and eyes.

1 2 The term eumelanin (e\3=good) covers brown to

black pigment found in skin, hair, feathers, insect

cuticles and melanomas. Phaeomelanin ty>ai6 ?= dusky)

covers a broad spectrum from violet to yellow. These pigments have been found only in feathers and mammalian hair, including human red hair. Both types of melanins

are formed in melanocytes from oxidation of tyrosine by tyrosinase, but they differ in that phaeomelanins synthe­

sis includes the incorporation of a molecule of cysteine.

Phaeomelanins and eumelanins may occur together and are responsible for most mammalian coloring (Thomson, 1974). Eumelanins are irregular, crosslinked polymers, containing 5,6 -dihydroxyindole units that are conjugated with proteins. Rheomelanins are soluble pigments of different colors which are naturally formed in the body from pre­ cursors present in the body as for example pigments formed in the cells of adrenal medulla from the oxidation of epinephrine and norepinephrine (Seiji, 1967). Formation of rheomelanins has also been demonstrated in human blood plasma in vitro from: (-)-epinephrine, (-)-norepinephrine,

(-)-epinephrine methyl ether, L-dopa, dopamine and L-a- methyldopa (Hegedus and Altschule, 1970) .

Neuromelanin is a pigment present in neurons of some areas of the central nervous system. It is believed to be of different origin than that of other melanins.

Unfortunately, investigators working in the field of melanins have not used a consistent terminology. The

following terms were proposed by Fitzpatrick et ajL. (1966) to define melanin containing cells, their precursors and related cells. Ontogenetic stages in the formation of melanin are also included. Melanocyte (Greek: Melan = black; Kytos=cell): a cell which synthesizes a specialized melanin-containing organelle, the melanosome. This term includes differen­ tiated cells which synthesize nonmelanized or partially melanized premelanosomes as terminal products. It is suggested that in albinism the melanocytes containing nonmelanized premelanosomes be called albino melanocytes.

Melanophore (Greek: Melan= black; Phore= bearing): a type of melanocyte which participates with other chromato- phores in the rapid changes of color in animals by intra­ cellular displacing (aggregation and dispersion) of melano­ somes . In human histology and pathology, a phagocytic cell that contains melanin but does not form the pigment (Stedman's Medical Dictionary). It should not be referred to as melanophore.

Melanoblast (Greek: Melan= black; Blastos= germ): a cell which serves at all stages of the life cycle as the precursor of the melanocyte (and/or melanophore). 4

Langerhans Cell; A distinctive cell of the mammalian epidermis and dermis presumed to belong to the

"melanocyte" series. It is revealed by gold impregnation and contains distinctive nonmelanized disc-like organelles. Premelanosome; All distinctive particulate stages in the maturation of melanosomes. Hence electron density is variable. It possesses an active tyrosinase system after the onset of melanin synthesis. ^-.Melanosome (Greek: Melan=black; Some=body) : A discrete melanin containing organelle in which melaniza- tion is complete; shown to be more or less electron-dense by electron microscopy; tyrosinase activity not usually demonstrable. Multiple melanosomes imbedded in supporting matrices/ for example, as in the macrophages and malpig- hian cells of mammals may be designated as melanosome complex. Accordingly, the terminology suggested by Fitz­ patrick et al. (1966) will be used in this paper. However, since the terms melanin granules and melanosomes are commonly used as synonymous, for simplicity these terms will be used interchangeably in the text. Tyrosinase (Mushroom, 1.10.3.1) also called poly­ phenol oxidase, o-diphenoloxidase and catecholase is a copper-containing enzyme which is localized in melanocytes and is responsible for the formation of melanin pigment. 5

It catalyzes two successive reactions.

Tyrosine ---- *• Dopa ---- ► Dopaquinone

Dopaquinone is polymerized to melanin nonenzymatically.

L-dopa is the best substrate for the enzyme. Tyrosinase is also responsible for the synthesis of catecholamines in insects (Sekeris and Karlson, 1966) and in the banana plant (Nagatsu et al., 1972). Before the discovery of tyrosine hydroxylase, tyrosinase was thought to be responsible for the synthesis of dopa from tyrosine in the biosynthesis of norepinephrine.

A summary of the characteristics of both enzymes is pre­ sented in table 1 .

The Melanocyte System

In mammals the melanocyte systems have different origins. The melanocytes of skin and hair are derived from melanoblasts which migrate from the embryonic neural crest early in development (Rawles, 1947). Melanocytes in the eye originate both from the optic cup (retinal melano­ cytes) and the neural crest (uveal melanocytes). There are indications that a small number of primary melano­ blasts from the neural crest undergo proliferation and differentiation to color local regions of the skin (Mintz,

1967). The differentiation of melanoblasts into melano­ cytes starts with the synthesis of melanosomes. This TABLE 1

COMPARISON OF THE PROPERTIES OF TYROSINE HYDROXYLASE AND TYROSINASE9

Tyrosine hydroxylase Tyrosinase

Distribution Organs Adrenal medulla, brain, sympatheti­ Skin, eye, melanoma cally innervated organs (heart, spleen, salivary glands, vas deferens, etc.)

Cell type Sympathetic nerve cell Melanocyte

Cofactor Tetrahydropterin (tetrahydrobiop- Dopa terin?), Fe++

Optimum pH 6.0 (5.8-6.3) 7.0

Substrate L-tyrosine ++ L-tyrosine + L-phenylalanine ++ D-tyrosine + Tyramine + Dopa +++

Km (M) L-tyrosine (DMPH4) 1 X 10-4 L-tyrosine 6 X L-phenylalanine (DMPH4) 3 X 10-4 D-tyrosine 8 X 10“310_3 DMPH4 5 X 10-4 L-dopa 5 X 10-4 D-dopa 4 X 10"3 L-tyrosine (tetrahydrobiopterin) 2 X 10“5 Tetrahydrobiopterin 1 X 10-4

Inhibitors L-a-methyl-p-tyrosine +++ 0 Dopa or other catechols + 0 Fe++ chelating agents, such as a,u',dipyridyl ++ 0 Cu-chelating agent, such as diethyldithiocarbamate 0 +++ Melanin pigment formation 0 +

aModified from Nagatsu (1973) cellular transformation is influenced by the genotype of the melanoblast which sets the minimal environment require­ ments necessary for differentiation as well as by the

adequacy of the local tissue environment which may be subject to change during development (Quevedo, 1971).

Melanocyte Characteristics Thin section electron microscopy studies show that melanocytes appear as distinctive cells. They are den­ dritic in nature possessing two or more dendritic processes,

there is an absence of desmosomes and they contain some filaments. Generally, melanocytes are located above but in close proximity to the basement membrane. The nucleus is round, slightly indented and has a double nuclear mem­ brane. The outer membrane is rough and it is separated O from the inner membrane by a clear zone approximately 500A wide. A nucleolus is also present and lacks a limiting membrane. Mitochondria are abundant and are dispersed throughout the cytoplasm. A well developed endoplasmic reticulum is located between the mitochondria and through­ out the cytoplasm of the melanocyte with both rough and smooth surfaced types present. The rest of the cytoplasm is clear with a few electron dense particles dispersed throughout. The Golgi complex is easily recognized, is well developed and seems to be important in the formation of melanin granules (see below) (Zelickson, 1967). 8

The dendritic cells of human and guinea-pig epider­ mis have been divided into three types: type I contains melanin granules in different stages of maturity; these cells are true melanocytes; type II does not contain melanin granules but possesses characteristic rod-shape granules (Langerhans cell) and type III does not contain any granules. The functional significance of the third type of dendritic cells is unknown (Zelickson, 1967). Melanocytes are found in the skin, eyes, inner ear and leptomeninges. Melanocytes in the skin are located at the epidermal-dermal junction. In the eyes they are present in the uveal tract, and in the central nervous system the cells are found at the junction of the pia and the arachnoid (fig. 1). Under ordinary conditions melanocytes probably multiply at a slow rate. This belief seems probable because when these cells are destroyed by physical or chemical means permanent depigmentation results. When these cells undergo neoplastic change to become malignant melanoma they multiply very rapidly (Lerner and Fitzpatrick,

1953). Several facts are in favor of the neural crest origin of the melanocytes: a) embryological studies in amphibia, birds, mice and human beings show that melano­ cytes are derived from the neural crest region; b) the 9

Figure 1. Human Melanocyte System (Modified from Brodthagen, 1972) EYE Uveal tract CNS Retina Leptomeninges

INNER EAR 0

HUMAN MELANOCYTE SYSTEM

MUCOUS MEMBRANE Epidermal —Dermal junction

SKIN Epidermal—Dermal junction H air Bulb Dermis

Figure 1

i 1! cells are dendritic in shape as are other nerve cells; c) the cells can be stained with silver, gold and methylene blue as can other nerve cells; these cells produce dopa which makes them related in function to the sympathetic nervous system cells producing norepinephrine; e) malignant melanomas (melanocyte tumors) do not respond to radiation therapy, resembling in this way most nerve cell tumors and f) abnormalities of pigmentation are frequently associated with abnormalities of the nervous system (neurofibromatosis)

(Rawles, 1947). Interestingly, although melanocytes have not been found in brain nuclei where the pigment neuromelanin is formed, an association between the pigmented areas and the presence of adrenergic system pathways have been found

(Bazelon et al_., 1967) . The interaction between melanocytes (also called pigment cells) and keratinocytes from the epidermis form a structural and functional unit called the epidermal melanin unit. This unit appears to be involved in several functions such as: formation of melanosomes, movement of melanosomes out into the melanocyte dendrites, attachment of melanocytes to keratinocytes and movement and break­ down of the melanosome complex within the keratinocyte

(Wiskwo and Szabo, 1973). 12

Studies in frog skin, using cytochalasin B,

suggest that a contractile microfilament system is in­

volved in the movement of melanin granules in the mel- anophore (Malawista, 1971). McGuire and Moellman (1972) demonstrated in dermal and epidermal melanocytes of Rana

pipiens that cytochalasin B prevents the dispersion of pigment granules by melanocyte stimulating hormone (MSH)

and causes centripetal movement of pigment granules that have been dispersed by MSH. Microfilaments are abundant

in the dendrites of epidermal melanocytes in which pig­ ment granules have been dispersed by MSH and also, the number of microfilaments is reduced in the dendrites of melanocytes lightened with cytochalasin B. McGuire and Moellman (1972) propose that dis­ persion of pigment granules is effected by way of the micro­

filaments; when microfilaments are destroyed, pigment gran­ ules move centripetally. Intact microtubules are required for pigment granule aggregation. More recently, Jifnbow and Fitzpatrick (1975), studied the activity of cytoplasmic filaments in human melanocytes after UV irradiation. They O found that human melanocytes in vivo contain 100 A in dia­ meter filaments. The filaments are primarily located in the 13 O perinuclear area and endoplasmic regions. These 100-A microfilaments appear to be involved in the elongation of the dendrites as well as in the movement and transfer of melanosomes. The melanocytic filaments appear also to be involved in the transfer of melanosomes from the melano­ cytes to the epithelial cells since there is an increase in the number of melanosomes transferred to the epithelial O cells during the rapid translocation of the 100-A filaments. These authors consider that microtubules are directly involved in the melanosome movement, although they may be involved in the elongation of the dendrites. The process of translocation, dispersion and aggregation, of melanosomes has been thoroughly studied in lower vertebrates such as fish, amphibians and reptiles.

In these animals the pigment cells are called melanophores.

Like melanocytes, the melanophores are also of neural origin. Jacobowitz and Laties (1968) observed direct innervation of the pigment cell where they found catechol­ amine containing fibers in anatomic proximity to conjunc­ tival and dermal melanocytes of a Teleost fish. In fish having directly innervated pigmented cells, electrical stimulation of a nerve to the skin produces localized blanching and cutting of the nerves produces immediate darkening (Lerner, 1971). 14

It must be pointed out, however, that melanophores in Teleost fish are substantially different from melano­ phores in mammals. Studies in normal mammalian skin (Falck and Rorsman, 1963), and of pigmented nevi and mela­ nomas (Falck et al., 1965), by histofluoremetric technique failed to demonstrate adrenergic innervation to melanocytes. Goldman and Hadley (1969), Gupta and Bhide (1967) and McGuire (1970), among others, demonstrated that melano­ cytes of all the animals they studied— fish, frogs, toads and lizards— have alpha- and/or beta-adrenergic receptors.

Also it has been found that stimulation of alpha-receptors leads to aggregation of melanosomes and skin lightening. Stimulation of beta-receptors produces dispersion of melanosomes, therefore producing darkening of the skin. Although some studies on the role of acetylcholine in the control of pigmentation in lower vertebrates have been done, the results have been inconsistent; acetyl­ choline darkens some fish but lightens about one of three Rana pipiens (Lerner, 1971). McGuire (1970) reported that epinephrine which has both alpha- and beta-adrenergic effects lightens skin previously darkened with MSH, and darkens lightened frog skin. On the other hand isoproter­ enol, a beta-receptor stimulant, produces only granule dispersion (darkening) whereas phenylephrine, an alpha- adrenergic agonist, causes only granule aggregation

(lightening). 15 Melatonin produces lightening of Rana pipiens dermal melanocytes in concentrations less than those re­ quired for epinephrine or acetylcholine. Alpha- and beta- adrenergic blockers do not prevent the effects of melatonin.

Melatonin does not have a direct effect on mammalian melanocytes. However, the possibility that melatonin may be acting centrally in the rat to prevent release of MSH from the pituitary has been suggested (Kastin and Schally,

1967). Novales and Davis (1967) reported darkening of frog skin and melanin dispersion in salamander cultured melanophores by cyclic-AMP. Also, caffeine, a phosphodi­ esterase inhibitor, caused dispersion of pigment granules. Abe et al. (1969 a,b) found that in skin from Rana pipiens 8-MSH, a -MSH, and ACTH increased cyclic AMP levels with the same potencies at which they stimulated darkening. On the other hand norepinephrine and melatonin inhibited the MSH-induced increase in cyclic AMP. The dispersion and aggregation of melanosomes which occurs through relatively fixed channels in pigment cells of fish, amphibians, and reptiles in response to MSH, caffeine, catecholamines, acetylcholine and melatonin have not been shown in mammalian system. Nevertheless, MSH has a pronounced darkening effect on the skin or hair of man, guinea pigs, hamsters and mice. There is evidence that darkening is associated with an increase in tyrosinase 16 activity which is most likely due to an increase in synthe­ sis and not to a simple activation of tyrosinase already present in the cell. The role of cyclic AMP in the induc­ tion of tyrosinase is still unknown, but since MSH in­ creases cyclic AMP in melanomas it is possible that the nucleotide is responsible for initiating tyrosinase synthe­ sis, although it is not known if increased tyrosinase synthesis leads to dispersion of melanosomes (Lerner,

1971). See fig. 2. In mammalian pigment cells or melanocytes MSH in­ creases cyclic AMP, induces tyrosinase synthesis and darkens skin. These steps may be related sequentially. Alpha- and beta-receptor sites have not been found in mammalian cells, but it is likely that the melanocytes are under neural control (Lerner, 1971). Ehinger and Falck (1970) suggest that melanocytes of rat iris possess an adrenergic as well as a cholinergic innervation. This concept of an adrenergic innervation is supported by the fact that heterochromia at times accom­ panies Horner's Syndrome in humans (Duke-Elder and Wyber, 1958), that has been observed in cats sympathectomized a few years previously and that white spots appear in the iris of pigmented rabbits a few months after sympathectomy

(Ehinger, Falck and Rosengren, 1969). Inhibition of pig­ mentation by sectioning of adrenergic nerves has been demonstrated (Bennet and Hausberger, 1938). Although a Figure 2. Mechanisms of hormonal control of pigmentation. A. MELANIN DISPERSION: ACTH, a- and 3-MSH and caffeine raise the intracellular levels of cyclic AMP in melanophores of amphibians, fish and reptiles to initiate dispersion of melanosomes.

B. MELANIN SYNTHESIS:. Injection of MSH into man, hamsters, mice and guinea-pigs increase tyrosinase activity, which in turn results in melanin formation. It is likely that these events are triggered by MSH, which raises cellular levels of cyclic AMP.

(From Lerner, 1971) 18 (A) MELANIN DISPERSION

cyclic A M P

(B) MELANIN SYNTHESIS

F ig u re 2 19 rapid mechanism for color changing is absent in mammals the depigmentation occurring after sympathectomy suggests

some kind of slowly acting trophic nervous influence on the

melanocytes.

Melanosomes Melanosomes are the final product of the synthesis

of melanin in the melanocyte (Fitzpatrick et al., 1966). In order to understand the origin of melanin it is neces­ sary to understand the process of formation of the melano-

some which is illustrated in table 2. Toda and Fitzpatrick (1971) outlined four stages in the development of the melanosomes. Stage I is characterized as a spherical

membrane limited vesicle that contains tyrosinase and at times some filaments but not melanin. In stage II (pre-

melanosomes), the organelle is oval and shows numerous membranous filaments, with or without cross linking, having

distinct periodicity. In stage III, the internal structure of the stage II melanosome becomes partly obscured by the

deposition of melanin. In stage IV, the oval organelle is electron opaque without a discernible structure. It is speculated that the earliest form of pre- melanosones in human black hair are large spherical vacu­

oles. The inner structure of these vacuoles consists of amorphous structures of protein matrices which transform into a filamentous or lamellar structure with a regular TABLE 2

FORMATION OF MELANOSOMES AND BIOSYNTHESIS OF MELANIN3

Nomenclature Organelle Morphology Biochemical Composition Probable Events

Ribosome RNA + Protein Biosynthesis of tyrosinase 100-150 A and formation of subunits containing tyrosinase D Golgi Complex 0.05 y Phospholipid + Protein Assembly of subunits and formation of unit membrane

Intermediate Vesicle © Phospholipid + Protein Stage I 0.5 y

Cross-linking of subunits Stage II and arrangement in a final characteristic structural Phospholipid + Protein form. Final product in Premelano some 0.7 X 0.3y + albino melanocyte Tyrosinase + Melanin Beginning of melanin formation Stage III 0.7 X 0.3 u

Melanosome / Melanin without measurable Melanin without measurable Stage IV 0.7 X 0.3 y tyrosinase activity tyrosinase activity. Final product of melanocyte

Typical values for human brown melanin granules a to From Seiji, 1967; Witkop, 1971 o striation pattern. The specific topographic relation between the large vacuoles and Golgi apparatus suggests that premelanosomes may be evolved from the Golgi apparatus. However it is not certain whether they arise from the en­ largement of small vesicles that pinch off from the cis- ternae of Golgi system or are contained and develop within a tubular smooth endoplasmic reticulum which is connected with the Golgi apparatus during melanogenesis (Jimbow and

Kukita, 1971). In studying the relationship of the Golgi apparatus to the formation of melanosomes in human cellular blue nevus, Allegra (1974) suggests that the process of melano­ genesis varies in relation to the maturity and differentia­ tion of the melanocytes in which it occurs. Thus, small melanosomes incapable of maturing into melanin granules are prevalently seen in embryonal melanocytes in which, characteristically, the Golgi apparatus is not present; whereas, the adult melanocyte contains both Golgi areas and maturing melanin. With the maturation process, the spherical vacuoles of the earliest premelanosomes become elongated and the inner structures arrange in a regular fashion. These morphological sequences are similar in various melanocytes of different origins (Jimbow and Kukita, 1971). There is no agreement about the morphology of the inner structure of premelanosomes from different origins. Jimbow and Kukita (1971) demonstrated that in human black

hairs however, the longitudinal sectioning of premelano­ somes shows two types of inner structure: a filamentous

and a lamellar structure. Both structures present the same striation pattern. The filamentous structures gener­

ally run parallel to the long axis; the lamellar ones are generally present in the center. They then join at the poles forming an oval shaped structure. In cross section

the inner structures are observed as concentrically

arranged dark lines. The nature of the striation patterns along the lamellae is interpreted as representing the alignment of macromolecules of proteins (Jimbow and Kukita, 1971).

Seiji (1967) suggests that the protein molecules consist not only of tyrosinase but also various proteins including other enzymes and constitutional protein moieties. Melanin synthesis begins with the deposition of granular substances on the protein matrix of the lamellar structure. As the process of melanization continues, the lamellar structures thicken and become electron dense.

Melanin deposition seems to continue until the spaces between the inner structures become filled. Melanin granules, or fully developed melanosomes, are generally described as electron-dense, amorphous bodies. However, when prepared under cytochemical conditions, melanosomes 23 appear to have a characteristic inner structure (Drochmans,

1967). In their study on human black hair, Jimbow and O Kukita (1971) using ultrasectioning (300-400A in thickness) , have found that two components can be distinguished in the fully developed melanosomes: the dense cortical shell or region, and the less dense central core. The central core consists of regularly folded lamellar structures disposed in the amorphous matrix. The cortical shell is present below the outer membrane of melanosomes and encloses the central core. It is composed of fine granular osmiophilic substances. In addition to those two structures these investi­ gators also found round electron-lucent structures. Serial sectioning suggests that they are globular shaped and O measure about 400A in diameter. Most of them appear to be attached to the outer layer of the central core and are protruding into the cortical shell. These globular bodies are also found in the melanosomes passed into keratino­ cytes and very rarely in premelanosomes (Jimbow and Kukita,

1971). Mottaz and Zelickson (1969) reported that some melanosomes of very dark brown hair display some small amount of a "moth eaten" effect which is absent in lighter brown hairs. This effect and the globular bodies observed by Jimbow and Kukita (1971) seem to be identical. The 24

nature and significance of these structures needs to be elucidated. A schematic representation of a fully devel­ oped melanosome is presented in figure 3.

Chemical Composition of Melanosomes

At the present time, the information available on the chemical composition of melanosomes, or melanin granules, is very limited. This is probably due to the intrinsic characteristics of melanins. Because of their insolubility in most solvents melanins do not permit a detailed study of the chemical nature of the melanin granules, unless it is done under very drastic conditions which may jeopardize the validity of the experimental results obtained under such conditions. The presence of succinic dehydrogenase and cyto­ chrome oxidase has been demonstrated in both mitochondria and melanin granules. Stein (1955) studied the chemical composition of melanin granules obtained from ox choroid and compared it to that of mitochondria. He found that melanin granules contain far fewer proteins than mito­ chondria and the content of iron was much higher in melanin granules than in mitochondria. Higher content of zinc

(Stein, 1955; Seiji et al. 1963) and of copper (Stein, 1955) in melanin granules of ox choroid and Harding-Passey and B-16 mouse melanoma have been found in comparison to the metal content of mitochondria. Tyrosinase was Figure 3. Schematic Representation of a fully developed hair melanosome. Note the dispos tion of electronlucent, globular bodies. (From Jimbow and Kukita, 1971). 26

gill

F ig u r* 3 27 localized mostly in melanosomes and not in mitochondria. No tyrosinase activity could be demonstrated in the mito­ chondrial fraction of mouse melanoma. These results demon­ strate than melanin granules are different from mitochon­ dria. Takahashi and Fitzpatrick (1966) reported the pres­ ence of large amounts of dopa in the melanosomal fraction of Harding-Passey melanoma. In B-16 mouse melanoma, the content of dopa was very low. This suggests genetic dif­ ferences in the melanosomal proteins of Harding-Passey and B-16 melanoma which could account for the characteristic difference in color of these tumours (Takahashi and

Fitzpatrick, 1966). Recently, the chemical compositions of hair melano­ somes has been elucidated. Borovansky and Duchon (1974) analyzed the chemical composition of human hair melano­ somes; Hall and Wolfram (1975) reported studies on poodle and human hair, and hair and squid melanins. Hair melano­ somes were highly melanized with a content of approximately

60% melanin in contrast to 30% for melanomas. In both studies (Borovansky and Duchon, 1974; Hall and Wolfram,

1975), hair melanosomes contained all 18 amino acids, although the content of arginine, glycine and proline was higher than that in melanoma melanosomes; however,, dopa was not detected (Borovansky and Duchon, 1974). Hall and 28

Wolfram (1975)also reports that the amino-acid composition of squid melanin protein is similar to the other melanin containing tissues. On the basis of these results they

suggest the existence of a specific protein in melanin containing tissues. This hypothesis is considered a spec­ ulation at the present time. Duchon et al. (1973) completed a chemical analysis of melanosomes isolated from centrifugation of: Harding-

Passey, B-16, Cloudman S91 and "Stanford" mouse melanoma homogenates, human and horse melanoma homogenates, chicken embryos' retinal pigment epithelium and ox choroid homo­ genates , the squid Loligo pealii ink sac and the Sepia Officinalis ink sac homogenates. Using acid hydrolysis (6N HC1 for 24 hours) they found that melanin forms 18-72% of the total weight of the melanosomes, with respect to their source. The amount of protein recovered varied from

5-61% expressed as the sum of all amino acids recovered. The; amino acid analysis revealed that the protein component of the ten different melanosomes studied consist of all known amino acids present in proteins. Also, 0.2-0.8% of 3,4-dihydroxyphenylanine (dopa) was found in most, but not all, of the melanosomes. Taurine (2-amino-ethane-l-sul- phonic acid) was found in Loligo pelaii (squid) and Sepia officinalis melanosomes hydrolysate. 29 The chemical composition of melanosomes seems to

vary according to species. Duchon et al. (1973) also found

that melanosomes from Loligo pelaii (squid) and Sepia officinalis have a very much higher content of alanine and much lower content of lysine than the other melanosomes examined. Ox choroid and "Stanford" mouse melanoma melanosomes were found to have a very high content of glycine and proline and very low content of leucine and valine. The level of lysine seems to be inversely propor­ tional to the content of melanin. According to these authors, the chemical composi­ tion of melanosomes may reflect genetic and species differ­ ences in the composition of melanosomal proteins in the course of phylogenesis, organogenesis and ontogenesis.

Neuromelanin

In man neuromelanin is sharply localized in three areas of the brain: The substantia nigra, locus ceruleus, and the dorsal nucleus of the vagus nerve, Adler (1939). The greatest quantities of melanin in human brain are found in the substantia nigra where most of the pigmented neurons are in the pars compacta. Less pigment is present in nearly all cells of the locus ceruleus but only a small fraction of the cells in the dorsal motor nucleus of the vagus are pigmented. ... 30

Substantia nigra is made up of medium size multi­ polar neurons which in the adult human brain contain sufficient melanin pigment to give the nuclear group its characteristic black appearance and the name. The pigment lies within cells in the living brain but it is often extracellular in postmortem or pathological material. It is also present in albino humans (Foley and Baxter, 1956). The relationship of this pigment to' the function of the substantia nigra is not well understood. The substantia nigra extends throughout the whole length of the midbrain from the upper border of the pons fibers into the caudal end of the diencephalon. Bazelon et al. (196 7) in serial sections from an adult human brain have demonstrated pigmented neurons in the substantia nigra and its associated nuclei as well as in a column extending the entire length of brainstem from the mesen­ cephalon to the nucleus retroambigualis. Histologically the pigment was characterized as melanin. Substantia nigra is regarded as being functionally related to the basal ganglia and, through the basal ganglia, to the cerebral cortex as part of the arc important in regulating tonus and in stabilizing voluntary movements.

It is significant that certain degenerative involvement of the basal ganglia such as occurs in Parkinson's syndrome is frequently accompanied by degeneration of the neurons in the substantia nigra (Crosby et al., 1962). 31

The pigment granules are usually dispersed through­

out the cytoplasm of the cell and extend into the axon hillock and initial portions of the axon and less commonly

in the dendrites. Melanin granules have not been des­ cribed in neuronal processes remote from cell bodies nor in synaptic endings and are not apparent in glial tissues in normal brains in man (Marsden, 1969). There are no reports of nuclear distribution of

neuronal melanin in the brains of non-primate mammals.

However, pigmentation in certain species such as the

gorilla (Adler, 1942), cat and dog (Brown, 1943), and

horse (Gillilan, 1943) has been reported. Within the pri­ mates there is a progressive increase in the intensity of pigmentation as the phylogenetic relation to man becomes

closer and the maximum intensity is seen in man. It also appears that the intensity of pigmentation is related to the degree of evolution of the brain (Marsden, 1969). Pigmented ganglion cells have been observed through­ out the amphibian brain and there is an apparent increase in melanin nerve cells of some amphibians at metamorphosis (Adler, 1939). The age at which pigment first appears in human neurons is somewhat controversial. Mann and Yates (1974) reported similar results to those obtained by Foley and

Baxter (1958) who suggested that most cells of the locus ceruleus become pigmented at or shortly after birth and 32

are regularly pigmented at 18 months of age. Nigral neurons are rarely pigmented at birth and at 18 months of

age are only slightly pigmented. In both substantia nigra and locus ceruleus, therefore, the onset of pigmentation appears to be associated with the beginning of regular

activity within each nerve cell type. The intensity of pigmentation increases in both the locus ceruleus and substantia nigra until adolescence. Both nuclei have

been said to be equally and maximally pigmented (Foley and Baxter, 1958; Fenichel and Bazelon, 1968) at a level which

remains fairly constant throughout adult life. Moses et

al. (1966) noted that the pigment content and character in

the substantia nigra and locus ceruleus of brains at

20-30 years was comparable to brains of 55-85 year-old

patients. Mann and Yates (1974) report that from birth to

60 years of age the mean degree of melanin pigmentation of both the substantia nigra and locus ceruleus is propor­ tional to the age of the person, although the amount of melanin present in the individual cells of a group at a

given age may vary considerably. The linearity of the relationship indicates that melanogenesis is not a random but rather a constant feature of those cells. Melanin granules first appear in the cytoplasm

adjacent to the nucleus (Foley and Baxter, 1956; 1958). 33

At this site there is usually an indentation and condensa­ tion of the nuclear membrane. There is no suggestion that the initial site of melanin deposition is related to the

Golgi apparatus, in contrast to the melanocyte system where the association is clearly established. The ultrastructure of human neuromelanin in the substantia nigra and locus ceruleus have been reported by D'Agostino and Luse (1964) , Duffy and Tennyson (1965) and Moses et al. (1966). Neuromelanin granules are identified by their high electron density and are composed of: a lipid globule, a finely granular matrix of medium density, and an electron-dense coarsely granular material (Marsden,

1969) . Similarities and differences between neuromelanin, lipofuscin and cutaneous melanin have been described

(D'Agostino and Luse, 1964, Duffy and Tennyson, 1965,

Moses et al., 1966, Van Woert and Ambani, 1974). A summary of the ultrastructure, histochemistry and spectroscopic characteristics of neuromelanin granules, cutaneous melanin granules and lipofuscin is presented in table 3. Van Woert et al. (1967) compared spectroscopically the pigment isolated from human substantia nigra, dopa melanin, melanin-like pigments prepared by the incubation of monoamineoxidase with various neurotropic amines. Also, a visceral melanin from the liver of an amphibian TABLE 3

ULTRASTRUCTURAL, HISTOCHEMICAL AND SPECTROSCOPIC CHARACTERISTICS OF NEUROMELANIN, CUTANEOUS MELANIN AND LIPOFUSCIN GRANULESa

Property Neuromelanin Cutaneous Lipofuscin Granule Melanin Granule Granule

Ultrastructure Acid phosphatase Acid phosphatase Acid phosphatase Small lipid globule No lipid globule Large lipid globule High and low elec­ High electron den­ Low electron density tron density sity melanin melanin melanin

Histochemistry Lipid stains- Lipid stains- Lipid stains+ PAS- PAS- PAS+ Acid fast- Acid fast- Acid fast+ H 2°2+ H2°2+ H 2°2“ AgNO 2 + AgNO3++ AgNO3- Thionine-yellow Thionine-green Nile blue-blue Pyrroles- Pyrrolest Nile blue-green UV fluorescence - - +

From Van Woert and Ambani, 1974 35

(Amphiuma tridactylum, Amphiuma: amphibia genus including

only congo snakes) and lipofuscin from human cardiac muscle were examined and compared. They demonstrated that

neuromelanin is different from melanins found in other parts of the body. However the substantia nigra pigment

is a melanin. Lipofuscin also contains a melanin compon­ ent, but unlike substantia nigra granules it also contains

a large fluorescent lipid fraction. The melanin component

may explain the electron microscopic similarities of this pigment to substantia nigra (Duffy and Tennyson, 1965;

Moses et al., 1966). In contrast to the synthesis of cutaneous melanin, practically nothing is known about the synthesis of neuro­ melanin. Since cells of substantia nigra are rich in

dopamine (Anden et al., 1965) and dopamine is oxidized to

form melanin both by auto-oxidation and by several differ­ ent enzymes, the amine has been suggested to be a precursor of neuromelanin (Van Woert and Ambani, 1974). In experiments with human substantia nigra, dopa and dopamine were recovered in the hydrolysate of neuro­ melanin (Nordgren et al., 1971). Since after hydrolyzing three times, 7 hours each time, no dopamine was found in

the hydrolisate, the authors suggest that the primary nucleus of neuromelanin formed is dopa melanin, while the dopamine melanin forms a shell on the melanin particles. 36

With the aid of infrared spectrometry, Maeda and Wegman (1969) compared brain melanin to melanin synthe­ sized from dopa, dopamine, norepinephrine and epinephrine.

The studies point to norepinephrine as a possible pre­ cursor of neuromelanin. Van der Wende and Sporlein (1963) demonstrated the oxidation of dopamine to a melanin by albino rat brain in vitro. Rodgers and Curzon (1975) obtained melanin formation in brain homogenates using dopamine, dopa, norepinephrine, epinephrine, and 5-0H- tryptamine as precursors. No melanin formation was ob­ served using ^C-tyrosine. They also found that melanin formation was highest in the substantia nigra. Van Woert and Ambani (1974) and Rodgers and

Curzon (1975) were unable to demonstrate tyrosinase activ­ ity in human substantia nigra. On the other hand mono- amineoxidase (MAO) is suggested as the enzyme responsible for the synthesis of neuromelanin (Van Woert et al., 1967). Evidence against a major role of MAO in neuro­ melanin formation has been presented by Rodgers and 14 Curzon (1975). The authors used C-tranylcypromine, a MAO inhibitor, and in vitro conditions for the argument. Evidence in favor of peroxidase as the enzyme responsible for neuromelanin synthesis has been presented by Van Woert and Ambani (1974), who found the highest peroxidase activity in substantia nigra and the lowest 37 activity in the temporal cortex of human brain. Peroxidase activities in the substantia nigra and caudate nucleus are reduced in brains from Parkinsonian patients. On the other hand, Rodgers and Curzon (1975) could not obtain melanin from tyrosine and peroxidase, nor could they in­ crease melanin formation from dopamine when homogenates from substantia nigra and pons were used. The same authors suggest a non-enzymic process of melanin formation.

They indicate however, that brain may contain an enzyme which loses its activity between death and autopsy. On the other hand, tissue obtained 2.5 hours after death did not show more activity for melanin formation than tissue obtained 3 days after death. It is clear from the evidence presented so far that both the precursor and enzyme responsible for the synthesis of neuromelanin remain to be established.

Melanin Synthesis

Melanin synthesis takes place in the melanocyte, specifically in the premelanosome. Tyrosine is the pre­ cursor, tyrosinase being the enzyme catalyzing the first two steps of the reactions, which take place in the presence of molecular oxygen (Lerner and Fitzpatrick,

1950). Tyrosinase is inhibited by copper chelating agents such as diethyldithiocarbamate (table 1). Compounds such as 8-methoxypsoralen increase melanin formation 38 presumably by increasing the activity of the enzyme

(Lerner and Fitzpatrick, 1953). Fig. 4 shows the most accepted melanin synthesis pathway. The steps in this pathway are: a) in the presence of tyrosinase and molecular oxygen, tyrosine is oxidized to dopa. This reaction is usually slow at onset, but after an induction period it proceeds very rapidly.

This is an irreversible reaction; b) dopa is enzymatically oxidized by a reversible reaction to dopaquinone(dopaquin- one thus formed may be reduced to dopa when a reducing agent such as ascorbic acid is present in the reaction). Further stages of the reaction proceed relatively rapidly in the absence of the enzyme although the reaction rates are increased in the presence of the enzyme; c) dopaquin­ one undergoes a spontaneous irreversible and rapid intra­ molecular change in which the nitrogen of the side chain attaches itself to the 6-position of the benzene nucleus with the formation of 5,6-dihydroxyindole-2-carboxylic acid (leucodopachrome); d) leucodopachrome is readily oxidized by a reversible reaction to the corresponding quinone (dopachrome). Dopachrome is a red substance

(Xmax = 305 my and 475my) and it is the first visible product formed in the reaction; e) under physiological conditions dopachrome decarboxylates and undergoes a rearrangement to form 5,6-dihydroxyindole (Xmax = 275 my and 298 my); f) the indole compound is rapidly oxidized to Figure 4. The Raper-Mason pathway of mela­ nin synthesis (Modified from Blois, 1971). 40

Tyrosinase Tyrosinase CH—COOH CH—COOH HO n h 2 n h 2

tyrosine 3,4-dihydroxyphenylalanine (dopa)

HO

CH—COOH COOH HO

dopa-quinone leucodopachrome

HO

COOH HO H

dopachrome 5 ,6 -dihydroxy indole

N

indole 5,6 quinone MELANIN

F ig u re 4 41 the corresponding quinone; g) the indole-5,6-quinone then polymerizes to melanin through polyindole quinone with the consumption of approximately one atom of oxygen. Relative­ ly little is known about the mechanism of polymerization

(Seiji, 1967). Some modifications have been introduced in the

Raper-Mason melanin synthesis pathway. Nicolaus (1962) suggested the direct incorporation of dopaquinone, dopa­ chrome and 5,6-indole quinone-2-carboxylie acid into melanin. Axelrod and Lerner (1963) suggested O-methylation plays a role in the conversion of tyrosine to melanin in vivo. The melanin synthesis pathway as suggested by these authors is shown in figure 5. Since patients with melanomas excrete increased amounts of homovanillic acid (Duchon and Gregora, 1962) , Axelrod and Lerner (1963) suggested the possibility that patients with melanomas excrete increased amounts of O-methyldopa and the methoxy- hydroxyindoles. Misuraca et al. (1969) indicate that phaeomelanins are formed in vivo by a deviation of the eumelanins pathway involving a reaction between cysteine and dopaquinone produced by the enzymatic oxidation of tyrosine. Misuraca et al. (1969) proposed the pathway as shown in figure 6. Figure 5. The site of action of catechol-O- methyl-transferase (COMT) and hydroxyindole-O-methyl transferase in the conversion of tyrosine to melanin

(From Axelrod and Lerner, 1963) NINVIIN

S ajn B ij

H

HO ax.M aj*tnrai«|A|l«a- •faputXmpA) x c c 1WOO I

(•HM JqMltop) O M M )!*-*'* i— e x . ■«)«fiMpXi)|p.c'C-Aira^m*-e ; - x o c I P!>» n HOOJn_/MvV /S/'0*H9 h o o »«y /N v ^ ^ ^ S ^ 14 1NOO L-O OH *--- \ ^ - O M T

H O OXXX >

iwo> h o oIIX: ho JUOC

JUD" Figure 6. Pathway of phaeomelanin synthesis (From Misuraca et al., 1969; modified according to Thomson, 1974). 45 OH

enzyme enzyme c h n h 2

c o 2 h tyrosine dopa

HO

OH cysteine }

V « 2 SCHoCHC00H c o 2h c o 2h dopaquinone cysteinyldopa

^ O 2H + n h 2 — 2e s c h 2c h c o 2 h

c h n h 2 CHNHr i A c o 2 h

o-quinone

N \ ^ - co2 H PHAEOMELANINS

CHNH, i * c o 2 h dihydrobenzothiazine

F ig u re 6 46

This pathway is consistent with the fact that both tyrosinase and dopaoxidase*activities of the enzyme have been detected in melanic and phaeomelanic hair follicles and feather papillae. A change in the cysteine content of melanocytes could explain the mechanisms leading to eumelanin or phaeomelanin (Misuraca et al., 1969). In this respect, Cleffmann (1964) showed that the increase in sulfhydryl content of melanocytes is one of the initial events leading to the production of phaeomelanins instead of eumelanins.

The presence of tyrosinase seems to be necessary only for dopaquinone formation, as in eumelanins synthesis. The divergence from eumelanin pathway follows immediately after dopaquinone formation and it is the presence or absence of cysteine which determines whether melanocytes produce eumelanins and/or phaeomelanins. Very little is known about the synthesis of phaeomelanins beyond the dihydrobenzothiazine stage (Thomson, 1974).

Synthetic Melanins Synthetic melanins, as most natural melanins, are amorphous brown to black substances, obtained either by an enzymatic process or by auto-oxidation. These synthetic

Tyrosinase and dopaoxidase activities refer to hydroxylation of tyrosine and oxidation of dopa, respec­ tively by the enzyme tyrosinase (Seiji, 1967). 47

melanins generally result from the oxidation of phenolic

compounds. Mushroom tyrosinase is the enzyme generally used

for synthesis of melanin. In our laboratory we have used peroxidase (horseradish) for the synthesis of melanin from

dopamine. The model systems for synthesis of melanins in vitro are generally set up to understand the synthesis of natural melanins. Most studies of melanin formation

in vitro have been oxidative processes, enzymic or other­ wise, such as the oxidation of 5,6-hydroxyindole, 3,4- dihydroxyphenylalanine, or 3,4-dihydroxyphenylethylamine.

In all of these cases, the intermediate formation of indole-5,6-quinone, has been assumed. Bu'Lock (1961) studied the nature of the reacting species, in particular the extension to which oxidation precedes, accompanies or

follows the polymerization steps in the formation of melanin from adrenochrome. Melanins derived from 3,4-dihydroxyphenylethyl­ amine, 3,4-dihydroxyphenylalanine and 5,6-dihydroxyindole in vitro, either enzymically or auto-oxidation, as well as from tyrosine enzymically in vitro have been studied by

Swan and associates to a great extent (Swan, 1963).

According to Swan (1973) if the Raper scheme of melanin synthesis were correct, it would be expected that the same melanin would be obtained from tyrosine, L-dopa, 48 dopamine, and 5,6-dihydroxyindole.1 However, this does not seem to be the case. From experiments on the decarboxy­ lation of melanin derived from [^CC^H] -dopa Swan and

Wagott (1970) concluded that the melanin might contain not only indole 5,6-quinone, indole-5,6-guinone-2-carboxylie acid and indoline - 5,6-quinone-2-carboxylic acid units, but also units which had not undergone cyclization. In dopamine melanine the proportion of uncyclized units

(approximately one in three) appears to be greater than in dopa melanin (Binns et al., 1970). The authors con­ cluded that in dopamine melanin there must be a consider­ able number of polymeric linkages at positions other than those thought to be involved in dopa melanin. From these results, no conclusive evidence for the structure of melanin is available. Several authors

(Blois, 1973) agree that the evidence available at the present time points to the probability that dopa-type melanins, natural or synthetic, are irregular polymers built up from several types of units. Mason (1967) defends the hypothesis of a regular polymer of melanin. According to Binns et al. (1970) , however, some physical and chemical differences between melanins derived from dopa, dopamine, and 5,6-hydroxyindole have been observed. Auto-oxidation of dopamine results in a black precipitate, soluble in dilute sodium hydroxide solution, 49 while dopa melanin remained in the solution. The i.r. spectra of these melanins did not show sharp peaks, but the peaks were not identical. Thathachari (1973) studied structure of melanin, both synthetic and natural using X-rays diffraction techniques. He concluded that there is a short range order at the molecular level in all melanins. The high density of the melanins is responsible for the observed contrast in the electron micrographs of melanin containing tissues. McGinnes and Proctor (1973) suggest that melanin is black because absorbed light is not reradiated, but is converted to rotational and vibrational degrees of freedom

(i.e. heat) and also that the relatively featureless spectrum of melanins from the far ultraviolet to the in­ frared means that such transition is available for any energy photon between these limits. As in natural melanins, free radicals have been detected in dopa melanin (Adams, et al., 1958) by electron spin resonance. Hegedus and Altschule (1970) synthesized soluble rheomelanins from several catechols in plasma.

All the rheomelanins obtained showed the presence of free radicals as expected. 50

Free Radical Properties of Melanin A chemical substance that has an odd number of

electrons and is therefore usually highly reactive and

unstable and cannot be isolated by ordinary methods is

generally referred to as a free radical. In contrast,

most chemical compounds have an even number of electrons

and are stable (Pryor, 1970). All melanins known so far in their original state

have free radical properties. They are considered stable free radicals. The chemical and biological significance of free radical development during the formation of

natural and synthetic melanins is unknown (Mason et al.,

1960). Correlation between pigmentation and free radical

content has been demonstrated in several species. Whole white Calliphora puparia (bottle fly larvae) have no free

radical content, whereas the black puparia gives an elec­ tron spin resonance signal. Also, in human hair samples of varied colors an increasing order of free radical con­ tent was observed. Reduction of the free radicals with ascorbic acid or l^S produces a decrease in the free radical content (Mason et al^. , 1960) . Van Woert et al. (1967) studied neuromelanin,

isolated neuromelanin pigment, Amphiuma (snake) melanin, 51 lipofuscin and several synthetic melanins by comparing them by electron paramagnetic resonance, infrared spectro­ scopy, ultraviolet spectroscopy and fluorescence micro­ scopy. The electron paramagnetic resonance revealed that there is no paramagnetic resonance absorption in the non- pigmented tissue of substantia nigra; however, the pigment showed the presence of a stable free radical species with a G-value of 2.005 and a line width of 11.7. The signal amplitude increases when the pigment is irradiated with visible light. Amphiuma melanin, dopa melanin and other synthetic pigments demonstrated similar characteristics of the electron paramagnetic resonance spectra. The line widths of the signals are greater prior to the removal of the protein by digestion, which seems to indicate that the synthetic melanins combine with the enzyme protein in the preparations so that the protein interaction with the unpaired electrons resembles that seen with the natural melanoprotein pigments. The free radical properties of human melanoprotein from hair and melanoma tissue obtained by Bolt (1967) are similar to those reported by Mason et al. (1960) and Van Woert et a2^. (1967) .

Lipofuscin also showed a similar spectra with a

G-value of 2.005, due probably to the insoluble pigment residue and it is not present in the lipid fraction (Van Woert et al., 1967). 52

A summary of electronparamagnetic resonance spectra from several melanins is presented in figures 7

and 8. The free radical property of melanin seems to be

a property of the pigment itself since, as reported by

Mason et al. (1960), there is no evidence of any detectable free radical in the intermediate products between 3,4- dihydroxyphenylalanine and indole-5,6-quinone, or formation of radicals during the enzymic transformation. Mason et al. (1960) ascribe the free radical property of dopa melanin to a semiquinoid form,

HO

which must be greatly stabilized by the possibility of resonance throughout the highly conjugated polymer and by steric restrictions in reactivity. The importance of the characteristics of free radi­ cals of melanins comes to light when one considers it as a possible mechanism of interaction of melanin with drugs which, per se, are electron donors. The interaction of melanins with drugs is well documented especially the interaction with chlorpromazine. Figure 7. Electron paramagnetic spin reson­ ance spectra of natural and synthetic melanins. (aVan Woert and Ambani, 1974; ^Borg, 1972; cMason et al., 1960) SUBSTANTIA NIGRA MELANIN0

SYNTHETIC L-DOPA MELANIN* AMPHIUMA MELANIN0

LIPOFUSCIN MELANIN COMPONENT0 y\r 38 GAUSS

10 GAUSS

w ith ascorbic acid

BOVINE UVEA MELANINb SEPIA MELANIN*

Figure 7 Figure 8. Electron paramagnetic spin reson­ ance spectra of human hair melanin before and after ultraviolet irradiation (From Mason et al., 1960). J J ~'v ' V

<— — >

77 GAUSS UV 77 GAUSS UV

HUMAN HAIR (BLACK) HUMAN HAIR (BLONDE)

Figure 8 <_n o\ 57

Figure 9. Charge-transfer interaction model reaction for compounds having free radical properties. An electron from the highest filled orbital of the donor supposedly goes to the lowest empty orbital of the acceptor.

/ CHLORPROMAZINE® + MELANIN® CHlORPROMAZINI+ -f MELANIN'

DONOR ACCEPTOR

F ig u re 9

U1 00 Karreman et al. (1959) postulate that due to the charac­ teristics of electron donor of chlorpromazine it may interact with electron acceptors by a mechanism called

"charge transfer." In "charge transfer" complexes an electron supposedly goes from the highest filled orbital of the donor to the lowest empty orbital of the acceptor (see figure 9). This transfer does not involve configura­ tional change. Karreman et al. (1959), then support the hypothesis of a charge-transfer mechanism for drugs acting in the central nervous system since they also found that d-lysergic acid diethylamide and serotonin are very good electron donors. On the other hand, Borg (1965) studied the formation of free radicals of imipramine. His results led him to speculate that at least part of the pharmaco­ logical actions shared by phenothiazines and structurally related drugs may stem from their transformation to free radicals in vivo. Bolt and Forrest (1968) suggest that, in vitro, melanoprotein accepts an electron from chlorpromazine to form a diamagnetic species; i.e., reduced melanoprotein, therefore melanoprotein participates in a complete charge- transf er reaction. These results are supported by Van

Woert (1968) experiments demonstrating that NADH is oxidized in vitro in the presence of melanin. This NADH oxidation is inhibited by phenothiazines, which form 60 charge-transfer complexes with melanin, which indicates that this oxidation is most likely related to the free radical properties of melanin.

Binding of Drugs to Melanin

Potts (1962, 1964) was the first to observe binding of drugs to pigmented tissues. He found that phenothiazines were concentrated in the pigmented struc­ tures of the rabbit eye, with little or no retention in corresponding structures of the eye of albino animals

(Potts, 1962). Although Potts1 experiments demonstrated binding of drugs to melanin in vivo and in vitro, the influence of pigmentation on the distribution and effects of drugs was first suggested by Chen and Poth (19 29). They observed that racial differences influence the mydriatic action of several drugs. The initial mydriatic effect of cocaine, euphthalmine, 1-ephedrine, dl-ephedrine, and d-pseudo- ephedrine was more marked in Caucasians, while the drugs were less effective in Negroes. These results were supported by Barbee and Smith (1957) who found ephedrine to be ineffective as a mydriatic in Negroes, while it was equally effective in Caucasians with blue or light brown eyes.

/ 61

Angenent and Koelle (1952, 1953) assumed that the differences in drug effects in pigmented and nonpigmented eyes had an enzymatic basis since a tyrosinase system capable of transforming catecholamines to melanins is found in the nonalbino rabbit eye but not in the albino rabbit eye. They suggested that the continuous and rapid destruction of the adrenergic neurotransmitter by the enzyme would reduce the effectiveness of drugs such as ephedrine which produce their effects through the release of the chemical mediator. Subsequently, Obianwu and Rand (1965) suggested that the poor mydriatic effect of ephedrine in pigmented eyes may be due to a lower content of norepinephrine in these irides, probably because of the diversion of cate­ cholamine precursors to melanin synthesis. These authors also suggested that the adrenergic innervation in the pigmented eye is poorer than in the nonpigmented one.

Seidehamel et al. (1970) showed that nonpigmented guinea-pig irides are more sensitive to ephedrine mydriatic effects than the pigmented ones. Patil et al. (1974) found that (-)- 14 C-ephedrme accumulates in greater amounts in pigmented guinea-pig irides than in nonpigmented ones. The drug disappears faster from the nonpigmented than from the pigmented iris. 62

The differential binding of cocaine in pigmented

and nonpigmented guinea-pig eyes was also observed. The 14 binding of (±)- C-cocaine is inhibited m the presence of (-)-^C-ephedrine, but it is not affected by chemical sympathectomy or reserpine (Patil, 1972). Using autoradiographic studies, Lindquist (1973) demonstrated binding of epinephrine, dopamine, norepin­ ephrine and serotonin to melanin granules in vitro, while 14 L-dopa and L-tyrosine were not. C-labelled catechola­ mines and serotonin bound superficial cervical lymph nodes and melanin structures in the pigmented mouse in vitro, whereas no accumulation was observed in albino mice. Jacquot et al. (1974), in studies on guinea-pigs found that forty hours after intravenous administration of

(±)-ephedrine-^C, intense accumulation of the drug could be seen autoradiographically in pigmented tissues such as choroid and skin. In vivo incorporation of (+)-ampheta- 14 3 mine-2- C (Harrison et al., 1974a) and of L-dopa- H, L- a-methyldopa- 3 H and (±)-isoproterenol-7- 3H (Harrison et al.,

1974b) into melanin and nonmelanin components of pigmented and albino guinea-pig hair was studied. Incorporation of the drugs into melanin and nonmelanin fractions was demon­ strated.

Wepierre et al. (1975) tried to correlate the mydriatic effect of several sympathomimetic amines to the pigmentation of the eye. In two different strains of rats, using autoradiography, they found that the non- 14 phenolic amines: (±)-ephedrine- C, (±)-norephednne- 14 14 methyl- C and (+)-amphetamine-7- C, had a high affinity for the pigmented tissues in pigmented rats. Tyramine, a phenolic compound, did not accumulate in the pigmented tissue. The nonpigmented eyes of albino rats did not accumulate any of the drugs. The authors suggest that in vivo, the accumulation of a-methylated sympathomimetic amines in pigmented tissues is related to the rate of metabolism of the drug. In in vitro experiments, Patil et al. (1974) demonstrated binding of tyramine to pig­ mented iris, but in this case a monoamineoxidase inhibitor was used.

The relationship between metals and melanin has been known for some time. Cotzias et al. (1964) and Cotzias (1974) have emphasized the importance of this relationship in the central nervous system. Potts and Au (1971) found that 129 hours after intramuscular admin- istration of 204 Thallium, the metal had accumulated m the uveal tissues of pigmented rabbits. No evidence of accumulation in cornea, aqueous, vitreous, sclera or blood was found. The accumulation seemed to increase with time, probably at the expense of other tissues. 64

There are very few reports on the binding of cate­

cholamine pigmented tissues in the brain, in spite of the

fact that the striking similarity of the distribution of neuronal melanin and that of the brain catecholamines

(Bazelon, et al., 1967) indicates some relationship between neuromelanin and the presence of dopamine norepinephrine in the nerve cells. Yshii and Friede (1968) have reported accumulation of tritiated norepinephrine in pigmented

nerve cells in human substantia nigra, in vitro. The authors concluded that the binding occurred at the surface of the cells since no binding was noted in the cytoplasm

and on the melanin granules. In contrast to these results, 3 Lindquist (1973) observed selective binding of H-dopamme

and 3 H-norepinephrine to melanin granules in the pigmented cells of human substantia nigra in vitro. No accumulation of radioactivity was seen in the cell membrane of pig­ mented cells. It is possible that differences in the thickness of the tissue sections used by each investigator may account for the differences in the results. The antimalarial drugs, chloroquine and quinine, are well known for their ototoxic properties. Potts (1964) showed that quinine binds in vitro to uveal melanin granules. Lindquist and Ullberg (1972) found a high 14 accumulation of C-chloroquine in the inner ear of fetuses from pregnant pigmented mice whereas no accumula­ tion was found in the inner ear of albino fetuses following 65 administration of the drug by intravenous injection to the pregnant mouse. The toxicity of aminoglycoside antibiotics to the inner ear has been the subject of numerous studies. The rate and extent of accumulation of these drugs in inner ear fluids (endolymph and perilymph) has been studied in guinea-pigs (Stupp et al., 1973). Evidence has been provided that the effect is due to the affinity of the antibiotics for melanin present in the cochlea/ mainly in the stria vascularis, one of the sites where endolymph is thought to be produced (Dencker et al., 1973).

Dencker et al. (1973) have found serious damage of the stria vascularis of young pigmented guinea-pigs treated with kanamycin, but not in albino animals. In normal animals melanin granules are evenly distributed in the strial cells; in kanamycin treated animals the granules were aggregated. In in vitro experiments using bovine iris melanin granules, Lindquist (1973) has shown that of the ototoxic drugs, kanamycin has the highest affinity for melanin, followed by chloroquine, quinine, dihydrostreptomycin, streptomycin and viomycin. Salicylic acid, a drug which may cause temporary hearing loss did not show affinity for melanin. Experiments conducted in rats and guinea-pigs to demonstrate the influence of pigmentation on the ototoxic­ ity of kanamycin and neomycin indicate that pigmented 66 animals may be more likely to suffer hearing impairment following ototoxic drug administration, however this has not been confirmed so far (Harpur and D'Arcy, 1975).

The importance of the affinity of phenothia2 ines and 4-aminoquinolines in the etiology of toxic retino­ pathy is well established. Potts (1962, 1964) was the first to demonstrate the binding of these as well as of several other polycyclic compounds to melanin in vitro.

Potts (1962) studied the binding of phenothiazines in the eyes of pigmented and nonpigmented rabbits following systemic administration. He found selective accumulation of substituted phenothiazines in the uveal tract, the pigmentation being a very important factor for the accumulation. Interestingly no accumulation of the parent compound, phenothiazine, could be demonstrated.

In other experiments Potts (1964) demonstrated the binding of several polycyclic compounds to melanin gran­ ules in vitro. He also demonstrated that the accumulation is a function of melanin per se, because treatment of the granules with proteolytic enzymes did not modify the binding of the polycyclic compounds and also, binding of similar characteristics are observed with synthetic mela- nins. Changes in pH or salt concentration did not modify the binding of the drugs to melanin. 67 Blois (1965/ 1968) demonstrated the binding of

35S-chlorpromazine and of iodoqume to melanin containing tissues in melanoma-bearing C3H mouse. 35 S-chlorpromazine achieved tissue concentrations at equilibrium in proportion to the melanin content of the tissue. Blois and Taskovich (1969) determined the selectivity of the binding of drugs to melanins, comparing it to the binding to other bio­ polymers using a dialysis procedure'. The ratios (ratio of radioactivity remaining in the dialysis bag to that re­ maining in the control after 50 hours of dialysis) of binding of drugs to desoxyribonucleic acid, ribonucleic acid, albumin, lipoprotein and melanin were determined.

Drug binding ratios were consistently higher for melanin than for the other polymers used. Lindquist and Ullberg (1972) and Lindquist (1973) have reported the most complete studies on the binding of drugs to melanins. Lindquist and Ullberg (1972) using autoradiographic methods to study the melanin affinity of chloroquine and chlorpromazine in albino and nonalbino mice demonstrated that 35 S-chlorpromazine was localized in several tissues including the melanin structures of the eye, Harder's gland, superficial cervical lymph nodes, skin, hair follicles, and brain. Accumulation in the brain decreased rapidly, but levels of the drug were high in melanin structures 90 days after injection. The drug was also accumulated in the fetal eye. 68

They showed that levels of 14 C-chloroqume m melanin structures were constantly high 90 days after injection, and, surprisingly, the concentration of the drug in the uveal tract was high even one year after administration of the drug. A leucocytic melanin transport through the body in normal Bantu negroes has been described (Wasserman,

1965). Satanove (1965) and Wasserman (1965) suggested that phenothiazine-induced general melanosis is caused by the leucocytic transport of melanin from the hyper- 35 pigmented skin to the viscera. High accumulation of S- chlorpromazine, or ^C-chloroquine was found in the super­ ficial cervical lymph nodes of pigmented mice after injec­ tion of the drugs. No accumulation was found in the corresponding lymph nodes of albino mice (Lindquist and Ullberg, 1972). The accumulated drugs in the lymph nodes seems to be located in melanosomes present in the nodes as demonstrated histologically and by electron micro­ scopy by Lindquist (1973). At the present time there is very little known about binding of drugs to pigmented tissues in the brain.

The lack of information probably is due to the fact that neuromelanin is not found in small laboratory animals such as guinea-pig, mouse, rat or rabbit (Marsden, 1961). A very important finding in this respect was reported by 69

Christensen et al. (1970) who found degeneration of pig­ mented nerve cells in the substantia nigra of patients with phenothiazine-induced tardive dyskinesia. Lindquist (1972, 1973) using autoradiography demonstrated in vitro uptake of 35 S-chlorpromazme m the pigmented neurons of human substantia nigra and locus coeruleus. These results emphasize the importance of above noted finding (Christensen et al., 1970). However,

Lindquist's experiments failed to show selectivity of the binding to the pigmented tissue since there was no com­ parison of the binding of the drug to nonpigmented tissues of the brain. The affinity of phenothiazines and 4-aminoquino- lines for melanins is generally accepted as the most important factor in the etiology of the toxic retinopathy caused by these drugs. Chloroquine and chloropromazine induced pigmentation disorders of the skin have also been related to accumulation of drugs in skin melanin (Satanove,

1965; Stewart et al., 1968). Most of the ocular changes have been associated with the intake of NP-207, thiori­ dazine, chlorpromazine and trifluoperazine (Sidall, 1967).

The first reports on retinotoxicity of a pheno- thiazine originated from a clinical test of piperidyl- chlorophenothiazine (NP-207) (Goar and Fletcher, 1957). 70

There are several reports on toxic chorioretinopathy in

man after thioridazine treatment, (Scott, 1963; Sidall,

1966, Cameron et al., 1972). Although it is the most used phenothiazine, very few reports on chlorpromazine retino­

pathy have been published. However, chorioretinopathy in patients treated with chlorpromazine has been reported

(Sidall, 1966). Investigations comparing the effects of

chlorpromazine, thioridazine and NP-207 in the cat have

demonstrated that although chlorpromazine content can be

increased 15 times in the retina (relative to calculated equal distribution in the whole body), no retinotoxicity

could be observed; NP-207 is highly retinotoxic with only a 10 times relative accumulation in the retina

(Meier-Ruge et al., 1966). Legros et slI. (1971) demon­ strated chlorpromazine induced modifications of the elec- troretinogram and retinal histologic lesions in the dog.

Electroretinogram changes with no histological lesions

in albino rats and no changes in retina in pigmented rats were shown by Legros et al. (1973). The authors suggest

that the ocular toxicity of chlorpromazine is not due

to its affinity for melanin. Many patients treated with phenothiazines develop what is called the eye-skin syndrome, which is described

as a combination of pigment deposits in the eye and hyper­ pigmentation of the skin, localized mainly in the face, arms, hands and legs. The color of the affected areas is described as diffuse violet to dark blue-black (Hays et al.,

1964). The identity of the pigment has not been estab­ lished, but the formation of melanin or a melanin-like substance has been suggested. Perry et al. (1964) indi­ cated that the pigment may be a metabolite of chlorproma­ zine. Although Bolt and Forrest (1968) support the con­ cept of chlorpromazine metabolism and hyperpigmentation of the skin, they consider that ocular opacities have a different but yet unexplained origin. Based on electro- microscopic studies, Zelickson (1965) reported that the pigment is not due to true melanin. He suggested that it is a drug metabolite, pseudomelanin, or some other form of pigment. Howard et al. (1969) induced experimental cata­ racts with chlorpromazine in both albino and nonalbino guinea-pigs. The authors concluded that chlorpromazine cataracts do not appear to be related to the presence of melanin in the eye. The phenomenon according to them may represent a foci of denatured protein resulting from the interaction of light with the drug, a photosensitizing agent, and lens protein, or possible deposits of the drug within the lens. In chlorpromazine treated drugs,

Tousimis and Barron (1970), demonstrated the presence of granules within the cytoplasma of corneal stromal ceils that seem to be associated to the prolonged oral admin­ istration of the drug. Similar structures were not 72 found in untreated dogs. In patients with chlorpromazine induced general melanosis the pigment has been found to be deposited throughout the reticuloendothelial system and viscera.

The identity of the pigment has not been established.

Greiner and Nicolson (1964) suggested that it is a mela­ nin because of its histochemical reactions. Satanove (1965) suggested that the phenothiazine induced general melanosis is due to leueocytic transport of the pigment from the hyperpigmented skin to the viscera. This suggestion is supported by Wasserman's (1965) demon­ stration of a leucocytic melanin transport system which was found to be a normal phenomenon in Bantu negroes. Arons et ad. (1968) demonstrated lack of melanin pigmentation in the skin of an albino psychiatric patient treated with chlorpromazine in high doses. Dopa and tyrosinase activity were observed within the epidermis.

Electron microscopic studies demonstrated that in epi­ dermal melanocytes with nonpigmented melanosomes the response as well as the metabolism of the drug were normal. It is very interesting that in the study reported by Zidek and Janku (1971) albino mice were found to be less resistant than pigmented ones to the toxic effects of chlorpromazine and isoniazide. 73

It is a well recognized fact that the extrapyra- midal disorders produced by neuroleptic drugs or encepha-

litic processes have in common the involvement of the pigmented centers or the midbrain, the neuromelanin of which has been altered within the cells or displaced from them (Forrest, 19 74). When pigmented neurons degenerate, their melanin granules are released and may be seen lying

free in the tissue or within phagocytes. Such changes occur irrespective of the cause and represents the out­ come of cell death. The system of pigmented neurons appears to be specifically damaged in Parkinsonism

(Marsden, 1969). Decrease in the very dense component of melanin granules in substantia nigra (Duffy and Tennyson, 1965), decrease in the number and melanin content of cells in the substantia nigra (Pakkenberg and Brody, 1965), and quan­ titative reduction of melanin pigment in substantia nigra (Roy and Wolman, 1969) of Parkinsonian patients has been reported. Hornykievicz (1973) demonstrated a reduction of dopamine concentration in the corpus striatum of Parkinson's disease patients. The change appears to result from destruction of the substantia nigra with loss of dopamine containing nigrostriatal fibers (Poirier and Sourkes, 1965). According to Marsden (1969) the appear­ ance of Parkinson's syndrome during the treatment with 74 a-methyldopa, reserpine and phenothiazines relates to the ability of these drugs to interfere with adrenergic mech­ anisms in the brain. As mentioned above, extrapyramidal disorders are known to be the most common adverse effects of phenothia­ zines . The main extrapyramidal disturbances caused by phenothiazines are akathisia, dyskinesia, and Parkinsonism. The mechanism responsible for the development of pheno- thiazine induced extrapyramidal symptoms is not known.

Forrest et al^. (1963) found depigmentation in substantia nigra and cell degeneration with neurophagia most pro­ nounced in the putamen in the brain of a patient who had been treated for several years with trifluopromazine and chlorpromazine. Christensen et al. (1970) in a study of

28 brains from dyskinesia patients, of which 21 had neuroleptic drug induced dyskinetic symptoms, found cell degeneration in substantia nigra in 96% of the cases, whereas only 25% of the control cases showed histological changes. In analyzing the origin of extrapyramidal diseases,

Forrest (1974) concludes that neuroleptic drugs, and especially chlorpromazine, among the phenothiazines, have a special affinity and attach themselves to neuromelanin of the extrapyramidal neurons producing charge-transfer complexes. The resulting compounds are bioelectrically inert and lose the electron trapping function of melanin. 75 Histopathological autopsy findings on dyskinetic and drug treated patients seems to confirm the causative relation­ ship between drug induced extrapyramidal disorders and the neuromelanin of the pigmented subcortical nuclei. Chlor- promazine-melanin complexes are eventually expelled as foreign bodies from the neuron, which itself may disinte­ grate in the process.

Pathophysiological Significance of Natural Melanins Although melanin itself is considered by many as only a protecting agent against ultraviolet radiation, when one considers the abnormalities of pigmentation, melanin assumes a different dimension. Most of the dis­ orders of pigmentation are genetically determined. Phenyl­ ketonuria and albinism are two of the most important biochemical disorders of pigmentation. In phenylketonuria there is a reduction of phenyl­ alanine hydroxylase, thus, the conversion of phenylalanine proceeds at a very low rate. The condition is inherited as a recessive trait, is characterized by decreased pig­ mentation of the hair and eyes and, if untreated, becomes associated with mental deficiency (Jervis, 1937). It is thought that phenylalanine, not being converted to tyro­ sine, may act as tyrosinase inhibitor causing a reduction of hair and eye pigmentation together with lessened 76 pigmentation in other locations such as the substantia nigra (Fitzpatrick et al., 1961). It has also been suggested that phenylalanine acts as repressor of tyro­

sinase synthesis and that the hypopigmentation of phenyl­ ketonuria is a combination of a repressor and a competi­ tive inhibitor effect of phenylalanine (Riley, 1974). Changes in the activity of tyrosinase, however, do not explain the paradoxical depigmentation of the substantia nigra in phenylketonuria. Albinism is due to a genetic defect in which tyrosinase is in an inactive form. The structural gene for tyrosinase appears to be located at the C locus, is responsible for the production of a tyrosinase with re­ duced activity (Coleman, 1962; Witkop, 1971). Hearing

(1973) suggested that the C locus may be regarded as regulatory rather than a structural locus for tyrosinase.

There are two types of oculocutaneous albinism described so far. Tyrosinase-negative oculocutaneous albinism, characterized by the presence of nonmelanized melanosomes within melanocytes and lack of histochemically demonstrable tyrosinase activity. As in the albino mouse, human tyrosinase-negative oculocutaneous albinism appears to involve a defect in tyrosinase. On the other hand, tyrosinase-positive oculacutaneous albinism in man is defined by the existence of partially melanized melanosomes 77 within melanocytes and evidence of tyrosinase activity within hypomelanotic hair bulbs incubated in tyrosine-

containing solutions. This type may result from limita­

tions in the availability of tyrosine within melanosomes

(Witkop, 1971). Pathological disturbances of pigmentation may be classified as conditions associated with an increase in pigmentation and those in which there is a loss of pig­ mentation. Hyperpigmentation includes: photosensitivity, acanthosis nigricans, urticaria pigmentosa; hyperpigmen­ tation associated with hormonal disturbances involving

ACTH and MSH, with mast cell tumors (mastocytosis) and with schizophrenia in patients not treated with pheno­ thiazines (Riley, 1974). According to Lerner and Fitzpatrick (1950) certain vitamins are involved in melanogenesis and skin pigmenta­ tion may result from deficiencies of vitamin A, nicotina­ mide and ascorbic acid. The hyperpigmentation may result from a decrease in sulfhydryl groups which, on the other hand, act as tyrosinase inhibitors (Seiji et al., 1969). Hypomelanosis may result from: a reduction in the number of pigment producing cells; a reduction in the efficiency of melanin biosynthesis; a reduction in the transference of pigment from melanocytes to keratinocytes 78 or a rapid loss of melanin from the epidermis (Riley,

1974). Depigmentation can be obtained by using chemical agents such as hydroquinone monobenzyl ether, p-hydroxy- anisole and other phenols. Frenk et al. (1968) have reported data which suggests that mercaptoethylamines, 2-mercaptomethylamine (cysteimine) and 2-mercaptoethyl-diethylamine, destroy melanocytes. Although no explanation for the mechanism of this action has been offered, it is suggested (Riley,

1974) that these sulfhydryl compounds reduce the ability of the cell to inactivate potentially dangerous free radicals as a consequence of the inhibition of melanin synthesis since melanin seems to act as free radical scavenger because of its free radical properties.

Development of tumors in the melanocyte system is also well known. There are benign tumors such as lentigo, Hutchinson's freckle, cellular naevi and juvenile melanoma, and the malignant tumors, malignant melanoma and xeroderma pigmentosa. The discussion of nature and characteristics of these tumors is beyond the scope of this paper.

Other pathological problems related to melanin involve metals. The importance of metals in extrapyra­ midal areas is indicated by the association of extra­ pyramidal disturbances to abnormalities of copper 79 metabolism in Wilson's disease and manganism. In patients with Wilson's disease there is accumulation of copper in the liver. Also, Leu et al. (19 70) described hyperpig- mented lesions on the lower portion of the legs in patients with Wilson's disease. Histological studies showed that the lesions were due to excessive melanin deposition rather than to the presence of copper or iron. The etiology of the increase in melanin pigment is unknown.

The deposition of copper in the brain in Wilson's disease (Cummings, 1968) and the therapeutic effects of chelating agents upon the extrapyramidal symptoms suggest that the extrapyramidal abnormalities are related to copper

(Curzon, 1975). Neurological disturbances, mostly of Parkinsonian type have been described in chronic manganism

(Mena et al., 1967). Lack of wool pigmentation has been found in copper deficient sheep. The phenomenon has been attributed to the inhibition of tyrosinase activity resulting from copper deficiency (Underwood, 1971). In relation to heavy metals, melanosis has been associated with prolonged use of medicaments containing arsenic, bismuth, gold or silver. The mechanism of melanosis seems to be related to binding of the metals to sulphydryl groups, eliminating the inhibitory action of these groups on the activity of tyrosinase. An increase 80 in melanin within the basal cell layer and dermal melano­ cytes is generally observed (Lorincz, 1954). Irradiation seems to increase this pigmentation since dark areas are limited to the exposed skin (Molokhia and Portnoy, 1973). Bleaching of the skin can be obtained by using cosmetic creams containing mercury. This is thought to be due to replacement of copper by mercury in tyrosinase and subsequent inactivation of the enzyme (Lerner, 1952).

Hyperpigmentation however may result, probably as a con­ sequence of binding of sulfhydryl groups by mercury (Burge and Winkelmann, 1970). A typical Parkinson's syndrome has occasionally been described in mercury poisoning

(Cohen, 1962). The preceding information reveals that melanin is not just a protecting shield against actinic radiation. Disturbances in melanin biochemistry result in phenyl­ ketonuria, for example, a serious problem that may cause mental retardation. The extrapyramidal syndrome as well as pigmentation disorders associated with drugs or metals are serious enough to deserve more attention not only from a clinical, but also from a pharmacological and toxicological point of view. 81

STATEMENT OF THE PROBLEM

It is obvious from the foregoing introduction that many aspects of melanin chemistry and biochemistry are still unexplained. There is very little information about the pharmacological and toxicological significance of the presence of melanins in the body. Because of the broadness of the field it was decided to limit this study to only certain aspects of the inter­ actions of drugs with melanins. These interactions may re­ late to the development of drug-induced side effects in melanin bearing tissues. It has been known for many years that atropine and atropine-like drugs, when applied to heavily pigmented eyes, exhibit unusually slow onset of mydriatic and cyclo- plegic effects. Although atropine is classified as a com­ petitive reversible blocker of the muscarinic receptor, there is no satisfactory explanation for the pigment- dependent initial slow effect and the long duration of its mydriatic and cycloplegic effects. The appearance of the extrapyramidal syndrome in patients who have undergone prolonged treatment with pheno­ thiazines and butyrophenones or who have been exposed to 82 toxic quantities of metals such as manganese is also well known. Although the possibility of neuromelanin involve­ ment has been suggested, there is very little evidence re­ ported so far, that the drugs or metals involved have affinity for neuromelanin.

On the other hand, the ability of phenothiazines and 4-amino-quinolines to bind melanins and, as a conse­ quence, to produce hyperpigmentation of the skin and retinopathy are well recognized. However, the quantitative characteristics of the binding of drugs to pigmented tissues, as well as the possibility that other drugs with the same pharmacological activity but with less affinity for melanin could be used as substitutes for the most toxic ones have not been explored so far. Therefore, the object of this study is to analyze some aspects of the binding of drugs to melanins in order to determine the following: a) Relation of the prolonged mydriatic effects of atropine to pigmentation of the iris. b) Character of binding of phenothiazines and related drugs to human brain pigmented and nonpigmented tissues in order to gain information about the selectivity of binding of drugs to brain tissue.

c) Capacity and affinity constants for binding of phenothiazines and related drugs to melanin granules 83

as compared to synthetic melanins, both in the absence

and presence of fluphenazine, thioridazine, haloperidol and clozapine. The results of these experiments may help to explain the prolonged atropine mydriatic effects and contribute to an understanding of the relationship be­ tween the presence of melanins in certain tissues and the localization and persistence of side effects produced by some drugs. CHAPTER II

METHODS AND MATERIALS

Experimental Animals Albino and nonalbino rabbits, of either sex, weigh

ing 2.3 to 4.0 Kg were used. The strains used were New Zealand White rabbits (Kings Wheel Rabbitry, Centerburg,

Ohio) and New Zealand Black rabbits (Three Springs Kennels Zelinopoe, Ohio). The animals were housed in individual cages and were allowed free access to food (Purina rabbit

chow) and water. A few days after arrival, the rabbits were classified according to the absence or presence of

atropinesterase (atropinase).

Bovine Iris Melanin Granules Bovine eyes were obtained on ice from a local

slaughter house (Ohio Packing Co., Columbus, Ohio) and on the same day tissues were processed for the granules.

Human Brain Post mortem adult samples of midbrain from adult humans containing substantia nigra and superior cerebellar peduncles and frontal cortex were obtained from the county morgue usually within 24 hours after death. The tissues 84 85 were carried on ice to the laboratory and kept frozen until used, usually for 5 to 10 days. Samples were taken from individuals with no apparent pathological brain lesion or record of drug medication at the time of death.

Experimental Procedures

Qualitative test for atropinesterase Serum atropinesterase was determined by a slight modification of the method of Van Zuphten (1972). Blood was withdrawn from the medial ear artery and allowed to settle at room temperature; then it was stored in cold (3-5°C) up to 24 hours. The resulting serum was treated for atropinesterase activity in the following way. Agar-cresol red media was prepared by adding 2.5 ml of 0.2% cresol red solution per each 100 ml of aqueous 1.5% R solution of agar-agar (Bacto-Difco Agar ) at 65°C. The pH was adjusted to 8.0 with 0.1 N NaOH. The mixture was placed in Petri dishes and allowed to solidify. Wells of approximately 6 mm in diameter were cut with a cork borer and the agar plugs removed. A 2% atropine sulfate solution was prepared and pH adjusted to 7.5 with 0.1 N NaOH. Twenty five yl of serum were placed in each well followed by the addition of 25 yl of atropine sulfate solution. For the control experiments, 25 yl of serum heated at 65°C for one hour were used. The Petri dishes were then incubated at 37°C until a yellow color around the bottom of the well 86

appeared or for a maximum of four hours. Atropinesterase degrades atropine to tropine and tropic acid; the latter

metabolite alters the alkaline pH of the medium to an acidic one. Thus, a change from red to yellow color around

the bottom of the well was indicative of the presence of atropinesterase in the serum.

3 Binding of H-atropine by Tissues Rabbits were killed by injecting air through the ear marginal vein and the irides were immediately dissected as described by Jacobowitz (1967). Tissues were rinsed twice in 5 ml each of cold physiological salt solution. Each iris was cut into two equal pieces and each piece was

placed in an oxygenated physiological salt solution at 37°C.

The composition of the solution in mM was: NaCl, 118;

KCl, 4.7; CaCl2 , 2.5; MgCl2, 0.54; NaH2P04 , 1.0; NaHCC>3 , 25; Glucose, 11. The chemicals were dissolved in demineral­

ized double distilled water. After an initial equilibration period of 15 minutes, the tissue was incubated with known specific activity of 3 H-atropine for a fixed time, then rinsed to remove excess radioactivity with two volumes of 5 ml each of cold physio­ logical salt solution. The tissue was blotted on filter paper, weighed and homogenized twice in 1 ml each of 0.4 N perchloric acid. The combined homogenate was centrifuged at 43,500 Xg, 0-4°C, for 30 minutes in a Sorvall Model 87

RC2-B centrifuge. The radioactivity in the supernatant was

counted in 12 ml of Aquasol (New England Nuclear) by liquid scintillation spectrometry in a Packard Tri-Carb Model 3375. Quenching was monitored by automatic external standardiza­

tion. The counting efficiency for the tritium varied 3 between 21 and 23%. The uptake of H-atropine by human iris and pigment epithelium was similarly studied.

When the binding of 14 C-imipramine to rabbit iris was studied, the irides were divided in four pieces and each piece was incubated with the drug for a fixed time. Binding of the drug to the tissue was determined as des­ cribed above.

Loss of radioactivity from tissues The rate of loss of ^H-atropine from rabbit iris and stomach fundus strips and of ^^C-imipramine from rabbit iris was determined by transferring the tissue every five minutes to a fresh oxygenated solution maintained at 37°C. At the end of the designated time the radioactivity re­ tained by the tissue was determined as described above.

3 Ocular penetration of H-atropine Rabbits were anesthetized with pentobarbital sodium, 30 mg/Kg, i.v., and 0.1 ml of 2% atropine solution 5 containing 5.32 X 10 dpm/0.1 ml, was placed in the left eye. At the end of the designated period, the eye was blotted to remove excess radioactivity. The animals were 88 killed by injecting air through the ear marginal vein, the eye was quickly dissected and rinsed three times in volumes of 10 ml each of cold physiological salt solution for 5 minutes. The aqueous humor was withdrawn and the radio­ activity was counted in Aquasol. The iris was isolated and the radioactivity in the tissue was determined as described above. The radioactivity in aqueous humor and iris from the contralateral eye was also determined for control purposes.

Pharmacologic experiments An isolated iris was suspended in a 10 ml organ bath containing physiological salt solution (of the same composition as described above) at 37°C. A thread was passed through the pupillary margin and tied to a glass hook; the opposite end of the iris sphincter was connected by means of a silk thread to a force displacement trans­ ducer (FT-03)(see figure 10). A baseline tension of 150 mg was maintained throughout the experiment and the change in tension developed by the iris sphincter was recorded on a

Grass Polygraph Model 7 (Grass Instruments, Quincy,

Massachusetts). After a 50 minute equilibration period, during which the tissue was washed at regular intervals, two cumulative dose-response curves to carbachol were obtained at 60 minute intervals. The tissue was thoroughly washed after the first dose-response curve and it was exposed 89

Figure 10. Tissue bath assembly used for recording tension changes of isolated rabbit iris. 90

T r a n s d u c e r

I r i s

3 7 . 5 C W a s h

Figure 10 to the antagonist for 60 minutes before recording the second dose-response curve, which was obtained in the presence of the antagonist. To check for changes in tissue sensitivity during the test period, the contra­ lateral iris sphincter without the antagonist was simultaneously set up in a separate tissue bath (Furchgott,

1967). PA 2 values were calculated according to the method of Arunlakshana and Schild (1959). Dose ratio, which is defined as the ratio of ED,.,, of the agonist with and with­ ou out the antagonist, was used to obtain the PA2 plots. Rabbit fundus strips were prepared according to the proced­ ure described by Furchgott and Bursztyn (1967). In other experiments on iris, recovery of the muscarinic blockade was studied by testing the sensitivity of the tissue to a dose of carbachol at regular intervals, a —5 control response to carbachol ED^q (1 X 10 M) was obtained and the tissue was thoroughly washed and incubated with a dose of atropine for 60 minutes. Sensitivity of the tissue to carbachol in the presence of atropine was ascer­ tained. The tissue was washed three times within 5 minutes and 10 minutes after the last wash the tissue sensitivity to carbachol was tested. The procedure was repeated several times. The contralateral iris sphincter, without atropine treatment, was used to check the sensitivity changes to carbachol during the testing procedure. 92 In vivo experiments Rabbits were classified according to the absence or presence of atropinesterase in the serum and they were

trained in restraining cages. Pupillary diameter in the presence of light (Bausch and Lomb illuminator, setting #1,

12 cm from the eye) was measured. After the instillation of 0.1 ml of a 2% atropine sulfate solution at zero time, the onset of the mydriatic effect, determined as the change of sensitivity to light, was measured every fifteen min­ utes for two hours. Thereafter, the pupillary diameter

was determined at regular intervals for 96 hours. In no case were the animals kept for more than four hours in

the restraining cages. At the end of the 96th hour the

animals were killed and the in vitro sensitivity of the iris sphincter to carbachol was tested as described above.

In a second series of experiments only atropines­

terase negative rabbits were used. After instillation of 0.1 ml of a 2% atropine sulfate solution (5.2 yCi/ml 3 H-atropine) the mydriatic effect was followed for 96 hours as described above. At the end of the 96th hour the animals were killed. The irides were isolated, weighed

and homogenized with 2 ml of 0.4 N perchloric acid, and the homogenate was centrifuged at 43,5000 Xg, at 0-4%C, for 30 minutes. The total radioactivity in the supernatant was counted in 12 ml Aquasol by liquid scintillation

3 93

spectrometry. Counting efficiency for tritium was 25%.

Thin layer chromatography In two atropinesterase-negative nonalbino rabbits,

0.1 ml of a 2% atropine sulfate solution (46.29 yCi/0.1 ml 3 H-atropine) was instilled in each eye. The higher radio­ activity per instillation was necessary to detect sufficient

counts in the tissue extract. The mydriatic effect of atropine was followed for 96 hours. At the end of the 96th hour the animals were killed and the irides isolated.

The two irides of each animal were combined and homogenized in 2 ml of a 0.4 N perchloric acid solution and the homogen-

ate was centrifuged at 43,500 Xg at 0-4°C, for 30 minutes. The precipitate was discarded and the supernatant was

placed in 2.5 ml vials. This extract was subjected to evaporation under constant air current at room temperature.

When the volume of the material was reduced to approximately

0.1 ml (after approx. 36 hr), the material was spotted on a silica gel plate. The plate was developed using a mixture of methylacetate, isopropanol, 25% ammonia (45:35:20) as solvent. The plate was divided into 1 cm segments and radioactivity was determined in each segment by liquid scintillation spectrometry as described above. 94 14 . . Binding of C-imipramine to human brain homogenates The pigmented tissue of substantia nigra weighing 0.3 to 0.5 g, or an equal amount of cerebral cortex, was homogenized with physiological salt solution (1 ml/0.1 mg) in a 5 ml glass Teflon homogenizer (Arthur Thomas, Phila­ delphia, Pa.) at 600 rpm with 8 to 10 strokes. The homogenate was subjected to sonication on ice for five minutes, using a 20% relative output of a sonic 300 dismembrator (Artek Systems Corp., Farmingdale, N.Y.).

Aliquots of the homogenate were incubated at 30°C under constant oxygenation. It was important that the oxygenat­ ing tubing did not come in close contact with the homogen­ ate since it causes some frothing. The incubation volume was 2.5 ml. After an equilibration period of 30 minutes, ^C-imipramine was added to the preparation for a final concentration of 12.8 nCi/ml (2.5 X 10 M ) . The incubation time with the drug was 30 minutes and the preparation was centrifuged at 43,500 Lg, 0-4°C, for 30 minutes. The radioactivity in the supernatant was determined and the radioactivity bound to the tissue was determined by differ­ ence from the initial amount in the incubation medium. 95

Localization of labelled drugs in tissues by discontinuous sucrose density gradient centrifugation

Rabbit iris Albino and nonalbino rabbit irides were incubated 14 in physiological salt solution with C-imipramine, 12.8

nCi/ml (2.5 X 10 — 6 M) for two hours, then rinsed twice in 5 ml of an isotonic (0.32 M) sucrose solution for three minutes. Excess radioactivity was blotted on a filter paper and the tissue homogenized in a glass Teflon homogenizer for 2 minutes with 4 ml of the isotonic sucrose solution. The total homogenate was layered on a discontin­

uous sucrose density gradient (1.0, 1.5, 2.0, and 2.5 M sucrose solution, 2.5 ml each). The gradient tubes were

centrifuged in a swinging bucket (Beckman Rotor SW36) at

105,000 Xg for one hour in a Beckman Model L ultracentri­

fuge. The sucrose layers were carefully separated using

Pasteur pipettes and the volumes of the fractions intern­ ally equalized using demineralized double distilled water.

Proteins were precipitated by adding 1 ml of 10% trichloro­ acetic acid solution to each fraction and removed by

centrifuging in a desk model centrifuge for ten minutes. The precipitate was subsequently washed twice with 0.5 ml of a 5% trichloroacetic acid solution. The supernatants were collected for determination of radioactivity. Protein content of the tissue was determined in the precipitate. Rabbit Retina Albino and nonalbino rabbits retinas were isolated and homogenized in 2.5 ml of cold physiological salt solution in a glass Teflon homogenizer, with 8 to 10 strokes, at 600 rpm. The homogenate was incubated with o —6 H-chlorpromazine, 244.2 nCi/ml (2.5 X 10 M) for one hour, under constant oxygenation at 37°C. After the incubation period the preparation was centrifuged at 43,500 Xg, at 0-4°C, for 30 minutes. The pellet was rinsed twice with 5 ml each of 0.32 M sucrose solution and was uniformly suspended in 4 ml of the isoosmotic sucrose solution. The suspension was layered on the top of a discontinuous sucrose density gradient prepared as described for rabbit iris. The rest of the procedure was similar to that described for rabbit iris.

Human Brain Pigmented tissue of substantia nigra from two different samples, weighing 0.5 to 0.6 g, or equal amount of a secretion of superior cerebellar peduncles were homogenized in a 0.32 M sucrose-potassium phosphate buffer solution, pH 7.4, at 600 rpm, in a glass Teflon homogenizer, with 8 to 10 strokes. The concentration of the tissue in the homogenate was 0.1 g/ml . The homogenate was sub­ jected to sonication on ice for five minutes, using a 20% relative output of a Sonic 300 dismembrator. Equal amounts 97 of homogenates from substantia nigra and cerebellar ped­ uncles were incubated at 37°C, under constant oxygenation in a Dubnoff incubator shaking at 100 rpm. After a 30 minute incubation period equivalent concentrations of 14 C-imipramine, 3 H-haloperidol, or 3 H-chlorpromazine were added to both preparations, and the incubation continued for an additional hour. The preparation was centrifuged at 43,500 Xg, 0-4°C, for 30 minutes. The pellet was rinsed twice with 5 ml each of 0.32 M sucrose solution and it was uniformly suspended in 5 ml of the same isoosmotic solution. The suspension was layered on a discontinuous sucrose density gradient (1.0, 1.5, and 2.0 M sucrose solutions; 3 ml each). The rest of the procedure was carried out as mentioned above for rabbit iris. The concentration of the drugs used was 6.25 X

10"6 M and the respective specific activity for each drug used was: 14 C-imipramine, 56.3 nCi/ml; 3 H-haloperidol, 3 56.3 nCi/ml; and H-chlorpromazine, 610.6 nCi/ml. Counting efficiency for ^Carbon and tritium was 72-74% and 23-29%, respectively, as established by automatic external standard­ ization. Protein was determined in the different fractions of the sucrose density gradient according to the method of

Lowry et al. (1951). 98

Preparation of bovine iris melanin granules Bovine irides from ten eyes were isolated, weighed,

cut in small pieces and placed in a 50 ml glass homogenizer

container (Arthur Thomas Co., Philadelphia, Pa.) with a loosely fitted Teflon pestle. Twenty ml of 0.1 M potassium

phosphate buffer, pH 7.4, were added. The Teflon pestle

was used in brief strokes without disintegrating the tissue. The extract containing melanin granules was passed through a silk cloth (Mesh #10XX) into centrifuge tubes. The process was repeated until the iris stroma, originally brown to black in color, became gray and the supernatant did not take additional pigment. In order to remove large

tissue particles the supernatant containing the pigment granules was centrifuged for five minutes at the lowest setting of an international model HN centrifuge. The super­

natant from the low speed centrifugation was centrifuged at

5,900 Xg at 0°C for 10 minutes in a refrigerated RC2-B SorVall centrifuge. The supernatant was discarded and the

pellet resuspended in potassium phosphate buffer. This process was repeated ten times to obtain a clear super­

natant. The weight of the residue was determined and a 250 mg/ml suspension of granules was prepared in potassium phosphate buffer. The purity of the material was examined by light microscopy and it mainly contained granules. The preparation was refrigerated and used after 24 hours. In 99 a given experiment, granules from 10 irides were used.

The average weight of 10 irides was approximately 3.55 g and the average yield of granules from ten irides was

21%.

Preparation of synthetic melanins

L-dopa melanin Synthetic L-dopa melanin was prepared according to a slight modification of the procedure described by Potts (1964) and Binns et al.(1970). Ten grams of L-dopa were dissolved in 3 L of 0.1 M potassium phosphate buffer

(pH 7.4) and 66 mg of mushroom tyrosinase were added to the solution. The solution was continuously stirred in a water bath at 37°C for four hours. Then, it was gassed with 95% oxygen and 5% carbon dioxide at room temperature for 90 minutes. Stirring was continued at 37°C for four hours. The preparation was left overnight at room tempera­ ture with continuing bubbling of oxygen. The insoluble melanin formed was centrifuged at 19,000 rpm and the residue was washed several times with distilled water. The melanin residue was finally resuspended in distilled water and freeze dried. The synthetic melanin obtained was kept under refrigeration until used. Synthetic L-dopa melanin was prepared by Mr. Raman Baweja (Shimada et al.,

1976). 100

Synthetic L-a-methyldopa melanin and D-a-methyldopa melanins These melanins were prepared following a similar procedure to that described for L-dopa melanin, except

2.5 g of either isomer of a-methyldopa was dissolved in 400 ml of 0.1 M potassium phosphate buffer (pH 6.9, optimum pH for the enzyme). Twenty five mg of mushroom tyrosinase were added.

Synthetic dopamine melanin Dopamine melanin was prepared by dissolving 2.5 g of dopamine HCl in 400 ml of 0.1 M potassium phosphate buffer (pH 6.9, optimum pH for peroxidase) and then adding

100 mg of horseradish peroxidase. The procedure followed was the same as that described for synthetic L-dopa melanin. A summary of the yield of melanins is presented in table 3.

Binding of 14 C-imipramine and 3 H chlorpromazine to bovine iris melanin granules Each 10 ml Erlenmeyer incubation flask containing 5, 10, 25 and 150 mg of granules in 2.5 ml of 0.1 M potassium phosphate buffer, pH 7.4, was placed in a Dubnoff incubator shaking at 100 rpm, at 37°C. After a 30 minute incubation period, "^C-imipramine, 24.7 nCi/ml (2.5 X 10 ^) was added and the incubation continued for an additional 60 minutes.

The suspensions of granules were centrifuged at 43,500 Xg, at TABLE 4

PREPARATION OF SYNTHETIC MELANINS

MELANIN SUBSTRATE r g ENZYME r g YIELD

L-DOPA melanina L-DOPA , 10.0 Tyrosinase , 0.066 14.1 % , . b Dopamine melanin Dopamine , 2.5 Peroxidase , 0.100 26.0 %

L-a-methyl-DOPA melanin L-a-methyl-DOPA, 2.5 Tyrosinase , 0.025 26.0 %

D-a-methyl-DOPA melanin D-a-methyl-DOPA, 2.5 Tyrosinase , 0.025 15.6 %

aPrepared in Potassium Phosphate Buffer pH 7.4

Prepared in Potassium Phosphate Buffer pH 6.9 101 102 0-4°C, in a refrigerated RC2-B Sorvall centrifuge for 30 minutes. Radioactivity in the supernatant was determined by liquid scintillation spectrometry in a mixture of 1 ml of distilled water and 13 ml of liquid scintillation solution (ACS, Amersham/Searle). The amount of drug re­ tained by the granules was determined by difference from a standard run under the same experimental conditions. The procedure for 3 H-chlorpromazme binding to melanin granules was similar to that described for imi- pramine except that 75 mg was the highest amount of gran- 3 ules used and the specific activity of H-chlorpromazme

— 6 was 244.2 nCi/ml (2.5 X 10 M). The equilibrium time was reduced to 30 minutes. Affinity and capacity for binding were calculated as described by Shimada et al. (1976).

Binding of 14 C-imipramine and 3 H- chlorpromazine to synthetic melanins Two, 4, 6, or 8 mg of synthetic L-dopa melanin were placed in each of four Erlenmeyer flasks containing 2.5 ml of potassium phosphate buffer, pH 7.4. After a 30 minute preincubation period ^C-imipramine was added to each flask —6 for a final concentration of 2.5 X 10 M (12.8 nCi/ml). Incubation was continued for an additional hour. The suspensions of melanin were centrifuged at 43,500 Xg, 103 0-4°C, for 30 minutes. Radioactivity was determined in

1 ml of the supernatant as described above for melanin granules. The amount of radioactivity retained by melanin was determined by difference from a standard run under the same experimental conditions. 3H-Chlorpromazine binding to four synthetic melanins was studied. The synthetic melanins used were: L-dopa, dopamine, L-a-methyldopa, and D-a-methyldopa melanins. One, 2, 4, 6 , and 8 mg of the melanins were placed in each of five Erlenmeyer flasks containing 2.5 ml of potassium phosphate buffer. After the preincubation 3 -5 period H-chlorpromazine, 244.2 nCi/ml (7.5 X 10 M) was added and the incubation continued for 30 additional min­ utes. After incubation, the melanin suspensions were cen­ trifuged and radioactivity was determined in the super­ natant. The amount of radioactivity retained by melanin was determined as described above. Affinity and capacity for binding were calculated as described by Shimada et al. (1976).

3 Binding of H-chlorpromazine to melanin granules and synthetic L-dopa melanin in the presence of other drugs The procedure followed was similar to that described 3 above for binding of H-chlorpromazine to both melanin granules and synthetic melanins. After a 30 minute pre­ incubation period, an appropriate amount of unlabelled 104 drugs was added to give a final concentration of 7.5 X

10 and incubation continued for 30 additional minutes.

3 — 6 At this time, H-chlorpromazine, 244.2 nCi/ml (2.5 X 10 M) was added to the preparation and incubation with the labelled drug continued for an additional 30 minutes. The suspensions were centrifuged and radioactivity was deter­ mined in the supernatant. The amount of radioactivity retained by the granules was determined as described above. 3 The proportion of H-chlorpromazine:unlabelled drug was

1:30. 3 For the study of the binding of H-chlorpromazine to L-dopa melanin an amount of the unlabelled drug to give -4 a final concentration of 7.5 X 10 M was added. After 30 3 minutes incubation with the unlabelled drug, H-chlor- -5 promazine 244.2 nCi/ml (7.5 X 10 M) was added to the preparation and the incubation with the labelled drug continued for 30 minutes. The radioactivity retained by melanin was determined as described above. The proportion 3 of H-chlorpromazine:unlabelled drug was 1:10. 3 Affinity and capacity for binding of H-chlorpro­ mazine in the presence of unlabelled drugs were calculated according to Shimada et al. (1976). 105 3 Binding of H-chlorpromazme to melanin granules extracted with organic solvents Bovine iris melanin granules were prepared as described above. The suspensions of granules originally

obtained were divided into three equal portions. One portion was extracted with ether; a second portion was extracted with acetone and the third portion was used as

control. The volume of each portion was generally 1-2 ml. The melanin granules suspensions were placed in 60 ml separation funnels and 50 ml of ether or acetone were added. The preparation was shaken for 15 to 20 minutes and was left in the organic solvent overnight. After

approximately 15 hours, the preparation was shaken for

10 minutes and then centrifuged in a RC2-B Sorvall centri­

fuge at 5,900 Xg at 0-5°C, for ten minutes. The super­ natant was discarded and the residue of granules was re­ suspended in 20 ml of potassium phosphate buffer, pH 7.4. The suspension was centrifuged again for 10 minutes.

This process was repeated three times. The supernatant was discarded, the granule residue was weighed and kept in potassium phosphate buffer until used. An aliquot of unextracted material treated in the same fashion was used as a control for the binding studies. 106

Suspensions of melanin granules containing 2, 5,

10/ 25 and 75 mg of granules total were prepared in potassium phosphate buffer. After a 30 minute preincuba- 3 tion period at 37°C, H-chlorpromazine was added to the

— 6 preparation for a final concentration of 2.5 X 10 M (244.2 nCi/ml), and the incubation continued for 30 additional minutes. The preparation was centrifuged at

43,500 Xg, 0-4°C, for 30 minutes. Determination of radio­ activity bound to the granules was done as described before,

Procedure for calculating binding constants The analysis of the binding data assumed that the binding of drugs to melanin is analogous to the adsorption of a drug on a solid exemplified by the type I Langmuir Isotherm. The amount of drug bound per milligram of melanin, r, should be related to the concentration of free drug, [°]free' by the equation

n k [D1free „ , r = ------Eq. 1 1 + k [D)free where n is the maximum number of binding sites in a class of mutually independent sites and k is a constant related to the affinity of the drug for the binding sites.

Rearrangement of equation 1 leads to

1 1 1 1 Eq. 2 nk [DIfree Equation 2 shows that a linear relationship should be obtained when 1 /r is plotted as a function of l/tD!free* The extrapolated y-axis intercept is 1/n and the slope is

1/nk. Such a treatment of the binding data permitted the calculation of capacity,n and k, affinity, for imipramine and chlorpromazine and, based on these values, a comparison of the binding of different drugs to melanin could be made since in the presence of an inhibitor of the binding, equation 1 is modified to

n k [D] r Eq. 3 1 + k[D] + k' [C] where k' is related to the affinity constant for the inhibitor and [C] is the concentration of free inhibitor.

Defining a term k app ^ as

k k app 1 + k'[C] and substituting into equation 3 one gets

r Eq. 4 or 1 1 1 1 + Eq. 5 r n nkapp 108 Thus, if the reciprocal plots obtained for the data in the presence and absence of an inhibitor intersect at the same point on the y-axis, the nature of the binding is characterized as competitive. In the case of non-competi­ tive inhibition, k remains the same as in equation 1 and n decreases; thus, reciprocal plots in the absence and presence of the inhibitor would have the same x-intercept, but different y-intercepts. The regression line parameters (slopes and intercepts) of Eqs. 2 and 5 with their respective 95% confidence intervals and the F ratio (indicative of the variance of the regression line) were obtained via com­ puter analysis. Computer generated intercepts and slopes yield capacity as 1 /y-intercept and affinity values as y-intercept/slope. For the other studies, comparisons between two means were done by using the student "t" test (Goldstein,

1964) .

Drugs

The following drugs were used: (a) Labelled drugs: 3 H-Atropme sulfate (G) Amersham/Searle S.A. 407 mCi/mM and Arlington Heights 350 mCi/mM Illinois Radiochemical Purity 98% 109

3 3 H-Chlorpromazine ( H-methyl) CEA, IRE, SORIN S.A. 37 Ci/mM (CIS) Radiochemical Purity 99% France ^H-Haloperidol- IRE, Belgium (Chlorophenyl-3- H) S.A. 10.5 Ci/mM Radiochemical Purity 97%

^C-Imipramine Amersham/Searle S.A. 4.5 mCi/mM and Arlington, Virginia 9.2 mCi/mM

(b) Unlabelled drugs: Atropine Sulfate Mallinckrodt Chemi­ cal Works, St. Louis Missouri L-a-methyl-dopa Merck Sharp & Dohme D-a-methyl-dopa Research Labs. West Point, Phila­ delphia

Carbamyl Choline Chloride Aldrich Chemical Co. Milwaukee, Wisconsin

Chlorpromazine. HC1 Smith Kline & French Labs Philadelphia

Clozapine (100-129) Sandoz Pharmaceuti­ cals Hanover, New Jersey

L-Dopa Merck Co. Rahway, New Jersey

Dopamine HCl Regis Chemical Co. 6 -OH-Dopamine. HBr Morton Groves, Illinois

Fluphenazine HCl Schering Corporation Bloomfield, New Jersey 110 Haldol (Haloperidol) McNeil Laboratories, MCN-JR-1625 Inc. Fort Washington Philadelphia

Peroxidase (Horseradish) Sigma Chemical Co. Type Crude St. Louis, Missouri

Thioridazine HCl Sandoz Pharmaceuti­ cals Hanover, New Jersey

Tropine Nutritional Biochem­ Tropic Acid ical Corporation Cleveland, Ohio Tyrosinase Type III Sigma Chemical Co. (Polyphenol Oxidase) St. Louis, Missouri (From Mushrooms)

The radiochemical purity of the labelled compounds was tested by thin layer chromatography at regular inter­ vals and no degradation of the material was detected.

Appropriate dilutions were prepared as needed and kept at

0°C until used. Freshly prepared solutions of the unla­ belled compounds were always used. Due to the insolubility of the drug in most solvents and because lactic acid is recommended by McNeil Labs. Inc. as the best solvent for the drug, haloperidol was prepared in a 2.5% lactic acid solution. Clozapine was prepared in citrate buffer con­ taining 1.2% (w/v) citric acid and 0.193% (w/v) sodium citrate dihydrate). All other solutions were prepared in saline. CHAPTER III

RESULTS

3 In vitro uptake of H- atropine by tissues Both pigmented and nonpigmented irides accumulated radioactivity when incubated with 3 H-atropine; however, the amount accumulated by the pigmented iris exceeded that accumulated by the nonpigmented iris (fig. 11). The 3 results could have been expressed as H-atropine dpm/iris; however, due to the differences in weight of the irides (20-23 mg, nonpigmented; 33-40 mg, pigmented iris), it was arbitrarily decided to normalize the results to common weight basis. Therefore, the results were expressed as 3 H-atropine dpm/25 mg iris. Since a single radiochromato­ graph peak corresponding to authentic atropine was obtained from the radioactivity extract from the iris obtained from either albino or nonalbino atropinesterase-positive animals, it is highly probable that the accumulated radio- • • 3 activity was essentially H-atropine. No differences in the accumulation of radioactivity by irides were noted when either atropinesterase-positive or enzyme-unclassified animals were used (see Appendix A).

Ill 112

3 Figure 11. A. In vitro accumulation of H-atropine (30 nCi/ml) by the pigmented and non­ pigmented rabbit iris. B. dpm per gram/dpm per milliliter (tissue/ medium) ratio for the accumulation of the drug. Rabbits were atropinesterase- positive. Each point represents an average of four to five observations with the S.E.M. t 1------1------1------1------r

30 nC/ml 20 (I0‘ *M )

Pigmented Iris n*4

*30 nC/ml Nonpigmented frit n» 4

20 30 4 0 SO 60 70

TIME (min) TIME (min)

Figure 11 114

Although the proportion of atropinesterase-negative animals in nonalbino rabbits was high (23%), only a small

percentage (8 %) of albino rabbits were atropinesterase- negative. Thus, it was impractical to study all four types of animals all the time. Where essential, experiments were verified by examining a limited number of irides from

atropinesterase-negative animals. The accumulation of 3 H-atropine (30 nCi/ml, 10 -5 M) by irides, when studied for 35 minutes in the atropinester­

ase-negative albino or nonalbino animals is shown in

table 5 : The values do not differ significantly (P> 0.05)

from those irides of atropinesterase positive animals (table 5). At any given time, after a 10-minute equilib­ ration of the drug, the ratio of radioactivity accumulated by the pigmented irides to those nonpigmented, varied 3 between 3 and 4 fold. The uptake of H-atropine by the pigmented iris follows a complex pattern. The amount of drug bound after 60 minutes was much higher than expected on the basis of a single equilibration process. The tissue/ medium ratios for the accumulation of drug were higher in the pigmented iris. 115 3 Accumulation of H-atropine by human irides and pigment epithelium Human eyes with light brown to brown color obtained from the eye bank of The Ohio State University were stored in the cold and irides were studied within 3 36 to 48 hours after death. The accumulation of H- atropine by human iris or pigment epithelium compared favorably with that of the pigmented rabbit iris. Results are summarized in table 5.

3 In vitro loss of H-atropine from tissues after repeated washings Figure 12 illustrates the loss of radioactivity when irides or stomach fundus strips were washed at 5- minute intervals. Little or no drug was lost after the repeated washing of the pigmented iris, whereas the radio­ activity was rapidly lost from the nonpigmented iris and fundus strips. Although the fundus strips accumulated less drug per unit weight than the nonpigmented iris, the rate of loss of radioactivity by these nonpigmented tissues were identical. Half of the initial radioactivity from the nonpigmented tissues was lost in approximately 14 minutes. TABLE 5 116

ACCUMULATION OF H-ATROPINE (30 nCi/ml, 10 M INCUBATED FOR 35 min) BY HUMAN IRIS, PIGMENT EPITHELIUM AND RABBIT IRIS

Tissue N dpm/25 mg of Tissue

Human

Iris (pigmented)a 6 20948 ± 5672b cl Pigment epithelium 2 26894 ±

Rabbit Pigmented iris atropinesterase- negative 4 24908 ± 4193b Nonpigmented iris atropinesterase- negative 4 6356 + 228 Pigmented iris atropinesterase- positive 4 16168 ± 2135b Nonpigmented iris atropinesterase- positive 4 5727 ± 376

a Color - light brown to brown. The accumulation was studied as described for the rabbit iris (see Methods section). The eyes were stored in cold and uptake was studied within 36 to 48 h after death. b P > 0.05 117

Figure 12. The rate of disappearance of ^H- atropine from irides and stomach fundus strips obtained from albino and nonalbino rabbits. Tissues were incubated with 30 nCi/ml of -^H-atropine (10"5m ) for 35 minutes and at regular intervals irides were transferred to the drug free physiological solution at 37°C. Ordinate represents the amount of % - atropine bound per 25 mg of iris. Tissues were obtained from atropinesterase-positive animals only. -ATROPINE dpmxlO/25 mg TISSUE 0.01 0.1 0 20 40 IE (min) TIME iue 12 Figure Albino rabbits 60 Nonalbino rabbits =4-5 4 n= (atropinesterase positive) 80 Pgetd Iris ^Pigmented NonpigmentedIris arpnseae positive)(atropinesterase 0 10 140 120 100

HStomach J Strips Fundus

119 3 Ocular penetration of H-atropine 3 After topical application of H-atropine, absorp­

tion of the drug into the aqueous humor and the iris was

studied up to 45 minutes (fig. 13). Only 0.3 and 0.2% of

the total radioactivity was found in the aqueous humor of the nonpigmented and pigmented rabbit eye, respectively. Our results are consistent with similar findings of pene- 14 tration of C-atropine in albino rabbit eye (James and Stiles, 1959). Either pigmented or nonpigmented iris accumulated only 0.16% of the drug. If it is assumed that the accumulated drug was not metabolized, only 3 ng of the drug were accumulated by either type of iris. At no time was the accumulation of the drug by the pigmented or non­ pigmented iris significantly different. The drug trans­ ported across the corner (in aqueous humor and iris) of albino and nonalbino rabbits eye was approximately the same (P>0.05). The aqueous humor and iris of the contra­ lateral eye accumulated very small amounts of radioactivity.

Antimuscarinic effects of atropine in the pigmented and nonpigmented rabbit iris and fundus strips Data obtained either from iris or fundus strips in the presence of increasing concentrations of atropine showed a parallel shift of the dose-response curves of the agonist carbachol to the right (fig. 14). In the pig­ mented iris a high dose of atropine produced significantly Figure 13. In vivo rate of uptake of radio­ activity after the topical application of 0 . 1 ml of 2 % atropine sulfate solution, by the aqueous humor and irides of albino and nonalbino rabbits. At any given time the total radioactivity found in the aqueous humor and iris was not significantly differ­ ent in both types of animals. Each point corres­ ponds to the mean of five observations with the S.E.M. dpm/IRIS OR AQUEOUS HUMOR SAMPLE 1000 50 - 1500 0 - 500 - Harpn S4, . m 2 Solution 2% ml 0.1 , S04 3H-atropine [0.1 ml = 5.32 x I05 dpm xI05 ml[0.1] 5.32 = imne Iris Pigmented qeu hmrfo nnlio rabbits nonalbino from humor Aqueous iue 13 Figure IE (min) TIME opgetd Iris Nonpigmented Aqueous humor from humor Aqueous 30 lio rabbits albino

~ 0.27 ~ 0. 8 .1 -0 - 0.09 - 0

% % OF TOTAL RADIOACTIVITY H* to 122

Figure 14. Dose-response curves of carbachol obtained in isolated pigmented and nonpigmented rabbit iris and fundus strips determined in the pre­ sence and absence (control) of atropine. A. Dose-response curves of carbachol obtained in isolated iris sphincter of albino and nonalbino rabbits. The tissue was incubated with atro­ pine, 2 X 10”^M, for 60 minutes and a second dose response curve to carbachol in the presence of the antagonist was obtained. Note the greater shift to right obtained with the nonpigmented rabbit iris. B. Dose-response curves of carbachol obtained in isolated fundus strips of albino and nonalbino rabbits. The tissue was incubated with atro­ pine, 2 X 10-®M, for 60 minutes and a second dose-response curve to carbachol in the presence of the antagonist was obtained. % MAXIMUM CONTRACTION 100 0 4 0 6 0 8 20 0 - 0 " CONTROL- NON-PIGMENTED IRISNON-PIGMENTED IMNE II Ty T PIGMENTED IRIS n = 5 ® k

‘ATROPINE 2XI0"8M WITH ABCO [ CARBACHOL iue 14 Figure CONTROL - m 0'7 I0 ] 10 - .1 ...i i. i SOAH STRIPS STOMACH ABN RABBITALBINO 0 - 0 i j u l NNABN RABBIT NON-ALBINO • - 10 I

5 - i...i IH ATROPINE WITH . i i i i.m m in n=5 xO"eM 2xiO 10 4 -

10

- 123 smaller dose-ratios than the ratio obtained in the non­ pigmented iris. PA 2 plots are illustrated in fig. 15. EDj-g values of carbachol from the two types of irides were not significantly different (P> 0.05). The dose-response curves of carbachol when repeated in the cpntralateral iris, without the antagonist, were shifted to the right by approximately 0.2 log units. The dose ratios were cor­ rected for the sensitivity changes. The slopes of the lines corresponding to the PA 2 plots obtained from non­ pigmented iris or fundus strips were close to the theoretical value of 1 (Arunlakshana and Schild, 1959).

However, the value from the pigmented irides did not give a straight line. When there are sites of loss, either for agonist or antagonist, unusual slopes for PA 2 plots are expected (Furchgott, 1972). On the various tissues examined, the PA2 values of atropine varied from 8 . 8 8 to 8.58, and these values are approximately the same as those observed for chinchilla rabbit iris sphincter and guinea- pig ileum (Taylor, 1974; Arunlakshana and Schild, 1959).

Recovery from Atropine Blockade

The sensitivity to ED^q concentration of carbachol from the nonpigmented iris was rapidly regained whereas 1 that from the pigmented rabbit iris was very slowly re­ covered. Results are summarized in fig. 16. The sensi­ tivity of the contralateral iris, which served as control, 125

Figure 15. Arunlakshana and Schild (19 59) plot of log (DR-1) of the agonist carbachol against atropine obtained from the tissues of albino and non­ albino rabbits. The slopes of the pA 2 plots from the nonpigmented iris, the fundus strip from albino rabbit and the fundus strip from nonalbino rabbit were 0.97, 0.95 and 1.02, respectively. (The slope of the regres­ sion line, if drawn, for the data from the pigmented iris is 0.84). As illustrated, all lines were drawn by unaided eye. Incubation of atropine with tissues was 60 minutes. P values compare the significance be­ tween two points at the same concentration of the antagonist. \

3 TTTTT]------1— r~rTT'n i|------1— Rabbit Iris Sphincter Rabbit Stomach Fundus Strip

P>0.05, 2 Nonpfgmented Iris n - 5 - 6----- Nonolbino n » 5 -6 -

P<0.005,

P>0.05, P>0.05 O

m i l l

ATROPINE (MOLAR)

Figure 15 Figure 16. In vitro rate of recovery from atropine blockade in the pigmented and nonpigmented iris obtained from atropinesterase-positive rabbits. The tissue was incubated with the antagonist for 60 minutes and the response to EDgn dose of car­ bachol (10“5m ) was tested at regular intervals (See Methods for recording of the tension changes of the iris sphincter in vitro). Note that recovery of the blockade is very fast in the nonpigmented iris. % % BLOCKADE 200 100 2 0 0 6 0 3 ) 5 ( iue 16 Figure IE (min) TIME ■ Atropine I x 10 '5 M > 105 min 105 > min 5 10 > M '5 x 10 I 7 M " l0 Atropine x 2 ■ Atropine • Q Atropine I x I0 '! M 5 4 min 4 5 M '! x I0 I Atropine Q o Atropine 2 x l0 '7 M 15 min 15 M '7 l0 x 2 Atropine o imne Iris: Pigmented t Iris: Nonpigmented ) 5 ( 0 8 2 Recovery - ) 5 ( * ) 4 ( 120 129 changed with time. In a series of experiments, the average decline in sensitivity of the nonpigmented iris after four repeated experiments was 2 0 % ± 6 %, whereas by the pigmented iris was 32% ± 7%. Since these two values were not signif­ icantly different, no corrections were made for the changes in sensitivity of the tissue during the course of these experiments.

In vivo mydriatic effects of atropine

Although no differences in the onset of action of atropine in the four types of rabbits could be observed, differences in the duration of action were very clear (fig. 17). Prior to the application of atropine, the pupillary diameter in the albino and nonalbino animals differed by 1.5 to 2.0 mm; the pigmented iris did not constrict in response to light as much as the nonpigmented iris. The duration of atropine mydriasis in the four types of rabbits can be arranged as: nonalbino atropinesterase- negative > albino atropinesterase-negative > nonalbino atropinesterase-positive > albino atropinesterase-positive.

At the end of the 96th hour after the application of atropine, the animals were killed and the sensitivity of the irides to carbachol in vitro was tested. The results are presented in table 6 . Compared to control, only the EDj-q of carbachol from the atropine treated, pigmented, atropinesterase-negative animals was significantly shifted Figure 17. Onset and duration of atropine mydriasis in albino and nonalbino rabbits. Atro­ pine sulfate (2 %, 0 . 1 ml) was instilled in the conjunctival sac at zero time and the response to light was measured at regular intervals. The time required for half of the maximal effect is indicated with the corresponding S.E.M. At the end of 9 6 hours the animals were killed and in vitro sensi­ tivity of the iris sphincter to carbachol was tested (table 6 ). Note the longest duration of atropine mydriatic effect in atropinesterase-negative non­ albino rabbits. RABBITS

Nonolblno otroplnose,-ve

t*296hr

Albino otropinase,-ve

t t =29.7 hr

Nonalbino otropinase.+ve

n=4

n=4 "'Si o~ — I »" n=4 | Atropine Albino otropinose,+ve

I Ig-J I____ I____ I____ I---- 1---- 1— I I I I I 0 I 5 10 20 30 40 50 60 70 80 90 100 TIME (hours) 1 3 1

Figure 17 TABLE 6

SENSITIVITY OF ISOLATED IRIS SPHINCTER TO CARBACHOL FROM ATROPINE TREATED (96 H BEFORE) AND UNTREATED RABBITS

Negative Log Molar ED^q with SEM

Type of Rabbit Atropinesterase Control N Atropine treated N P

Albino positive 5.63 ± 0.13 4 5.76 ± 0.16 4 > 0.05

Non-albino positive 5.79 ± 0.06 5 5.96 ± 0.13 3C > 0.05

Albino negative 5.72 ± 0.16 5 5.59 ± 0.12 3° > 0.05

Non-albino negative 5.72 ± 0.07 9 5.06b± 0.14 3C < 0.01

0.1 ml of 2% atropine was applied topically at 0 time and iris sphincter isolated 96 h after (refer to Fig. 17 for in vivo effects).

k The value also differs significantly (P < 0.05) from other values obtained from atropine treated tissues.

Because of some technical reasons only three out of four animal irides were tested. 133 to the right. The amount of drug retained by the tissue corresponds to a concentration of approximately 2 X 10“8m .

Fig. 18 illustrates the results obtained when % - atropine was applied topically to the eyes of albino and nonalbino atropinesterase-negative rabbits. The duration of the atropine mydriatic effects is similar to that ob- 3 tained with the first group of animals. The amount of H- atropine retained by the pigmented irides was significantly higher than that retained by the nonpigmented iris (P> 0.05), with the pigmented iris retaining about eight times more •%-atropine than the nonpigmented irides. The total amount of atropine retained per iris 96 hours after topical appli­ cation was 3.48 ng. Data if fig. 19 indicates that the total radio­ activity retained by the pigmented iris of atropinesterase- negative rabbits 96 hours after topical application is mainly ^n-atropine as determined by thin layer chromato­ graphy. In these experiments ^H-atropine was applied in both eyes. Half time of the atropine mydriatic effect was greater than 96 hours, as in the previous experiments in vivo.

Binding of ^C-imipramine by isolated rabbit iris

Both pigmented and nonpigmented rabbit irides accumulated radioactivity when incubated with ^C-imipra- mine. However, the amount of radioactivity accumulated by the pigmented iris (fig. 2 0 ) exceeded that accumulated by Figure 18. Duration of atropine mydriatic effect after instillation of 0 . 1 ml of 2 % ^H- atropine (0.52 uCi total) in atropinesterase- negative rabbits. Animals were killed at the 96th hour and the radioactivity remaining in the irides was determined. As compared to the non­ pigmented iris, the pigmented iris retains a sig­ nificantly higher (P< 0.05) amount of the drug. PUPIL DIAMETER (mm) 0 1 4 8 9 5 6 2 3 0 7 -toie( ) % (2 H-Atropine 20 Rabbits(Atropinesterase Negative) 0 4 TIME (hr) 0 8 0 6 iue 18 Figure Albino Nonalbino |2=43. 57 .5 6 ± .5 3 = 4 t|/2 I n=4 100 2>96hv 4 * n

o> after 96hrs after Radioactivity in iris Rabbits Albino Nonalbino Rabbits

5 3 1 136

Figure 1 9 . Thin layer chromatography of iris extract from pigmented atropinesterase-negative rabbits treated 96 hours before with 0.1 ml of a 2% 3H-atropine solution (46.29 yCi/0.1 ml 3H-atropine). Although the reference 3 H-atropine contained a small amount of impurity corresponding to R f 0.31, the iris extract gave only a single peak radioactivity corresponding to the main peak of atropine. H H CPM x 10 200 I 50 0 1.4 Impurity % ehlctt :45 Methylacetate Ammonia (2 5 % ) •• ) % 20 5 Ammonia(2 spoao: 35 Isopropanol: ovn SystemSolvent Rf 0.31 Rf Distance From Origin (cm) Origin From Distance

5

Slc Gl lt ) Plate Gel (Silica iue 19 Figure Atropine Reference 10 Rf 0.76 Rf Rf 0.79 Rf

rs Extract Iris 15 137 138

Figure 20. In vitro accumulation of -^C- imipramine by the pigmented and nonpigmented rabbit iris. The concentrations of l^C-imipramine used were: 1.28 nCi/ml (2.5 X 10“^M), 4.26 nCi/ml (7.5 X 10”7m ) and 12.80 nCi/ml (2.5 X 10"%) . Each point represents the mean ± S.E.M. of 4-7 observations. 139

RABBIT IRIS NONPIGMENTED 0 - - 0 36 PIGMENTED » - • n = 4 - 7

O X 2 4 2 a. a LU 20 12.80 nCi/mJI Z 2 CC< 161C CL «§ ♦O 12

l.28nCi/mH

3 0 6 0 !90 120 140 TIME (min)

Figure 20 140 the nonpigmented one. Regardless of the concentration of ^C-imipramine used, after 15 minutes of equilibration the ratio of radioactivity accumulated by the pigmented iris to the nonpigmented iris varied between 1.5 and 2.0. Equilibrium appears to be reached at 120 minutes, although the plateau is almost complete after 60 minutes of incuba­ tion. In both types of irides the tissue/medium (T/M) ratios were highest at the lowest concentration of imipra- mine. When the concentration of the drug was increased the ratios decreased. In perfused lungs, a decrease in T/M ratios with increasing concentrations of imipramine were observed by Junod (1972). Thus, the process of accumulation appears to be a saturable one.

In vitro loss of 14 C-imipramine from isolated rabbit iris Fig. 21 illustrates the loss of radioactivity from pigmented and nonpigmented rabbit irides when washed with drug-free physiological salt solution at 5-minute intervals.

The loss of radioactivity from both types of irides seems to follow the same pattern; however, the amount of radio­ activity retained by the pigmented iris is, at all times, greater than that retained by the nonpigmented ones. The loss of radioactivity seems to stabilize between 60 and

120 minutes. At these times, the radioactivity retained by the pigmented iris is approximately 2.5 times greater than that retained by the nonpigmented ones. Half of the 141

Figure 21. Rate of disappearance of -*-4 C- imipramine from isolated pigmented and nonpigmented rabbit iris. Tissues were incubated with 12.8 nCi/ml (2.5 X 10“6m) of imipramine for one hour and at five- minute intervals the irides were transferred to drug- free physiological solution and the amount of drug remaining in the tissue was determined. The ordinate represents the amount of l4 C-bound per iris. PIGMENTED

NONPIGMENTED

40 60 80 100 120 TIME (min)

Figure 21 143 initial radioactivity retained by the nonpigmented tissue is lost in approximately 14 minutes.

Antimuscarinic effects of imipramine in the pigmented and nonpigmented rabbit iris The parallel shifts of the dose-response curves of the agonist by imipramine indicate a competitive blockade

(fig. 22). Based on K„ values, imipramine appears to be a more effective blocker in the nonpigmented iris than in the pigmented one, by a factor of 23.

14 Binding of C-imipramine to brain homogenates When incubated with 14 C-imipramine the substantia nigra homogenate accumulated more drug than the cortex homogenate. At higher amounts of tissue homogenates, the accumulation of labelled drug by the homogenate of sub­ stantia nigra was significantly higher than that from the cortex (fig. 23).

14 Localization of C-imipramine in isolated rabbit iris by discontinuous sucrose density gradient centrifugation It was found that the heaviest melanin granules settle at the bottom of the density gradient. A small amount of lighter granules remains suspended in the 2.5 M sucrose layer. The distribution of radioactivity along the gradient follows different patterns in each type of 144

Figure 22. Dose-response curve of carbachol obtained in isolated pigmented and nonpigmented rabbit iris determined in the presence and absence (control) of imipramine. The tissue was incubated with imipramine, 1 X 10“% , for 60 minutes and a second dose-response curve to carbachol was obtained in the presence of the drug. The dissociation constant (Kg) was calcu­ lated according to the method of Furchgott (1967).

? CONTRACTION (%) 100 40 0 6 0 8 20 n=5 Control IMNE NONPIGMENTED PIGMENTED r« kQ=3.8 Dose-Ratio ABT RI SPHINCTER IS IR RABBIT With 1 X Imipramine 3.6 0 M 10 CARBACHOL(M)

iue 22 Figure -7 n = 5 Control r® oe-Ratio Dose- 44- ) 3 -8 4 (4 0 6 Imipramine 10" 5 M

With — With

5 4 1 146

Figure 23. In vitro accumulation of -^C- imipramine by homogenates of human substantia nigra and cerebral cortex. Tissue homogenates were incubated with 12.8 nCi/ml (2.5 X 10“°M) of -^C- imipramine for 30 minutes. 147

HUMAN BRAIN **

SUBSTANTIA NIGRA 0— -O CORTEX n = 4 -5 * p

Figure 23 148 iris (fig. 24). In the nonpigmented iris the greatest 14 amount of C-imipramine is found in the lightest sucrose layers, 0.32 M and 1.0 M. Approximately 43% of the total amount of drug remains in the top gradients, whereas 29% separates in the 1.0 M sucrose layer. On the other hand, in the pigmented iris, the largest amount of drug (70% of the total bound) is found in the sediment containing melanin granules. The total amount of labelled drug present in the pigmented iris homogenate (17,333 ± 1811 dpm) is significantly greater (P< 0.05) than that in the nonpigmented iris (8918 ± 1484 dpm). The protein content of the pigmented iris was always greater than that in the nonpigmented iris (P< 0.05). The distribution of proteins in the gradient follows a similar pattern in both types of irides. The largest amount of protein is found in the top gradients (0.32 M sucrose) and in the sediment at the bottom of the 2.5 M sucrose fraction (fig. 24).

3 Localization of H-chlorpromazine in isolated rabbit retina by discon­ tinuous sucrose density gradient dentrifugation

As in the rabbit iris the heaviest melanin granules of the rabbit retina settle at the bottom of the density gradient. However, one distinct band of pigmented material separates at the 2 . 0 M sucrose layer with the pigmented 149

Figure 24. Localization of C-imipramine in isolated pigmented and nonpigmented rabbit iris by discontinuous sucrose density gradient centrifuga­ tion. Rabbit irides were incubated with 12.8 nCi/ml of 14c_j_mipramine (2 . 5 X 10“6m) for 120 minutes. Four milliliters of the iris homogenate were fractionated on a discontinuous sucrose density gradient (see Methods). In the pigmented iris, approximately seventy per cent of the total 14C- imipramine is accumulated in the pigmented fraction. Each point represents the mean ± S.E.M. of five observations. RABBIT IRIS DPM x 10 /fraction DPMxIoVfraction Sucrose 0 5 10 15

hj 0.32

2.0

2.5

;>h NONPIGMENTED PIGMENTED 2 0 n=5 n=5 0 2 Protein mg/fraction Protein mg/fraction

C-Imipramine Protein Content

Figure 24 151

rabbit retina. A less pigmented band separated at the

1.5 M sucrose layer. The distribution pattern of radio­

activity in each type of retina is different (fig. 25). 3 In the albino rabbit retina the greatest amount of H-

chlorpromazine is found in the lightest sucrose layers,

0.32 M and 1.0 M. Approximately 4 3% of the total bound drug remains in the top gradients, whereas 31% separates in the 1.0 M sucrose layer. In the pigmented rabbit retina, the largest amount of drug (41% of the total bound drug) is found in the sediment of melanin granules at the bottom of the gradient. 3 It was also found that the amount of H-chlorpro- mazine bound in the 2 . 0 M sucrose fraction (16% of the total drug bound) obtained from the pigmented tissue is significantly higher (P< 0.05) than that obtained from the corresponding fraction in the albino rabbit retina. The total amount of drug bound to the pigmented retina (3.55 ±

0.20 nM) is significantly greater (P< 0.05) than in the nonpigmented rabbit retina (2.82 ± 0.11 nM).

The subcellular distribution of proteins in various fractions of the tissue was similar (P> 0.05) in both albino and nonalbirio rabbits retina, except in the granules sediment, which contained the largest amount of protein. 152

3 Figure 25. Localization of H-chlorpromazine in isolated pigmented and nonpigmented rabbit retina by discontinuous sucrose density gradient centrifugation. The tissue was incubated in physiological salt solution with 244.2 nCi/ml (2.5 X 10“®M) of -^H-chlor- promazine for 60 minutes. Four milliliters of the retina homogenate were fractionated on a discontinu­ ous sucrose density gradient (see Methods). In the pigmented retina, approximately 40 percent of the total -^H-chlorpromazine is accumulated in the pigment fraction which sediments at the bottom of the 2.5 M sucrose. The fraction of pigmented rabbit retina which separates at the 2.0 M sucrose solution accumu­ lates significantly higher amounts of -^H-chlorproma- zine than the corresponding fraction from the nonalbino rabbit retina (P< 0.05). Each point represents the mean of five observations ± S.E.M. ^H-chlorpromazine bound is expressed as nanomoles per fraction. RABBIT RETINA CPZ nM/FRACTION CPZ nM/FRACTION 1.50 0 .5 0SUCROSE 0 .5 0 .50

V__ 0.32

2.0

V, 2.5

NON-PIGMENTED PIGMENTED 0.75 0.50 0 2 5 0.25 0.50 0.75 PROTEIN mg/FRACTION n=5 PROTEIN mg/FRACTION 3H-CHL0RPR0MAZINE (CPZ) PROTEIN CONTENT

Figure 25 154 14 Localization of C-imipramine, -^H-haloperidol and ^H-chlorproma- zine in human brain by discontin­ uous sucrose density gradient centrifugation In the human brain, the distribution of radio­ activity follows the same pattern as protein content. A distinct pigmented band, presumably of neuromelanin, con­ sistently separated at the 1.5 M sucrose fraction, in the interface with the 2 . 0 M sucrose fraction. The largest amount of drug was retained in the 14 top gradients (0.32 M sucrose). The amount of C-imipra- mine, for example, bound separated in the 1.5 M sucrose fraction of substantia nigra was found to be significantly higher (P<0.05) than in the corresponding fraction from the nonpigmented brain tissue. This pattern of binding of drug was similar for both 3 H-haloperidol and 3 H-chlorproma- zine (fig. 26, 27, and 28). Although the protein content in the 1.5 M sucrose fraction did not differ in both substantia nigra and superior 3 cerebellar peduncles, H-chlorpromazine binds 1.78 times 14 3 more than C-imipramine and 2.3 times more than H- haloperidol in this fraction. When the drug/protein ratios in the 1.5 M sucrose fraction from the superior cerebellar peduncles were compared, no differences in the binding among the drug were found (P > 0.05). In some of the samples a very small amount of pigment was observed to separate at the bottom of the 155 /

Figure 26. Localization of 14 C-imipramine in human brain by discontinuous sucrose density gradient centrifugation. Homogenates of substantia nigra and superior cerebellar peduncles were incubated with 56.3 nCi/ml (6.25 X 10“®M) of l^c-imipramine for 60 min­ utes. Five milliliters of the tissues homogenates were fractionated on a discontinuous sucrose density gradient (see Methods). The fraction of substantia nigra homogenate separating in the 1.5 M sucrose solution binds significantly higher (P < 0.05) amounts of l^c-irniprandne than the corresponding fraction in the nonpigmented superior cerebellar peduncles. Each point represents the mean of five observations ± S.E.M. 3 H-Chlorpromazine bound is expressed as nanomoles per fraction. HUMAN BRAIN nM/FRACTION nM/FRACTION 0.5 1 2 4 6 SUCROSE T 1---1-- 1--- 1— « | I I I | I I I |

0.32

2.0

—j ______i i__ i ft i i i i i i i i i SUPERIOR 0 .5 SUBSTANTIA 0.5 I 2 4 6 ' CEREBELLAR NIGRA PROTEIN mg/FRACTION PEDUNCLE PROTEIN mg/FRACTION n=5 ,4C-IMIPRAMINE PROTEIN CONTENT 6 5 1

Figure 26 3 Figure 27. Localization of H-halopendol m human brain by discontinuous sucrose density gradient centrifugation. Homogenates of substantia nigra and superior cerebellar peduncles were incubated with 56.3 nCi/ ml (6.25 X 10”^M) of ^H-haloperidol for 60 minutes. Five milliliters of the tissue homogenates were fractionated on a discontinuous sucrose density gradient (see Methods). The fraction of substantia nigra homogenate separating in the 1.5 M sucrose solution binds significantly higher (P < 0.05) amounts of ^H-haloperidol than the corresponding fraction of the nonpigmented superior cerebellar peduncles. Each point represents the mean of five observations ± S.E.M. %-Chlorpromazine bound is expressed as nanomoles per fraction. HUMAN BRAIN nM/FRACTION nM/FRACTION 0.5 0.5 1.0 2 4 6 SUCROSE — i 1----1----r-it i | i i i | i"i i |

0.32

1.0

1.5

2.0

I . 1 .... i | i i i i | i | 0 5 SUPERIOR SUBSTANTIA 0.5 I 2 4 6 CEREBELLAR NIGRA PROTEIN mg/FRACTION PEDUNCLE PROTEIN mg/FRACTION n = 5 3H-HAL0PERID0L PROTEIN CONTENT 8 5 1 Figure 28. Localization of ^H-chlorpromazine in human brain by discontinuous sucrose density gradient centrifugation. Homogenates of substantia nigra and superior cerebellar peduncles were incubated with 610.6 nCi/ ml (6.25 X 10“®M) of ^n-chiorpromazine f°r 60 minutes. Five milliliters of the tissues homogen­ ates were fractionated on a discontinuous sucrose density gradient (see Methods). The fraction of substantia nigra homogenate separating in the 1.5 M sucrose solution binds significantly higher (P < 0.05) amounts of ^H-chlorpromazine than the corresponding fraction of the nonpigmented superior cerebellar peduncles. Each point represents the mean of five observations ± S.E.M. -^H-Chlorpromazine bound is expressed as nanomoles per fraction. HUMAN BRAIN CPZ nM/FRACTION CPZ nM/FRACTION 6 4 2 1.0 0.5 0 SUCROSE 0.5

0.32

2.0

SUPERIOR SUBSTANTIA Q 6 4 2 1.0 0.5 0 0.5 1.0 2 ■ 4 6 CEREBELLAR NIGRA PROTEIN mg/FRACTION PEDUNCLE PROTEIN mg/FRACTION n=5 3H -CHLORPROMAZINE (CPZ) PROTEIN CONTENT 160

Figure 28 161 gradient. However, this was not a consistent finding. A detailed summary of these results is presented in table 7.

Binding of '^C-imipramine to bovine iris melanin granules and synthetic L-dopa melanin Fig. 29 shows the equilibrium time determination for the binding of ^C-imipramine to different concentra­ tions of bovine iris melanin granules. Equilibrium is reached in approximately 60 minutes.- Data on the affinity, capacity and the linear re­ regression parameters upon which they are based for l^C-imi- pramine binding to melanin granules and synthetic L-dopa melanin are shown in fig. 30. The capacity of melanin granules to bind 14 C-imi- pramine is about l/50th that of synthetic L-dopa melanin, but, on the other hand, the affinity of the drug for the binding sites in the granules is higher than that for the synthetic melanin by a factor of 250.

Binding of ^H-chlorpromazine to bovine iris melanin granules and synthetic melanin Fig. 31 shows the equilibrium time determination for the binding of ^H-chlorpromazine to different concen­ trations of melanin granules. Equilibrium reaches very fast. Equilibrium time was considered to be 30 minutes. When the data on binding of ^H-chlorpromazine to melanin granules was treated according to the method des­ cribed by Shimada et al. (1976), a regression line with slope 1.41 and y-axis intercept of 0.55 were obtained TABLE 7

BINDING OF RADIOLABELLED DRUGS TO HUMAN BRAIN. LOCALIZATION BY DISCONTINUOUS SUCROSE DENSITY GRADIENT CENTRIFUGATION3

3 Sucrose ^^C-Imipramine (nMoles) S-Haloperidol ( nMoles) H-Chlorpromazine ( nMoles) Fraction (M) C. Peduncles S. Nigra C. Peduncles S. Nigra C. Peduncles S. Nigra

0.32 4.760b 3.280 4.130b 2.990 2.370 1.970 ± 0.200 ± 0.370 ± 0.170 ± 0.250 ± 0.200 ± 0.070

1.00 1.690 3.061C 1.229 2.440c 1.465 2.644C ± 0.300 ± 0.310 ± 0.201 ± 0.160 ± 0.170 ± 0.290

1.50 0.120 0.363C 0.086 0.280 0.376 0.647d,e ± 0.020 ± 0.050 ± 0.019 ± 0.030 ± 0.060 ± 0.020

2.00 0.034 0.026 0.006 0.014 0.023 0.024 ± 0.005 ± 0.004 ± 0.002 ± 0.005 ± 0.006 ±0.005 g Each value represents the mean of five observations ± S.E.M. Statistically greater than the binding of drug in the corresponding fraction of substantia nigra (P < 0.05). Statistically greater than the binding in the corresponding fraction of superior cerebellar peduncles (P < 0.05). Statistically greater than the binding in the corresponding fraction of superior cerebellar peduncles (P < 0.01). 14 3 Statistically different from the binding of C-imipramine, and H-haloperidol to Fraction III (1.5 M sucrose) of the substantia nigra (P < 0.05). 162 Figure 29. Determination of equilibrium time for -^C-imipramine binding bovine iris melanin granules. Increasing concentration of melanin granules were incubated with 2.5 X 10“^M (27.2 nCi/ ml) l^c-imipramine. At fixed times radioactivity accumulated by the granules was determined. Maxi­ mum binding usually occurs in 60 minutes. C-IMIPRAMINE DPMXIO 100 150 0 5 0 0 20 0 4 MELANIN GRANULES iue 29 Figure IE mn ) (min TIME 0 6 0 8 100 120 # 5 mg 5 # 5mg I50m 25 mg 25 n =4 n = 3 n =4 164 140

165

Figure 30. Binding of l^C-imipramine to bovine iris melanin granules and synthetic L-dopa melanin. A. The regression line represents binding of 14c- imipramine to bovine iris melanin granules which were incubated with 24.7 nCi/ml (2.5 X 10“®M) of ^^C-imipramine for 60 minutes. Affinity was calculated as intercept/slope and capacity as the reciprocal of the y-axis inter­ cept (Shimada et al., 1976).

B. The regression line represents binding of 14c- imipramine to synthetic melanin which was incubated with 12.8 nCi/ml (2.5 X 10”6m ) of 14c-imipramine for 60 minutes. I/M DRUG BOUND PER mgxlO r- MELANIN GRANULES E L U N A R G N I N A L E M - r 0 4 20 0 3 J I/M FREE DRUG X 10 s 100 0 8 0 6 0 4 0 2 ______Capacity=2.5 Capacity=2.5 ' 5M- |0 X 5 = Affinity 9.8 ! ______n= 4 I ______x O ,0Moles/mg IO‘ I ______

I UJ CD C Q l iue 30 Figure 20 N I N A L E M A P O D - L C I T E H T N Y S Oo I/M FREE DRUG 10*X aaiy .l O* Moles/mg *8 lO l.llx Capacity= fiiy=354X10 "' 3 M 0 1 X 4 5 =3 Affinity n =5

166 Figure 31. Determination of equilibrium for ^H-chlorpromazine bovine iris melanin granules. In­ creasing concentrations of melanin granules were incubated with 2.5 X 10“®M (244.2 nCi/ml, ^n-chlor- promazine. At fixed times radioactivity accumulated by the granules was determined. Maximum binding usually occurs in 30 minutes. H-CHLORPROMAZINE DPMxlO 0 1 15 0 5 0 20 40 MELANIN GRANULES iue 31 Figure TIME(min) 0 6 n=4 0 8 100

lOmg 5 - 120 100 100 mg 25mg 5 5 mg 140 168

169 (fig. 34). Data related to the affinity and capacity for

^H-chlorpromazine are shown in table 8 . When compared with synthetic melanins, the capacity of melanin granules to bind ^n-chlorpromazine is about l/60th, l/90th, l/120th and l/180th that of dopamine melanin, L-dopa melanin, L-a-methyldopa melanin and D-a-methyldopa melanin, respectively. The order of capacity of synthetic melanins to bind 3 H-chlorpromazine, accordingly was: ' D-a-methyldopa melanin > L-a-methyldopa melanin > L-dopa melanin > dopamine melanin (fig. 32). On the other hand, the affinity of ^H-chlorpromazine for the binding sites in melanin granules is greater than that for L-dopa melanin by a factor of 15, and greater than the affinity for dopamine, L-a-methyldopa and D-a-methyl­ dopa melanins by a factor of 6 , 12, and 24, respectively. 3 The order of affinity of H-chlorpromazme for synthetic melanins being: Dopamine melanin> L-a-methyldopa melanin> L-dopa melanin> D-a-methyldopa melanin (fig. 32). The data also shows that the affinity of %-chlor- promazine for L-a-methyldopa melanin is two times greater than for D-a-methyldopa melanin. In all cases the slope of 3 the regression line obtained for binding of H-chlorproma­ zine to L-dopa melanin was greater than those obtained with the other synthetic melanins. A summary of the regression lines corresponding to the binding of ^H-chlorpromazine to synthetic melanins is shown in figure 33. TABLE 8 170

SUMMARY OF THE BINDING CONSTANTS OF 3 H-CHLORPROMAZINE TO BOVINE IRIS MELANIN GRANULES IN THE ABSENCE AND PRESENCE OF OTHER DRUGS

AFFINITY CAPACITY DRUG M Moles/mg granules

3 H-Chlorpromazine 3.88 X 10® 1.82 X 10 ®

• 3 P — If) H-Chlorpromazine + 8.03 X 10 5.01 X 10

3 H-Chlopromazine + 1.04 X 10® 4.14 X 10 ^ Thioridazine (1:30) ^H-Chlorpromazine + 6.94X10® 4.92 X 10 ® Haloperidol (1:30) ®H-Chlorpromazine + 4.73 X 10® 7.07 X 10 ^ Clozapine (1:30)

3 H-Chlorpromazine + 2.98 X 10® 1.42 X 10 ® 171

Figure 32. Binding of ^H-chlorpromazine to synthetic melanins. The regression lines represent the binding of ^H-chlorpromazine to synthetic melanins. Increasing amounts of synthetic melanins were incubated with 244.2 nCi/ml (7.5 X 1 0 “ 5 jy[) Qf 3H-chlorpromazine for 30 minutes. Affinity was calculated as intercept/ slope and capacity as the reciprocal of the y-axis intercept (Shimada et a^L. , 1976) . [I/M CPZ per mg MELANIN] xlO‘ T 4 - AO 0 6 0 7 0 3 0 5 20 10 L-Dopo melanin L-Dopo j CAPACITY * 1.56x10” moles/mg 0 5 5 - 0 3 5 2 0 2 15 10 A F F I N I T Y * 2 . 5 4 X I 0 ’ M- 1 i i i i i SLOPE =0.24 (0.23-0.27) n*5 Dopamine-melamn 0 5 O 5 - 5 3 0 3 5 2 0 2 15 10 CAPACITY *1.21X10 moles /mg 003 AFFINITY =599XI0 J ____ SLOPE * 0 . 1 A ( 0 1 2 - 0 . 1 6 ) o I _____ n*5 I _____ t M" [ I/M FREE C P Z ] x 10"' x ] Z P C FREE I/M [ I ____ iue 32 Figure I L-a-m ethyl-dopa melanin ethyl-dopa L-a-m 0 5 30 35 -5 0 5 - 5 3 0 3 - 5 2 0 2 15 10 J CAPACITY *2. l7xlO-'moles/mg A F F I N I T Y * 3 - 2 3 x l O ’ M ~ ' _____ © SLOPE *0.19(0.17-022) I ____ n»5 I L ' I I D-a-m ethyl-dopa melanin ethyl-dopa D-a-m 0 5 5 3 0 3 5 2 0 2 15 10 AFFINITY *1.60x10’NT* CAPACITY *3.27xlO moles Ang 1

© SLOPE =0.14(013-016} 1111 n*5 173

Figure 33. Summary of regression lines repre­ senting the binding of ^H-chlorpromazine to synthetic melanins. Suspensions of synthetic melanins were incubated with 244.2 nCi/ml (7.5 X 10-^M) 3H-chlor- promazine for 30 minutes. Affinity was calculated as intercept/slope and capacity as the reciprocal of the y-axis intercept (Shimada et al., 1976).

I 5 - [l/M CPZ p e r m g MELANIN] x |0 ‘7 40 30 60 20 50 70 lM[l/M DOPA MELANIN N A L E M A P O -D L YTEI MELANINS SYNTHETIC RE P]x| “8 |0 x CPZ] FREE Figure 33 y . 5 0 5 30 25 20 15 L ELANIN Y M H T DOPA E M - a - L IN N A L E M DOPAMINE OA MELANIN DOPA ' D- METHYL Y H T E -M -a .D

174

175

Binding of 3 H-chlorpromazine to bovine iris melanin granules in the presence of some phenothi- azines and other related drugs

Table 8 shows the affinity and capacity values for 3 H-chlorpromazine binding to bovine iris melanin granules in the absence and presence of drugs. Table 9 shows the linear regression parameters obtained from the binding data, with the corresponding 95% confidence intervals. 3 When drugs were introduced as inhibitors of H-

chlorpromazine binding to melanin granules, the analysis of the data showed that the y-axis intercept of the regres­

sion lines obtained from experiments using fluphenazine and

thioridazine (1:30) are statistically different (P>0.05)

from that obtained with 3 H-chlorpromazine alone. The inter­

cept values, with the respective confidence intervals, ob­

tained from experiments using haloperidol, clozapine or dopamine (1:30) fell within the range of that obtained 3 with H-chlorpromazine alone. In all cases the slope of the lines obtained with 3 H-chlorpromazine in the presence of any of the other drugs was always greater than that obtained from the binding of 3 H-chlorpromazine per se (table 9). Haloperidol and cloza­ pine seem to be more effective inhibitors of the binding 3 of H-chlorpromazine than fluphenazine, thioridazine or dopamine (fig. 34) . 176 TABLE 9 SUMMARY OF THE LINEAR REGRESSION PARAMETERS FOR THE LANGMUIR TREATMENT OF THE BINDING OF 3 H-CHLORPROMAZINE TO BOVINE IRIS MELANIN GRANULES IN THE ABSENCE AND PRESENCE OF OTHER DRUGS

SLOPE INTERCEPT X 10" 9 (c.i.)a (c.i.)a

H-Chlorpromazine 1.41 0.55 (1.28 - 1.55) (-0.23 - 1.33) 3 H-Chlorpromazme + 2.48b 1.99° Fluphenazine (1:30) (2.26 - 2.70) (1.34 - 2.65) 3 H-Chlorpromazme + 2.33b 2.42C Thioridazine (1:30) (2.06 - 2.59) (1.68 - 3.16) 3 H-Chlorpromazme + 2.93b 2.03 Haloperidol (1:30) (2.57 - 3.28) (0.96 - 3.12) 3 H-Chlorpromazme + 2. 99b 1.41 Clozapine (1:30) (2.62 - 3.36) (0.25 - 2.57) 3 H-Chlorpromazme + 2. 37b 0.71 Dopamine (1:30) (2.03 - 2.71) (-0.63 - 2.04)

C.I.: 95% confidence intervals. j_ 2 Statistically different from H-Chlorpromazine binding regression line slope, P< 0.05. c Statistically different from 3 H-Chlorpromazine binding regression line intercept, P<0.05. Figure 34. Summary of regression lines repre­ senting the binding of ^H-chlorpromazine to bovine iris melanin granules in the absence and presence of some phenothiazines and related drugs. Suspen­ sions of melanin granules were incubated with 244.2 nCi/ml (2.5 X 10"6m ) of 3H-chlorpromazine in the absence and presence of unlabelled drugs in a % - chlorpromazine:drug proportion of 1:30. 178

MELANIN GRANULES

WITH CLOZAPINE WITH HALOPERIDOL WITH FLUPHENAZINE WITH THIORIDAZINE // WITH DOPAMINE

CONTROL 3H -C P Z

Q> 0 7 V. 3H —CPZ: Inhibitor - 1:30

I ULl J -20 0 20 40 60 80 100 120 140 [l/M FREE CPZ] x|0"v-8

Figure 34 179 3 A comparison of the binding of H-chlorpromazine

to melanin granules both in the absence and presence of other drugs is shown in figures 35 and 36.

3 Binding of H-chlorpromazine to synthetic L-dopa melanin in the presence of some phenothiazines and other related drugs Table 10 summarizes affinity and capacity values 3 for the binding of H-chlorpromazine to the melanin m the absence and presence of some phenothiazines and related drugs. Although the ratio chlorpromazine:drug is smaller than in the case of melanin granules, the concentrations 3 used were higher. The concentrations of H-chlorpromazme -5 -4 and the inhibitors were 7.5 X 10 M and 7.5 X 10 M respec- 3 tively. This change in the concentration of H-chlorproma- zine in the incubation medium was necessary due to the high capacity of binding of L-dopa melanin. When a final con- centration of 3 H-chlorpromazine of 2.5 X 10 ""6 M was used at the lowest concentration of synthetic melanin, 2 mg, approx­ imately 95% of the drug added to the medium was already bound. Table 11 shows the linear regression parameters obtained from the binding data with the corresponding 95% confidence intervals.

Of the four drugs introduced as inhibitors of the 3 H-chlorpromazine binding to synthetic L-dopa melanin, the comparison of the y-axis intercepts of the regression lines 3 Figure 35. Binding of H-chlorpromazme to bovine iris melanin granules in the absence and presence of some phenothiazines and related drugs. 3 Regression lines represent the binding of H- chlorpromazine to bovine iris melanin granules in the absence and presence of some phenothiazines and related drugs in a 3 H-chlorpromazine:drug propor­ tion of 1:30. Melanin granules were incubated with ^H-chlorpromazine 244.2 nCi/ml (2.5 X 10“6m) for 30 minutes, after incubation with the unlabelled drug for 30 minutes. Affinity was calculated as intercept/slope and capacity as the reciprocal of the y-axis intercept (Shimada et al., 1976). 0 10 4 -20 0 20 40 60 80 10 2 10 20 0 20 40 60 80 10 2 10 20 0 20 40 60 80 10 2 140 120 100 0 8 0 6 0 4 0 2 0 0 -2 140 120 100 0 8 0 6 0 4 0 2 0 0 -2 140 120 100 0 8 0 6 0 4 0 2 0 0 2 - 140 120 100 0 8 0 6 0 4 0 2 0 0 2

I I [I/M CPZ per mg GRANULES]xlO'* 5 2 20 H I I I I I 1 I I I X / I I I I I I I I •H-CPZ:CLDZAPNE » o o n»4 :0 > 1:30 h 'H-CPZ n>6 0 'H-CPZ: HALOPERIDOL ______1:30 1:30 4 * n I I /T // I I I I EAI GRANULES MELANIN / FE CPZ]xlO'# FREE l/M [ Figure 35 0 ’H-CPZ:FLUPHENAZINE ___ 1 _____ 1:30 1:30 « / n«6 I I I I I I I CD © / 'H-CPZ -6 n ______I I \IA *H-CPZ:THIORIOAZINE __ I ______1:30 * / n*5 I I I I © ______'H-CPZ n*6 I ____ I _____

140 I 181 182

3 Figure 36. Binding of H-chlorpromazme to bovine iris melanin granules in the absence and presence of dopamine. 3 Regression lines represent the binding of H- chlorpromazine to bovine iris melanin granules in the absence and presence of dopamine in a proportion of 3 H- chlorpromazine:dopamine of 1:30. Melanin granules were incubated with 3 H-chlorpromazine 244.2 nCi/ml (2.5 X 10"®M) for 30 minutes, after incubation with unlabelled drug for 30 minutes. Affinity was calculated as inter­ cept/slope and capacity as the reciprocal of the y-axis intercept (Shimada et al., 1976). -20 [l/M CPZ per mg GRANULES] XIO'9

20 0 H -C P Z : DOPAMINE : Z P -C H 0 0 0 0 0 10 140 120 100 80 60 40 20 [l/M FREE CPZ ] XI0-8 CPZ ] FREE [l/M 1:30 EAI GRANULES MELANIN = « n=5 iue 36 Figure

CPZ P -C H n=6 3 8 1 184

TABLE 10

SUMMARY OF THE BINDING CONSTANTS OF 3 H-CHLORPROMAZINE TO SYNTHETIC L-DOPA MELANIN IN THE ABSENCE AND PRESENCE OF OTHER DRUGS

AFFINITY CAPACITY DRUG m ”^ Moles/mg Melanin

3 H-Chlorpromazine 2.55 X 10^ 1.60 X 10 ^

3 H-Chlorpromazine + 2.16 X 10^ 1.11 X 10 ^ Fluphenazine (1:10)

3*H-ChlorpromazineH-Chlorpromazine + 2.33 X 10^ 0.98 X 10 ^ Thioridazine (1:10)

3^H-ChlorpromazineH-Chlorpromazine + 1.22 X 10^ 1.62 X 10 ^ Haloperidol (1:10)

3*H-ChlorpromazineH-Chlorpromazine + 1.44 X 10^ 1.86 X 10 ^ Clozapine (1:10) 185 TABLE 11

SUMMARY OF THE LINEAR REGRESSION PARAMETERS FOR THE LANGMUIR TREATMENT OF THE BINDING OF S-CHLORPROMAZINE TO SYNTHETIC L-DOPA MELANIN IN THE ABSENCE AND PRESENCE OF OTHER DRUGS

SLOPE INTERCEPT X 10“ 7 DRUG (C.I.)a (C.I.)a

3 H-Chlorpromazme 0.25 6.26 (0.23 - 0.27) (4.38 - 8.14) 3 H-Chlorpromazine + 0.40b 8.45 Fluphenazine (1:10) (0.37 - 0.43) (6.48 - 10.42) 3 H-Chlorpromazme + 0.44b 10.18c Thioridazine (1:10) (0.41 - 0.46) (8.84 - 11.53) 3 H-Chlorpromazine + 0.51b 6.19 Haloperidol (1:10) (0.47 - 0.55) (3.87 - 8.52) 3 H-Chlorpromazme + 0.37b 5.38 Clozapine (1:10) (0.35 - 0.40) (3.80 - 6.95)

C.I.: 95% Confidence Intervals

Statistically different from H-Chlorpromazine binding regression line slope , P < 0.05

Statistically different from %-chlorpromazine binding regression line intercept, P< 0.05. 186 corresponding to haloperidol, clozapine and fluphenazine show no statistical differences among them; therefore the 3 capacity of synthetic melanin to bind H-chlorpromazme is not affected in the presence of these other drugs. However, a statistically different intercept was obtained in the presence of thioridazine. This indicates that this drug modifies the capacity of synthetic L-dopa melanin to bind H-chlorpromazine. The slope of the regression lines obtained in the presence of the binding inhibitors was always significantly greater (P<0.05) than that obtained 3 with H-chlorpromazine in the absence of the other drugs

(table 11). Haloperidol appears to be a more effective 3 inhibitor of the binding of H-chlorpromazine to synthetic

L-dopa melanin than fluphenazine, thioridazine or clozapine 3 (fig. 37). A comparison of the binding of H-chlorproma­ zine to synthetic L-dopa melanin both in the absence and presence of the other drugs is shown in figure 38.

3 Binding of H-chlorpromazme to bovine iris melanin granules extracted with organic solvents Data on the affinity and capacity of melanin gran­ ules extracted with ether or acetone is shown in table 1 2 .

The corresponding linear regression parameters are shown in table 13. When compared to control granules processed in the same fashion, the intercept of the regression line 187

Figure 37. Summary of regression lines repre­ senting the binding of 3 H-chlorpromazine to synthetic L-dopa melanin in the absence and presence of some phenothiazines and related drugs. Synthetic L-dopa melanin was incubated with 244.2 nCi/ml (7.5 X 1 0 “ 3 m) of 3 H-chlorpromazine in the absence and presence of unlabelled drugs in a 3 H-chlorpromazine:drug proportion of 1 :1 0 . [l/M CPZ per mg MELANIN] x |0 '7 40 0 8 60 30 50 70 5 lM FE CPZj cr8 ic jx Z P C FREE [l/M 0 1 iue 37 Figure -OA MELANIN L-DOPA / / // / / 5 0 5 0 35 30 25 20 15 iS ,. WT CLOZAPINE ■WITH

- PZ'.Inhibitor H-C T THIORIDAZINE ITH W IH L HE I E ZIN A EN PH FLU WITH HALOPERIDOL ITH W

188 189

3 Figure 38. Binding of H-chlorpromazine to synthetic L-dopa melanin in the absence and presence of some phenothiazines and related drugs. 3 A. Regression lines represent the binding of H- chlorpromazine to synthetic L-dopa melanin in the absence and presence of some phenothiazines and related drugs in a proportion of ^H-chlor- promazinerdrug of 1:10. Synthetic melanin was incubated with -^H-chlorpromazine, 244.2 nCi/ml (7.5 X 10“5m ) for 30 minutes after incubation with the unlabelled drug for 30 minutes. Affinity was calculated as intercept/slope and capacity as the reciprocal of the y-axis inter­ cept (Shimada et al., 1976). L-DOPA MELANIN

® © © 70 Ta X 60 z < 50 _ l UJ 40 •h-cpz 'H-CPZ 'H-CPZ •H-CPZ C7> n =5 n = 5 n=5 n*5 E 30 a)k_ CL N 2 0 a. O '*H-CPZ: HALOPERIDOL 'H-CPZ: THIORIDAZINE •H-CPZ-.FLUPHENAZINE 'H-CPZ: CLOZAPINE 10 1:10 1:10 1:10 s n=4 n=5 n * 6 /! J I I I J -5 10 15 20 25 -5 10 15 20 25 -5 0 5 10 15 20 25 -5 10 15 20 25

[l/M FREE CPZ]xlO'8

VO o Figure 38 191

TABLE 12

SUMMARY OF BINDING CONSTANTS OF 3 H-CHLORPROMAZINE TO ISOLATED BOVINE IRIS MELANIN GRANULES EXTRACTED WITH ORGANIC SOLVENTS

Affinity Capacity TYPE OF GRANULES Mm- 1 Moles/mg granules

Control 1.46 X 108 2.31 X 10" 9 (Unextracted)

Ether extracted 0.98 X 108 3.85 X 10" 9

Control (Unextracted) 0.39 X 10 8.45 X 10

Acetone extracted 1.71 X 10® 1.80 X 10~ 9 192

TABLE 13

SUMMARY OF LINEAR REGRESSION PARAMETERS FOR THE LANGMUIR TREATMENT OF THE BINDING OF 3H-CHLORPROMAZINE TO ISOLATED BOVINE IRIS MELANIN GRANULES EXTRACTED WITH ORGANIC SOLVENTS

Type of granules SLOPE INTERCEPT X 10” 9 (C.I.)a (c .i.)a

Control 2.96 0.43 (Unextracted) (2.70-3.23) ^-O.29-1.16)

Ether extracted 2.64 0.26 (2.39-2.89) (-0.52-1.04)

Control 2.89 0.11 (Unextracted) (2.63-3.16) (-0.73-0.95)

Acetone extracted 3.27 0.56 (2.65-3.89) (-0 .0 2 -0 .2 2 )

a (C.I.), 95% Confidence Intervals 193 obtained using either ether or acetone extracted granules did not differ from the control values (P > 0.05). Although there appear to be differences in the 3 affinity of H-chlorpromazine for the acetone extracted granules, as compared to the controls, no statistically significant differences were found between the regression parameters (table 13). This is probably due to experi­ mental variability, as shown by the dispersion of the points around the regression lines (fig. 39). Figure 39. Binding of -^H-chlorpromazine to bovine iris melanin granules unextracted and extracted with organic solvents. 3 A. Regression lines represent the binding of H- chlorpromazine to bovine iris melanin granules unextracted and extracted with acetone (See Methods). Melanin granules were incubated with ^H-chlorpromazine 244.2 nCi/ml (2.5 X 10" for 30 minutes. Affinity was calculated as intercept/slope and capacity as the reciprocal of the y-axis intercept (Shimada et al., 1976)

B. Regression lines represent the binding of ^H- chlorpromazine to bovine iris melanin.granules unextracted and extracted with ether (see Methods). The granules were incubated with H chlorpromazine 244.2 nCi/ml (2.5 X 10”^M) for 30 minutes. 0 2 4 6 8 10 2 0 20 0 -20 100 80 60 40 20 0 20 [I/M CPZ per mg GRANULES] xIO'9 25 20 lM RE P]xCa lM RE CPZ] x 10 FREE [l/M CPZ] xlCTa FREE [l/M CTN EXTRACTED ACETONE o • n = 5 / V OTO CONTROL CONTROL MELANIN GRANULESMELANIN iue 39 Figure TE EXTRACTED ETHER = 'V ' n=5

n =5 ,-8

CHAPTER IV

DISCUSSION

Differential Mydriatic Effects of Atropine

When compared to the nonpigmented iris, the marked accumulation of ^H-atropine by the pigmented iris is very similar to that observed with other liposoluble drugs like chlorpromazine (Potts, 1964), cocaine (Patil, 1972), ephedrine (Patil et al., 1974) and 3-phenethylamine (Patil and Jacobowitz, 1974). However, at high concentration of 3 H-atropine, the equilibrium profile for the accumulation in pigmented iris differed from the above drugs. A rapid equilibrium is reached within 30 min and when the incuba­ tion is continued there is an increased accumulation.

Since irides from atropinesterase positive and the enzyme unclassified animals show similar accumulation patterns, the unusual uptake equilibration profile in vitro may not be enzyme dependent. Many structurally unrelated drugs are accumulated by melanin granules (Potts, 1964; Patil, 1972; Patil et al., 1974) and synthetic L-dopa melanin (Shimada et al_., 1973).

Hence, the accumulation of drugs by the pigmented iris is

196 197 attributed to the binding of drugs to the melanin. Shimada 3 et al. (1973) studied the binding of H-atropine to synthe­ tic L-dopa melanin. The binding is a saturable process, with the affinity and the capacity of the drug for the binding being 0.2 X 10^M ^ and 17.9 X 10 ^ moles/mg, respectively. 3 In vitro, the rate of loss of H-atropine by the two types of irides is also similar to that observed with cocaine and ephedrine (Patil, 1972; Patil et al., 1974). In repeated washing, drugs are rapidly lost from the non- pigmented iris. The th for the loss of radioactivity from the nonpigmented iris incubated with 14 C-cocaine, (-)- 14 C- 3 ephedrine and H-atropine were 10 min, 12 min and 14 min, respectively. Despite the fact that the values for th are approximately equal, the loss of the former two drugs from the tissue follows a single exponential decline while that of the latter is quite complex. The drugs accumulated by the pigmented iris, on the other hand, are retained and only a fraction of the total drug is lost in washings. This suggests that the interaction between the drug and melanin may not be a simple, reversible process; but the binding appears to be irreversible.

Previously, several theories were proposed and discussed to explain the well-known iris color-dependent, mydriatic effects of drugs (Chen and Poth, 1929; Angenent and Koelle, 1952, 1953; Obianwu and Rand, 1965; Seidehamel

et al., 1970). The unequal effects of mydriatics are not race dependent because albino Africans respond more like

Caucasians with light color iris (Emiru, 1971) . An

atypical muscarinic receptor of the pigmented iris sphinc­

ter could explain the observed initial slow onset of mydriasis. This explanation is less likely because (i) the unequal mydriatic effects in humans or in guinea pigs are also observed with nonmuscarinic drugs like cocaine

and ephedrine (Chen and Poth, 1927; Seidehamel et al., 1970); and (ii) in vitro, the sensitivity of the pigmented or the nonpigmented iris sphincter to the quaternary, nonliposoluble muscarinic agonist carbachol in the two types of irides is identical (table 6 ). If the EDj-q re­

flects the affinity of the drug to the muscarinic recep­ tors, it can be concluded that the receptors in the two types of irides are similar. In the case of the pigmented and non-pigmented iris no differences in the pA2 values were obtained. The characteristics of the line obtained for the pigmented iris can be explained on the basis that at low concentrations of the antagonist, the gradient of the drug to the pig­ ment is very small, whereas at higher concentrations the loss of drug to the pigment makes it a less effective blocker in the pigmented iris. This observation should explain the slow onset of action of the drug in the heavily pigmented eye. Equal PA2 values of atropine on the stomach

fundus strips either obtained from albino or nonalbino rabbits emphasize the pigment-dependent effects of the drug. In vivo, however, the topical application of atro­ pine to the pigmented and the nonpigmented rabbit eye

failed to show the differences in the onset of action.

For several reasons the accurate comparisons of in vivo data in rabbits with that in humans could not be made.

(i) The rabbit cornea is only half as thick as the human cornea and, therefore, relative to that seen in humans, the rate of penetration of drugs through rabbit cornea will be faster. The fast initial drug interaction kinetics could obscure the differences in the onset of action in the two types of rabbits. (ii) The initial pupillary diameter of the pigmented and the nonpigmented eye differ. Although equal amounts of the drug are trans­ ported across the corner in both types of rabbits, the resting position of the iris sphincter could affect the initial drug effects. In humans, this factor does not exist; the initial pupillary diameter of the albino and the nonalbino Africans are equal (Emiru, 1971). (iii) In vitro the drug effects were accurately quantitated, while such a quantitation in vivo had a limitation. If the small differences in onset of action of atropine in vivo in the pigmented and the nonpigmented eye do exist, it 200 could be easily obscured by the measurement error.

The data on the duration of atropine mydriasis

are particularly interesting (fig. 17). The shortest

duration of action in albino atropinesterase positive animals could be explained by the rapid destruction of

the drug by the enzyme (Werner, 1965; Kalow, 1962;

Ecobichon and Comeau, 1970). In the nonalbino, atropin­

esterase positive animals, the enzyme will limit the

action but a part of the drug which is not destroyed by

the enzyme could still be stored by the pigment. The stored' drug could be released to produce the muscarinic block, but again the enzyme will limit the free concentra­ tion of the drug (fig. 17). In albino atropinesterase negative animals, a relatively prolonged effect of the drug is understandable as the enzyme will not destroy the

drug. These results in albino atropinesterase negative and positive animals are identical to those reported before (Werner, 1965; Kalow, 1962) . Since the longest

duration of action was observed in nonalbino atropinester-

ase-negative animals, this indicates the role of both the

enzyme and the pigment in atropine mydriasis. Data from

experiments using albino and nonalbino atropinesterase 3 negative rabbits indicated that when 2% H-atropine was applied topically, at the end of the 96th hour a signific­

ant amount of radioactivity was found in the pigmented 201 iris, whereas the nonpigmented iris retained little or no radioactivity. This radioactivity was demonstrated to be authentic atropine in a separate series of experiments on atropinase-negative nonalbino rabbits. Thus, the long duration of the mydriatic effect of the drug is explained. 3 The accumulation of H-atropine by the pigmented human iris is very similar to that accumulated by pigmented iris obtained from atropinesterase-negative rabbit (table

5). Since anatomically the pigment cells and the smooth muscle cells of the sphincter lie in very close proximity to each other (Walls, 1967), it seems logical that the drug could be slowly released from the pigment cell to produce a prolonged muscarinic block. Thus, the long duration of mydriatic effects of the drug is explained.

These prolonged effects in nonalbino atropinesterase- negative animals are very similar to the prolonged effects seen in nonalbino humans who lack the enzyme (Goodman and

Gilman, 1971; Havener, 1966). In vitro studies on the recovery of the muscarinic block by atropine from the pigmented and the nonpigmented iris also support the role of pigment in the recovery of the antimuscarinic effect

(fig. 16) . Our interpretation of differential mydriatic effects of atropine in albino and nonalbino rabbits is summarized in figure 40. Since the initial rate of ' Figure 40. Schematic representation which explains the effects of atropine in the pigmented and the nonpigmented eye. The transport of atro­ pine across the cornea in both types of eye appears to be equal. The transported drug could be partly destroyed by the atropinesterase in aqueous humour, or partly stored in the pigment cell and its con­ stituents and only a fraction of the transported drug will block the muscarinic receptor. The dura­ tion (t%) of drug effect of atropine in four types of rabbits is albino atropinesterase-positive, 3.8 h nonalbino atropinesterase-positive, 12.4 h; albino atropinesterase-negative, 29.7 h; nonalbino atropin- esterase-negative > 96 h. Thus, the relative contri bution of the enzyme and the pigment in the duration of atropine mydriasis is quite clear. The drug stored by the pigment could be slowly released to produce a prolonged block in nonalbino atropinester­ ase negative animals. The duration of action of atropine in nonalbino atropinesterase-negative animals is similar to that seen in human eye which contains melanin and lacks the enzyme. Nonpigmented Iris Atropinesterase

Atropine

Pigmented Atropinesterase Pigment Iris Cell__

Melanin Cornea Atropine

Figure 4 0 atropine transport across the cornea (fig. 13) is the same

in albino and nonalbino rabbits, when atropine is applied

topically in the eye, after absorption the following may

happen. In the albino rabbits, the atropine will interact

with muscarinic receptors, but it will also be metabolized

by atropinesterase in the aqueous humor or the serum after

it enters the circulation. In the nonalbino rabbit, on the other hand, melanin becomes an additional factor in determining the fate of atropine. After the drug is bound to melanin it may be slowly released, prolonging in this way the blockade of the muscarinic receptors. In the absence of enzymatic degradation, in the nonalbino rabbit, a significant amount of drug is retained in the pigmented

tissue. Since the pigment cells and iris muscles lie in close proximity (Walls, 1967) , slow release of the drug

from the pigment may permit a more prolonged blockade of

the muscarinic receptors of the iris, leading, therefore,

to a longer duration of the mydriatic and cycloplegic

effects.

Although the role of atropinesterase in selecting

the rabbit for research has been emphasized previously and in this study, more importantly we wish to emphasize the value of selecting nonalbino animals for ocular research because of the far greater frequency of individuals with pigmented eyes in the population at large. 20 5 Binding of 14 C-imipramine to pig­ mented and nonpigmented tissues

The results clearly indicate that, as compared to

nonpigmented iris, the pigmented one accumulates large

amounts of 14 C-imipramine. Although several tissue com­

ponents may contribute to the total accumulation of drug in the iris, data from the discontinuous sucrose density 14 gradient experiments indicate that C-imipramine is not randomly accumulated by tissue fractions. In the nonpig­ mented iris, which lacks melanin, a relatively large percent of the drug is bound by low sucrose density frac­ tions where synaptosomes separate (Whittaker et al., 1964).

Since imipramine is known to interact with the transport

site in the neurone, the high accumulation by low density- sucrose fractions is expected. In the pigmented iris, the

drug is accumulated by both and the low density sucrose fraction accumulates about 70% of the total drug bound by

the iris. Since synthetic L-dopa melanin and melanin 14 granules accumulate C-imipramine, the observed binding

in pigmented tissue must largely be to the endogenous melanin. On repeated washings very little drug is lost from the pigmented iris, while half of the total drug was lost

from the nonpigmented tissues in 14 min. The retention of the high amount of drug by the pigmented iris is very

similar to the other liposoluble drugs like cocaine, 206 ephedrine and 3-phenethy1amine (Patil, 1972; Patil and

Jacobowitz, 1974; Patil et al., 1974). However, the rate of loss of 14 C-imipramine from the nonpigmented ins is unlike the drugs which are lost from the nonpigmented iris at a single exponential rate. Considering the various tissue components with which ^C-imipramine is postulated to interact (Roth and Gillis, 1974; Hunt et al., 1975; Bickel and Borner, 1974), a complex rate of loss of the drug from the nonpigmented tissue is understandable. The affinity constant values obtained in our ex­ periment for binding of imipramine to melanin granules are comparable to those obtained in rat liver fractions

(mitochondria, microsomes, etc.) (Bickel and Steele, 1974), to whole blood, red blood cells and red blood cell mem­ branes (Bickel, 1975) and human serum albumin (Sharpies, 1975). On the other hand melanin granules showed to have a capacity to bind imipramine similar to that of rat liver 4 but it is about 1 0 times greater than that of blood elements. This fact should be considered in the light of the kinetics of imipramine in the body in relation to the onset of appearance and duration of imipramine effects. The relevancy of unequal drug binding by pigmented and nonpigmented irides in relation to the pharmacological effect was examined in the in vitro iris sphincter prepara­ tion. The antimuscarinic effect of imipramine was used 207 as a test parameter. Results indicate that imipramine is apparently a weaker antimuscarinic substance in the pig­ mented iris. If the drug is rapidly bound by the pigment, there should be a drop in the free concentration of the drug to interact with the receptor and thus the degree of receptor blockade will be small. The converse should occur in the nonpigmented eye to produce a greater degree of the receptor blockade. These results obtained with imipramine are in agreement with those obtained in our laboratory with atropine (see above). Melanins obtained from a variety of sources, including neuromelanin from the human substantia nigra, show characteristic electron spin resonance signals which indicate the presence of free radicals in the polymer (Van Woert et al., 1967). These free radicals of melanins probably interact with and retain many substances (Karreman et al., 1959; Mason et ad., 1960; Potts, 1964; Blois, 1971). In fact, Borg (1965) demonstrated the formation of imipramine free radicals. He suggests that these free radicals, as for chlorpromazine, may be related to the pharmacological actions of these drugs. On this basis a selective accumulation of imipramine in the substantia nigra was anticipated. Results from the tissue homogenate

(fig. 23) indicate a significantly higher accumulation of the drug by the substantia nigra. 208 Lindquist (1972) used autoradiographic technique 35 to demonstrate the localization of S-chlorpromazme by

the human brain neuromelanin. It was implied that the accumulation may be important in understanding the extra- pyramidal side effects produced by tricyclic psychotropic drugs. However, the accumulation does not necessarily mean a toxicity, which is determined by as yet unknown mechanisms. In any case, imipramine is known to produce mild extrapyramidal symptoms resembling those of chlorpro- mazine. After medication of imipramine, if the drug is accumulated by neuromelanin and released to produce weaker blockade of dopamine receptors of the caudate nucleus, it may explain the symptoms produced by the drug.

Although no reference to prolonged mydriatic effects of imipramine has been made, the fact that imipra­ mine preferentially accumulates in melanin emphasizes our contention of the relevance of the pigment for the distri­ bution, and therefore for the effects of drugs.

Localization of labelled drugs in human brain and rabbit retina by discontinuous sucrose density gradient centrifugation 14 C-Imipramine, 3 H-haloperxdol and 3 H-chlorproma- zine were found to selectively bind the pigmented fraction

of substantia nigra which separates at the 1.5 M sucrose layer of the density gradient. However, no differences in 209 3 binding of C-imipramine and H-haloperidol to this 3 fraction were found. H-chlorpromazine was found to bind

the pigmented fraction significantly more than haloperidol

and impramine. However, the differences in binding obtained in our experiments do not correlate with the reported severity of extrapyramidal syndrome produced by

these drugs. Although imipramine has been reported to produce only very mild extrapyramidal symptoms, haloperidol and chlorpromazine are well known for inducing Parkinson's

syndrome after prolonged use. An inverse relationship between antimuscarinic effects in the central nervous system and ability to produce Parkinson's syndrome among neuroleptic drugs has been suggested (Snyder et al., 1974). Haloperidol is known to produce the most serious extra­ pyramidal disturbances. Haloperidol has also been reported to be an excellent blocker of dopamine receptors in the central nervous system (Carlsson and Lindquist, 1963). Snyder et al. (1974) have suggested that blockade of dopamine receptors by phenothiazines and butyrophenones may explain the prominent extrapyramidal side effects of these drugs. By blocking the dopamine receptors in the corpus striatum these drugs produce a functional deficiency of dopamine. It is suggested that the Parkinsonian side- effects of phenothiazines are the result of blockade of 210 dopamine receptors in the striatum, whereas the anti­ schizophrenic action may be related to dopamine receptors in other areas of the brain. In light of the evidence presented so far on binding of phenothiazines to brain tissue (Lindquist, 1972, 1973) and degeneration of substantia nigra in patients with prolonged treatment with phenothiazines (Forrest et al., 1964; Christensen et al., 1970) and based on our experimental results, we suggest that accumulation of the drugs in neuromelanin may be a prerequisite to the appearance of the extrapyramidal symptoms. Accumulation of the drug, accompanied by "inactivation" of neuromelanin by phenothiazines followed by degeneration of the neuron have been suggested by Forrest (1974) as part of the mechanism underlying the depigmentation of substantia nigra found in patients with drug induced dyskinesia. 3 Our results on the differential binding of H- chlorpromazine in pigmented and nonpigmented rabbit retina and the selective accumulation of the drug in the pigmented fraction of the tissues indicate that drug induced retino­ pathies are not the result of a random accumulation of drug in the eye. Since albino retinas do not accumulate polycyclic compounds and do not develop retinal damage

(Meier-Ruge et al., 1966; Meier-Ruge and Cerletti, 1968), drug-induced retinotoxicity in the pigmented eye appears 211

to depend on the melanin content of the uvea.

Binding of 3 H-chlorpromazine to bovine iris melanin granules and synthetic melanins 3 Our results on binding of H-chlorpromazine to bovine iris melanin granules and synthetic melanins demon­ strated that the drug has more affinity for the binding sites in melanin granules than in synthetic melanins; 3 however, the capacity of synthetic melanins to bind H- chlorpromazine is much greater than that of melanin 3 granules. The affinity of H-chlorpromazine to bind melanin granules is approximately 1 , 0 0 0 times greater than that for rat liver cell fractions (Bickel and Steele,

1974) and human serum albumin (Sharpies, 1975). The nature of the two systems may account for the differences, but does not explain them. It may be suggested that there are more binding sites in the synthetic melanins or that they are more readily accessible to the drug if one assumes that synthetic melanins are just melanin polymers, without the probable complexity of the melanin granules. It would not be proper, however, to consider synthetic melanins 3 just simple polymers. Our experiments on binding of H- chlorpromazine to synthetic melanins suggest more than that. 212

Although one would expect that melanins synthe­ sized from L-dopa, dopamine, L-a-methyldopa or D-a-methyl- dopa might be structurally different polymers, our results suggest that the nature of the polymer may have great influence on its ability to bind other compounds. For 3 example, the capacity of dopamine melanin to bind H- chlorpromazine is smaller than that of the other melanins. Dopamine differs from the other precursors in lacking the carboxyl and the a-methyl groups, respectively. On the 3 other hand, the affinity of H-chlorpromazine for binding sites in dopamine melanin is greater than for the other synthetic melanins, and the affinity of the drug for L-a- methyldopa melanin is two times greater than for the melanin obtained from the D-isomer. In looking at the structure of the precursors used to synthesize the melanins and the corresponding intermediate compound, one could speculate that the presence and stereochemistry of the carboxyl group, as well as the presence of substituents in the side chain may determine, in part, the number or availability of binding sites in the synthetic melanins polymer (fig. 41). These speculations would agree with the hypo­ thesis of an irregular melanin polymer formed from the polymerization of intermediates, including the indole-5,6- quinone, of the melanin synthesis. 213

Figure 41. Pathway of melanin synthesis show­ ing different precursors and probable intermediates formed corresponding to the dopachrome of the L-dopa melanin synthesis according to the Raper-Mason pathway. 214

HO M e la n in 1 HO COOH COOH

L - do pa dopachrom e

HO Melanin2 HO XX °H h2

dopam ine

M elanin.

COOH H9N

L- CL -methyldopa L-O'-methyldopachrome (?)

COOH- — > Me la n in . H0'4\<>’ .0 3Xk nmv ' nc o o h H CH3 c h 3

D “ CL -methyldopa D-CK-methyldopachrome (? )

Figure 41 215

Lack of knowledge of the structure of melanin makes an analysis of binding characteristics and differ­ ences between natural and synthetic melanins on the basis of chemical structure impossible at the present time. However, a detailed study of the binding of drugs to synthetic melanins may help to understand the mechanisms of binding of drugs to natural melanins. One could assume that if the binding of drugs to melanin granules only takes place in the melanin moiety one could expect differences in the binding when one com- 3 pares the binding of H-chlorpromazine to intact melanin granules with that to melanin granules which have been subjected to extraction with organic solvents. Our results 3 on the binding of H-chlorpromazine to intact granules and granules extracted with organic solvents do not confirm these assumptions. No differences in binding between both kinds of granules were obtained. The very high liposolu- bility of the drug may account for the rapidity with which the binding equilibrium with chlorpromazine is reached in melanin granules (fig. 31), but it does not explain the lack of differences in binding when the granules are extracted with organic solvents. It has been demonstrated that phenothiazines affect membrane permeability in various biological systems (Guth and Spirtes, 1964; Domino et al., 1968; Van Woert,

1970) . These drugs reduce surface activity and they 216 accumulate and cause lysis of membranes similar to the action of detergents. If chlorpromazine produces this detergent effect in intact melanin granules, promoting its own penetration and binding, one would expect the same re­ sults when other agents such as organic solvents are used. Therefore, in granules extracted with organic solvents, binding of chlorpromazine is expected to be similar to that in intact granules as demonstrated by our results. 3 The analysis of the data on binding of H-chlor- promazine to melanin granules revealed that the phenothia- zine fluphenazine and thioridazine significantly decrease 3 the capacity of melanin granules to bind H-chlorpromazme

(P<0.05) and also the apparent affinity constant of the drug. This is suggestive that the fluphenazine and thioridazine mechanism of inhibition of chlorpromazine binding is very complex. A combination of competitive and noncompetitive inhibition seems to be taking place. Although an explanation on the basis of structure-activity relationship is attractive, our experimental design and results do not allow us to account for it. On the other hand, clozapine, haloperidol, and dopamine are demonstrated to be competitive inhibitors of 3 H-chlorpromazine binding. Haloperidol and clozapine were better inhibitors than the other drugs. Because of the different chemical structure of the compounds, generaliza­ tions about the nature of the binding are not allowed. 217 All the drugs showed ability to bind the melanin granules; 3 however, the fact that the ability to inhibit H-chlor­

promazine differs from one drug to another indicates that

other factors more than their affinity for melanin are

involved in the binding of these drugs. Physicochemical

characteristics of the drugs seem to be very important

factors in determining the binding of the drugs. 3 Binding of H-chlorpromazine to synthetic L-dopa

melanin is competitively inhibited by haloperidol, cloza­

pine, and fluphenazine. Thioridazine appears to inhibit 3 the binding of H-chlorpromazine to synthetic L-dopa melanin by a combination of competitive-noncompetitive inhibition. Haloperidol seemed to be a better inhibitor

than the other drugs. It is interesting that clozapine does not seem to be as good an inhibitor of chlorpromazine binding to synthetic L-dopa melanin as it showed to be for

binding of chlorpromazine to melanin granules. 3 These findings on the binding of H-chlorpromazine

to melanin granules and synthetic melanins in the absence

and presence of other drugs emphasizes the idea that the

binding of drugs to melanin involves more than the simple affinity of the drugs for melanin. However, although the

differences in binding among drugs and between the two systems do not seem important enough to object to synthetic melanins as suitable models for the study of binding of drugs to melanins in the body, it points out the problems 218

of extrapolating results obtained in studies with synthetic

melanins to natural melanins.

All of the results obtained in these studies point

to the fact that melanins are not inert compounds with no

specific role in the body. Due to the drastic conditions

necessary in most cases for the demonstration of the forma­

tion of free radicals, the electron paramagnetic properties of many drugs are not known. However, the fact that the stable free radical properties of melanins have been demonstrated (Mason et ciL., 1960; Van Woert et al. , 1967) and drugs such as phenothiazines have been demonstrated to form free radicals and interact with melanin forming charge- transfer complexes suggests that the formation of free radicals may be a factor determining the interaction of drugs with melanin. In fact, free radicals have been demonstrated for several drugs, imipramine, salicylic acid, opiates, etc. (Borg, 1965; 1972). These facts, together with the evidence on binding of drugs to melanin presented so far, indicate that the binding to melanins may be an important element of the pharmacokinetics of drugs. Therefore, due to the conse­ quences that the binding of drugs to melanin may provoke, and due to the fact that the.population at large have melanin, we recommend that studies on the binding of drugs be conducted in pigmented animals and in pigmented tissues for drugs with possible therapeutic use. CHAPTER V

SUMMARY

In studying the differential mydriatic effects of atropine in albino and nonalbino rabbits it was found that: In vitro experiments: a. In irides from albino and nonalbino atropinester- ase positive rabbits, the pigmented iris accumu- 3 lated larger amounts of H-atropine than the nonpigmented ones. On repeated washings little

or no radioactivity was lost from the pigmented iris, whereas in the nonpigmented iris radio­

activity was rapidly lost (t% 14 min). 3 b. Pigmented human irides accumulated H-atropine in an amount comparable to that accumulated by

pigmented iris obtained from atropinesterase-

negative rabbits. c. After topical application of a 2% solution of 3 the drug, the rate of penetration of H-atropine through the cornea was similar in both albino and nonalbino rabbits, and after 40 minutes only 0.1% of the drug was bound. 220

d. pA£ values from nonpigmented iris and fundus strips

varied between 8.58 and 8.88 with slope values close

to the theoretical value of 1. The PA2 value of atropine in the pigmented iris was 8.82, but at

higher concentrations atropine was a less effective blocker than that observed in the nonpigmented iris.

The unusual PA 2 plot obtained from tHe pigmented

iris is attributed to loss of the drug to the pigment.

e. Recovery of nonpigmented iris from atropine blockade

was faster than that of pigmented iris. The degree of atropine blockade was evaluated by introducing

an EDgg dose of catechol every fifteen minutes. In vivo experiments were conducted in atropinesterase positive or negative albino and nonalbino rabbits. It was found that: a. The order of half-times of atropine mydriasis in the four types of rabbits was: Nonalbino atropin- esterase-negative > albino atropinesterase-negative >

nonalbino atropinesterase-positive > albino atropin-

esterase-positive. In the nonalbino atropinesterase-

negative rabbits, the half-time of duration of

action of atropine was 96 hours. This value is comparable to the half-time mydriatic effect of

atropine in humans. b. In atropinesterase negative rabbits, ^H-atropine retained in the iris 96 hours after topical appli

cation was higher in the pigmented iris than in

the nonpigmented ones. With the aid of thin

layer chromatography it was confirmed that the radioactivity retained in the tissue was atropine c. The prolonged duration of atropine effects in non

albino rabbits is explained on the basis of accu­ mulation of the drug in the eye pigment from which it is slowly released onto the muscarinic receptors producing a prolonged local blockade.

14 In in vitro experiments using C-imxpramxne: a. Both pigmented and nonpigmented rabbit irides

accumulated 14 C-imipramine. The ratio of radio­ activity accumulated by the pigmented iris to

the nonpigmented one varied between 1.5 and 2.0.

The loss of radioactivity from the pigmented iris

is faster than from the pigmented one, but the

loss of radioactivity from the nonpigmented iris

does not follow a simple exponential process. b. Imipramine is a less effective blocker of muscar­ inic receptors in the pigmented rabbit iris than in the nonpigmented iris. c. When homogenates of human substantia nigra were incubated with 2 X 10 (12.8 nCi/ml) of ^ C - imipramine they accumulated significantly higher

amounts of the drug than cerebral cortex homogen­

ates . d. 14 C-imipramine accumulated by the pigmented i iris. .

was mainly localized in the heavy melanin gran­ ules fraction separated by discontinuous sucrose

density gradient centrifugation.

e. The capacity of bovine iris melanin granules to bind ^C-imipramine is about l/50th that of syn­ thetic L-dopa melanin, whereas the affinity of

the drug for the binding sites in the melanin granules is higher than that for the synthetic melanin by a factor of 250.

Albino and nonalbino rabbits retinas were incubated — 6 3 with 2.5 X 10 M (244.2 nCi/ml) H-chlorpromazine. When fractionated on a discontinuous sucrose density gradient two pigmented bands were separated and the largest amount of drug (41% of total bound) in the pigmented retina, was found in the sediment of melanin granules at the bottom of the gradient. Little or no drug was found in the corresponding nonpigmented fraction of the albino rabbit retina. In in vitro studies on binding of drugs to human 3 3 brain, C-imipramine, H-haloperidol and H-chlor- promazine were incubated with homogenates of substantia nigra and superior cerebellar peduncles and the homogen­ ates were fractionated on a discontinuous sucrose den­ sity gradient. In these experiments: a. A distinct pigmented band, presumably of neuro­

melanin, consistently separated at the 1.5 M

sucrose fraction (at the interface with the 2 . 0 M

sucrose fraction). b. The amount of "^C-imipramine, "^H-haloperidol or 3 H-chlorpromazine separated at the 1.5 M sucrose

fraction was found to be significantly higher (P<0.05) than in the corresponding fraction

from the nonpigmented superior cerebellar

peduncles homogenates. Presumably the melanin

fraction binds more drug than the corresponding fraction of nonpigmented brain tissue. 3 d. The amount of H-chlorpromazine bound to the 1.5 M sucrose fraction from substantia nigra was significantly higher (P< 0.05) than the amount of 14 C-imipramine or 3 H-haloperidol bound to the

same fraction of substantia nigra. No differ­

ences in the binding of the drugs to the 1.5 M sucrose fraction of superior cerebellar peduncles

homogenate were found. The binding of 3 H-chlorpromazine to bovine iris melanin granules and to synthetic L-dopa, dopamine, L-a-methyl- dopa and D-a-methyldopa melanins was compared. In these experiments: a. Compared to synthetic melanins the capacity of 3 bovine iris melanin granules to bind H-chlor­

promazine was l/60th, l/90th, l/120th, and l/180th that of dopamine melanin, L-dopa melanin, L-a-methyldopa melanin and D-a-methyldopa melanin, 3 respectively. The affinity of H-chlorpromazine for the binding sites in the melanin granules

is greater than for any of the synthetic melanins

used. 3 b. The capacity of synthetic melanins to bind H-

chlorpromazine was as follows: D-a-methyldopa

melanin > L - a-methyldopa melanin> L-dopa melanin 3 > dopamine melanin. The affinity of H-chlor-

promazine for synthetic melanins was: dopamine melanin > L-a -methyldopa melanin > L-dopa melanin > D-a-methyldopa melanin. c. Haloperidol and clozapine were more effective 3 inhibitors of the binding of H-chlorpromazme to melanin granules than fluphenazine or thiorid­

azine or dopamine. Haloperidol, clozapine and

dopamine behaved as competitive inhibitors of the binding, whereas fluphenazine and thiorida­ zine showed a combination of competitive-noncom­ petitive inhibition. Haloperidol, clozapine and fluphenazine behaved 3 as competitive inhibitors of the binding of H- chlorpromazine to synthetic L-dopa melanin.

Thioridazine showed a combination of competitive- noncompetitive inhibtion. . Haloperidol was dem­ onstrated to be the most effective inhibitor and clozapine the least effective. Treatment of melanin granules with organic solvents, such as ether and acetone, did not 3 affect the binding of H-chlorproraazme to the granules. The detergent effect of phenothiazines, probably already affecting the normal granules may account for the lack of differences in binding obtained in granules extracted with organic solvents as compared to control granules. APPENDIX A

3 In vitro Uptake and Loss of H-Atropine from Tissues

Uptake and loss of atropine from tissues was studied in isolated irides and stomach fundus strips from albino and nonalbino rabbits unclassified for atropinester- ase. The irides were incubated with increasing concentra­ tions of the drug. The procedure followed was the same as described in Methods for atropinesterase-positive rabbits. 3 At all concentrations of H-atropine used, pig­ mented irides accumulated significantly larger amounts of drug than the nonpigmented ones. No differences were observed when the radioactivity accumulated by the unclas­ sified rabbit irides was compared to that accumulated by the atropinesterase-positive rabbits for both albino and nonalbino strains(fig. 42). Figure 43 illustrates the loss of radioactivity from albino and nonalbino rabbit iris stomach strips. The rate of loss of radioactivity from unclassified pigmented and nonpigmented rabbit iris was found to be similar to that in the atropinesterase-positive rabbits. Little or no drug was lost from the pigmented irides, whereas in the nonpigmented iris radioactivity disappeared very rapidly 226 227

o Figure 4 2. in vitro accumulation of H- atropine by the pigmented and nonpigmented iris of atropinesterase-positive and enzyme-unclassified rabbits. The accumulation of the drug by irides obtained from serum atropinesterase-positive rabbits were not significantly different than that selected at random on unclassified animals. Each point repre­ sents an average of 4 to 5 different observations with S.E.M. 0 1 H-ATROPiNE dpm x 10725 mg IRIS 40 0 3 5 3 20 25 10 15 0 5 560 6 45 5 Arpnseae positive Atropinesterase — Nnimne Iris Nonpigmented o imne Iris Pigmented • Unclassified 4-5 - 4 n- 15 3 nc/ml 3 iue 42 Figure

IE (min) TIME 30 10 nc/ml

-at—

0 3 nc/ml 30nc/mi I' M) (IO'5 228 229

(t%, 14 min). The loss of radioactivity from stomach

fundus strips shown in this figure corresponds to atropin­

esterase-positive rabbits.

3 Accumulation of H-atropine by irides obtained from atropine pretreated rabbit eyes Atropine sulfate, 0.1 ml of 4% solution, was applied to one eye of both albino and nonalbino rabbits unclassified for atropinesterase. Twenty four hours after the application, the animals were killed and the irides 3 isolated, and incubated with H-atropine. The procedure followed was the same as described in the Methods section

for accumulation of ^H-atropine by tissues. The contra­

lateral iris was used as a control. The results from this series of experiments are 3 summarized in table 14. It can be noted that H-atropine

accumulation was reduced only in the pigmented iris.

Antimuscarinic effects of atropine on pigmented and nonpigmented rabbit iris The blocking effects of a 2 X 10 ”8 M atropine dose were tested against a single ED^q dose of carbachol in both pigmented and nonpigmented rabbit irides. The procedure followed was the same as described in Methods (Functional

experiments). The results are summarized in table 15. 3 Figure 43. Rate of disappearance of H-atro- pine from iris and stomach fundus strips obtained from atropinesterase-positive and enzyme-unclassi­ fied albino and nonalbino rabbits. 3 Tissues were incubated with 30 nCi/ml of H- atropine (10~5 M) for 35 min and at regular intervals it was transferred to the drug-free physiological salt solution at 37°C. Irides obtained from atro- pinesterase-unclassified rabbits were selected at random. Stomach fundus strips were obtained from atropinesterase-positive animals only. Each point on the graph represents the mean of 4 to 5 observa­ tions with S.E.M. 50 X -ATROPINE dpmxlO/25 mg TISSUE 0.01 20 imne Iris Pigmented (unclassified) 0 4 TIME(min) opgetd Iris Nonpigmented iue 43 Figure (unclassified) lio rabbitsAlbino 0 6

Nonalbino rabbits =4-5 - 4 n= (atropinesterase positive) 80 Pgetd Iris ^Pigmented — Nonpigmented Iris — atropinesterase positive)(

100 2 140 120

'•* Strips HStomach Fundus

231 232

TABLE 14

IN VITRO, ACCUMULATION OF 3 H-ATROPINE (30 nc/ml, 10~5 M, INCUBATED FOR 35 min) BY IRIDES FROM RABBITS TREATED TOPICALLY (24 hr before) WITH 0.1 ml OF 4% ATROPINE

dpm/25 mg iris Tissue Control* Atropine treated

Nonpigmented iris 7456 ± 1214 7228 ± 437 > 0.05

N = 6

Pigmented iris 28623 ± 3102 20334 ± 2940 > 0.01

N = 6

aContralateral iris from the same animal without the atropine treatment. 233 TABLE 15 COMPARATIVE BLOCKING EFFECTS OF ATROPINE AGAINST A SINGLE DOSE ED9q OF CARBACHOL ON THE PIGMENTED AND NON- PIGMENTED IRIS OBTAINED FROM ATROPINESTERASE POSITIVE RABBITS

Mean tension (in mg with SEM) after carbachol (10-5m ) Tissue With atropine

Control3 2 X 10"8M

Nonpigmented iris 239 ± 42 50 ± 6 ii 25

Pigmented iris 240 ± 50 190 ± 55° N = 5

Sensitivity of contralateral control iris (to carbachol) during the testing procedure was increased by 106% ( ± 8 ) and 115% ( ± 5) for the nonpigmented and pigmented iris, respectively.

Incubation time = 60 min.

P < 0.05 when compared with that from nonpig­ mented iris. APPENDIX B

Binding of 14 C-Imipramine to Isolated Irides

from 6 -Hydroxydopamine Treated Albino

and Nonalbino Rabbits

Rabbits were given two doses, 15 mg/Kg each, of

6 -hydroxydopamine (Regis Chemical Co.) at 24 hours inter­ vals. Each dose was infused during a two hour period in the ear marginal vein, using a polyethylene tubing con­ nected to a syringe. The 6 -hydroxydopamine solutions were prepared in 0.1% sodium metabisulfite. The syringe cylinder was tightly wrapped with aluminum foil as addi­ tional protection of the 6 -hydroxydopamine solution from photooxidation. At the end of each injection period, 0.1 ml of a 2 % solution of 6 -hydroxydopamine was topically applied to each eye. At the end of 48 hours the animals were killed. The iris were isolated as described by Jacobowitz (1967). The irides were incubated with 2.5 X

- 6 14 10 M (12.8 nCi/ml) of C-imipramine for 60 minutes and the loss of radioactivity from the tissue was determined as described above in the Methods section.

The results are summarized in table 16. When com­ pared to normal rabbit iris, the amount of drug retained

234 235 by the pretreated rabbit iris, no significant differences were found except for the nonpigmented rabbit iris after 15 minutes of washing. In this particular instance, the amount of drug retained by the 6 -hydroxydopamine treated rabbit iris was significantly higher than in the normal rabbits (P<0.05). At 120 minutes after washings, the amount of drug retained by both pigmented and nonpigmented iris from pre­ treated rabbits was similar to the corresponding iris in the normal group of rabbits. An explanation for the increase in retention of drug in nonpigmented iris from

6 -hydroxydopamine treated rabbits after 15 minutes of washings is not available at the present time. TABLE 16

EFFECTS OF 6 -OH DOPAMINE PRETREATMENT ON THE BINDING OF 1 4 C-IMIPRAMINE TO ISOLATED PIGMENTED AND NONPIGMENTED RABBIT IRIS

14 C-imipramine DPM/iris

5 min 15 min 35 min 60 min 1 2 0 min

Normal Nonpigmented iris 13056 ± 721 9542 ± 1054 7668 ± 660 7127 ± 712 6817 ± 1392 n = 5

Pigmented iris 25786 ± 3450 21310 ± 1 2 1 0 — 18977 ± 2933 17317 ± 3036 n = 5

6 -OH-dopamine n = 5 Nonpigmented iris 13081 ± 607 20308 ± 3238a 11927 ± 2825 6778 ± 901 6228 ± 1518

Pigmented iris 23528 ± 5706 30630 ± 3504 24021 ± 1815 15705 ±1758 18246 ± 2306 n = 5

Statistically different from normal rabbit iris (P<0.05). 236 BIBLIOGRAPHY

Abe, K., Butcher, R. W. , Nicholson, W. E., Baird, C. E., Liddle, R. A. and Liddle, G. W.: Adenosine 3',5'- monophosphate (cyclic AMP) as the mediator of the actions of melanocyte stimulating hormone (MSH). Endocrinology, 8 £: 362-368, 1969. Abe, K., Gobison, G. A., Liddle, G. W., Butcher, R. W. , Nicholson, W. E. and Baird, C.E.: Role of cyclic AMP in mediating the effects of MSH, norepinephrine and melatonin on frog skin. Endocrinology, 85, 674-682, 1969. Adams, M., Blois, M. S. and Sands, R. H.: Paramagnetic resonance spectra of some semiquinone free radicals. J. Chem. Phys., 28_: 774-776, 1958. Adler, A.: Melanin pigment in the central nervous system of vertebrates. J. Comp. Neurol., TO.1 315-329, 1939. Adler, A.: Melanin pigment in the brain of the gorilla. J. Comp. Neurol., T6_: 501-507, 1942. Allegra, S. R.: Golgi apparatus and melanogenesis. Ultrastructural study of a human cellular blue nevus. Lab. Invest., 3_4: 258-267, 1974. Anden, N., Dahlstrom, A., Fuxe, K. and Larrson, K. : Further evidence for the presence of nigro-striatal dopamine neurons in the rat. Am. J. Anat., 116; 329-333, 1965. Angenent, W. J. and Koelle, G. B.: The destruction of norepinephrine by the dopa-oxidase system of ocular tissue. Science, 116: 543-544, 1952. Angenent, W. J. and Koelle, G. B.: A possible enzymatic basis for the differential action of mydriatics on light and dark irides. J. Physiol. (London), 119: 102-117, 1953.

237 238 Arons, B., Kosek, J. C. and Forrest, I. S.: Chlorpromazine therapy in a female albino patient: Clinical, histo- chemical and biochemical observations. Life Sci., 7: 1273-1280, 1968. Arunlakshana, 0. and Schild, H. 0.: Some quantitative uses of drug antagonists. Brit. J. Pharmacol., 1_4: 48-58, 1959. Axelrod, J. and Lerner, A. B.: O-methylation in the conversion of tyrosine to melanin. Biochim. Biophys. Acta., ^7!L: 650-655, 1963.

Barbee, R. F. and Smith, W. 0 .: A comparative study of mydriatic and cycloplegic agents in human subjects without eye disease. Amer. J. Ophthal. , 44_z 617-622, 1957. Bazelon, M., Fenichel, G. M. and Randall, J.: Studies on neuromelanin. I. A melanin system in the human adult brainstem. Neurology, 3/7: 512-519, 1967. Bennet, G. A. and Hausberger, F. X.: Uber den einfluss des adrenalins auf den pigmengealt der iris bei neurogener experimenteller heterochromie. Arch. exp. Path. Pharmak., 188: 40-52, 1938. Bickel, M. H.: Binding of chlorpromazine and imipramine to red cells, albumins, lipoproteins and other blood components. J. Pharm. Pharmacol., 21_z 733-738, 1975. Bickel, M. H. and Borner, H.: Uptake, subcellular distribu­ tion and transfer processes of imipramine and its metabolites formed in rat liver perfusion systems. Naunyn-Schmiedeberg1s Arch. Pharmacol., 284: 339-352, 1974. Bickel, M. H. and Steele, J. W.: Binding of basic and acidic drugs to rat tissue subcellular fractions. Chemico. Biol. Interact., 8^: 151-162, 1974.

Binns, F . , King, J.A.G., Mishra, S. N., Percival, A., Robson, N. C., Swan, G. A. and Wagott, A.: Studies related to the chemistry of melanin. XIII. Studies on the structure of dopamine melanin. J. Chem. Soc. C.: 2063-2070, 1970. Blois, M. S.: On chlorpromazine binding in vivo. J. Invest. Dermatol., 45: 475-481, 1965. 239 Blois, M. S.: Melanin binding properties of iodoquine. J. Invest. Dermatol., 50^: 250-253, 1968.

Blois, M. S.: Physical studies of the melanins. In, Biology of Normal and Abnormal Melanocytes, T. Kawamura, T. B. Fitzpatrick and M. Seiji (Editors), pp. 125-139, University Park Press, Baltimore, 1971. Blois, M. S. and Taskovich, L.: The reversible binding of some aromatic and cyclic compounds to biopolymers in vitro. J. Invest. Dermatol., 5_3: 344-350, 1969.

Bolt, A. G.: Interactions between human melanoprotein and chlorpromazine derivatives. I. Isolation and purifi­ cation of human melanoprotein from hair and melanoma tissue. Life Sci., 6^: 1277-1283, 1967.

Bolt, A. G. and Fo-rest, I. S.: In vivo and in vitro interaction of chlorpromazine and melanin. Recent Adv. Biol. Psychiat., 10: 115-120, 1965. Borg, D. C.: Free radicals for imipramine. Biochem. Pharmacol., 1£: 115-120, 1965. Borg, D. C.: Electron spin resonance applied to free radicals of physiological, pharmacological or bio­ chemical interest. In, Biological Applications of Electron Spin Resonance. H. M. Swartz, J. R. Bolton and D. C. Borg (Editors), pp. 266-350, Wiley-Inter- science, New York, 1972. Borovansky, J. and Duchon, J.: Chemical composition of hair melanosomes. Dermatologica, 149: 116-120, 1974.

Brothadgen, H.: Vitiligo. In, Research Progress in Organic, Biolgical and Medicinal Chemistry, Vol. 3, Part II. U. Gallo and L. Santamaria (Editors), pp. 640-648, American Elsevier Publishing Co. Inc., New York, 1972. Brown, J. C.: Pigmentation of certain mesencephalic tegmental nuclei in the dog and cat. J. Comp. Neurol., 81: 249-257, 1944. Bu'Lock, J. D.: The formation of melanin from adrenochrome. J. Chem. Soc., 52-58, 1961.

Burge, K. M. and Winkelmann, R. F.: Mercury pigmentation. Arch. Dermatol., 102: 51-61, 1970. 240 Carlsson, A. and Lindqvist, M.: Effect of chlorpromazine or haloperidol on formation of 3-methoxyramine and normetanephrine in mouse brain. Acta Pharmacol. Toxicol., 20_i 140-144, 1963. Cameron, M. E., Lawrence, J. M. and Olrich, J. G.: Thioridazine (Melleril) retinopathy. Brit. J. Ophthal., 56: 131-134, 1972.

Chen, K. K. and Poth, E. J.: Racial differences as illus­ trated by the mydriatic action of cocaine, Euphthal- mine and ephedrine. J. Pharmacol. Exp. Ther., 36: 429-445, 1929. Cleffman, G.: Function-specific changes in the metabolism of agouti pigment cells. Exp. Cell Res., 3^: 590-600, 1964. Christensen, E., Moller, J. E. and Faurbye, A.: Neuro- pathological investigation of 28 brains from patients with dyskinesia. Acta Psychiat. Scand., £ 6 : 14-23, 1970. Cohen, M. M. and Flora, G. C. : Cerebral intoxication. In, Clinical Neurology, Vol. 2, 2nd Edition, A. B. Baker (Editor), pp. 980-1053, Hoeber-Harper, New York, 1962.

Coleman, D. L.: Effect of genic substitution on the incorporation of tyrosine into the melanin mouse skin. Arch. Biochem. Biophys., 96_: 562-568, 1962. Cotzias, G. C.: Levodopa manganese and degenerations of the brain. Harvey Lectures, 6J3: 115-147, 1974. Cotzias, G. C., Papavasiliou, P. S., Van Woert, M. H. and Sakamoto, A.: Melanogenesis and extrapyramidal diseases. Fed. Proc., 2J3: 713-718, 1964. Crosby, E. C., Humphrey, T. and Lauer, E. W.: Correlative Anatomy of the Nervous System, Ch. 5, pp. 260-265, The MacMillan Co., New York, 1962. Cumings, J. N.: Trace metals in the brain in Wilson's disease. J. Clin. Path., 21: 1-7, 1968. Curzon, G.: Metals and melanins in the extrapyramidal centers. Pharmac. Therap. B. , 1_: 673-684, 1975. D'Agostino, A. G. and Luse, S. A.: Electron microscopic observations on the human substantia nigra. Neurology, 14: 529-536, 1964. 241 Dencker, L., Lindquist, N. G. and Ullberg, S.: Mechanism of drug-induced chronic otic lesions. Role of drug accumulation on the melanin of the inner ear. Experientia, 29_: 1362-1363, 1973. Domino, E. F., Hudson, R. D. and Zografi, G.: Substituted phenothiazines: Pharmacology and chemical structure. In, Drugs Affecting the Central Nervous System, Vol. 2, Alfred Burger (Editor), pp. 327-397, Marcel Dekker, Inc., New York, 1968. Dreyfuss, J. , Beer, B., Devine, D. D. , Roberts, B. F. and Schreiber, E. C.: Fluphenazine-induced parkinsonism in the baboon: Pharmacological and metabolic studies. Neuropharmacol. , 11: 223-230, 1972. Drochmans, P.: Ultrastructure of melanin granules. In, Advances in Biology of the Skin, Vol. 8 , The pigment­ ary System, W. Montagna and F. Hu (Editors), pp. 169- 177, Pergamon Press, Oxford, 1967. Duchon, J. and Gregora, U.: Homovanillic acid and its relation to tyrosine metabolism in melanoma. Clin. Chim. Acta, 443-446, 1962. Duchon, J., Borovansky, J. and Hach, P.: Chemical composi­ tion of ten kinds of various melanosomes. In, Pigment Cell, Vol. 1, Mechanisms in Pigmentation, V. Riley (Editor), pp. 165-170, Karger, Basel, 1973. Duffy, P. E. and Tennyson, V. M.: Phase and electron- microscopy observations of Lewy bodies and melanin granules in the substantia nigra and locus coeruleus in Parkinsons' disease. J. Neuropathol. Exp. Neurol., 24: 398-414, 1965. Duke-Elder, S. and Wyber, K.: System of Ophthalmology. Vol. 3, Part II. Congenital Deformities, Henry Kimpton (Editor), pp. 813-819, London, 1958. Ecobichon, D. J. and Comeau, A. M.: Genetic polymorphism of plasma carboxylesterases in the rabbit: Correlation with pharmacological and toxicological effects. Toxicol. Appl. Pharmacol., 27: 28-40, 1970. Ehinger, B. and Falck, B.: Innervation of iridic melano- phores. Z. Zellforsch., 105: 538-542, 1970. 242 Ehinger, B., Falck, B. and Rosengren, A. M.: Adrenergic denervation of the eye by unilateral sympathectomy. Albretch V. Graeffes Arch. exp. Ophthal., 177 : 206- 211, 1969

Emiru, V. P.: Response to mydriatics in the African. Brit. J. Ophthal., !55: 538-543, 1971 Falck, B., Jacobson, S., Olivercrona, H. and Rorsman, H.: Pigmented nevi and malignant melanomas as studied with a specific fluorescent method. Science. 149; 439-440, 1965 Falck, B. and Rorsman, H.: Observations on the adrenergic innervation of the skin. Experientia, 19^: 205-206, 1963 Fenichel, G. M. and Bazelon, M.: Studies on neuromelanin. II. Melanin in the brain stems of infants and children. Neurology, 18: 817-820, 1968 Fitzpatrick, T. B., Seiji, M. and McGugan, A. D.: Melanin pigmentation. New Eng. J. Med., 265; 328-332, 1961

Fitzpatrick, T. B., Quevedo, W. C., Levine, A. L., McGovern, V. J., Mishima, Y. and Oettle, A. G.: Terminology of vertebrate melanin-containing cells: 1965. Science, 152: 88-89, 1966 Foley, J. M. and Baxter, D.: Observations on the morphology of the cells of the locus ceruleus and substantia nigra. J. Neuropath, exp. Neurol., 15^: 219-221, 1956 Foley, J. M. and Baxter, D.: On the nature of pigment granules in cells of the locus ceruleus and substantia nigra. J. Neuropath, exp. Neurol., 17: 586-598, 1958

Forrest, F. M.: Evolutionary origin of extrapyramidal disorders in drug treated mental patients, its significance and the role of neuromelanin. In, Advances in Biochemical Pharmacology, Vol. 9, Phenothiazines, I. S. Forrest, C. J. Carr and E. Usdin (Editors), pp. 255-268, Raven Press, New York, 1974 Forrest, F. M., Forrest, I. S. and Roizin, L.: Clinical, biochemical and postmortem studies on a patient treated with chlorpromazine. Rev. Agressology, £: 259, 1963 (Cited by Van Woert and Ambani, 1974). 243 Frenk, E., Pathak, M. A., Szabo, G. and Fitzpatrick, T. B.: Selective action of mercaptoethylamines on melano­ cytes in mammalian skin* Arch. Dermatol., 9^7 : 465- 477, 1968. Furchgott, R. F.: The pharmacological differentiation of adrenergic receptors. Ann. N.Y. Acad. Sci., 139: 553-570, 1967. Furchgott, R. F.: The classification of adrenoceptors (adrenergic receptors). An evaluation from the standpoint of receptor theory. In, Catecholamines, M. Blashko and E. Muscholl (Editors) p. 283, Springer- Verlag, New York, 1972 • Furchgott, R. F. and Bursztyn, P.: Comparison of dissocia­ tion constants and of relative efficacies of selected agonists acting on parasympathetic receptors. Ann. N. Y. Acad. Sci., 144; 882-899, 1967- Gillilan, L. A.: The nuclear pattern of the non-tectal portions of the midbrain and isthmus in ungulates. J. Comp. Neurol., 78^: 289-364, 1943* Goar, E. L. and Fletcher, M. C.: Toxic chorioretinopathy following the use of NP-207. Amer. J. Ophthal., 44: 603-608, 1957.

Goldman, J. M. and Hadley, M. E . : In vitro demonstration of adrenergic receptors controlling melanophore responses of the lizard, Anolis Carolinensis. J. Pharmacol. Exp. Ther., 166: 1-7, 1969. Goldstein, A.: Biostatistics: An Introductory Text. The MacMillan Co., New York, 1964.

Goodman, L. S. and Gilman, A.: The Pharmacological Basis of Therapeutics, 4th Edition, p. 529. The MacMillan Co., New York, 1970. Greiner, A. C. and Nicolson, G. A.: Pigment deposition in viscera associated with prolonged chlorpromazine therapy. Canad. Med. Assn. J., 9jL: 627-635, 1964. Gupta, I. and Bhide, N. K . : Nature of adrenergic receptors on the skin melanophores of Rana Tigrina. J. Pharm. Pharmacol., 19: 768-769, 1967. 244 Guth, P. S. and Spirtes, M. A.: The phenothiazine tranquil­ lizers: Biochemical and biophysical interactions. Int. Rev. Neurobiol., 1_% 231-278, 1964 .

Hall, K. and Wolfram, L. J.: Isolation and identification of the protein component of hair melanin. J. Soc. Cosmet. Chem. , 26_: 247-254, 1975.

Harley-Mason, J.: Melanins. In, Comprehensive Biochemis­ try, Vol. 6 , M. Florkin and E. H. Stotz (Editors), pp. 254-257, Elsevier Publishing Co., New York, 1965.

Harpur, E. S. and D'Arcy, P.F.D.: The influence of pigment­ ation of rats and guinea-pigs on the ototoxicity of kanamycin and neomycin. Experientia, 31^ 1323-1324, 1975. Harrison, W. H. , Gray, R. M. arid Solomon, L. M.: Incorpora­ tion D-amphetamine into pigmented guinea-pig hair. Brit. J. Dermatol., 91^: 415-418, 1974a. Harrison, W. H., Gray, R. M. and Solomon, L. M. : Incorpora­ tion of L-dopa, L-a-methyldopa and DL-isoproterenol into guinea-pig hair. Acta dermatovener., 5£: 249- 253, 1974b. Havener, W.: Ocular Pharmacology, p. 197, The C. V. Mosby Co., St. Louis, 1966. Hearing, V. J.: Tyrosinase Activity in subcellular fractions of black and albino mice. Nature New Biol., 245: 81-82, 1973 . Hegedus, Z. L. and Altschule, M. D.: The formation of rheomelanins in human blood plasma from catecholamines, from L-dopa and from some of their derivatives. Arch. Int. Physiol. Biochem., 443-459, 1970. Hornykiewicz, 0.: Dopamine in the basal ganglia: its role and therapeutic implications (including the chemical use of L-dopa). Br. Med. Bull., 29_: 172-178, 1973 .

Howard, R. 0., McDonald, C. J., Brendann, D. and Creasey, W. A.: Experimental chlorpromazine cataracts. Invest. Ophthal., 8^ 413-421, 1969. Hunt, P. F., Kannengiesser, M-H. and Raynaud, J-P.: The nature of ^n-imipramine binding to synaptosomes. Biochem. Pharmacol., 24: 681-685, 1975. 245 Ishii, T. and Friede, R. L.: Tissue binding of tritiated norepinephrine in pigmented nuclei of human brain. Amer. J. Anat., 122: 139-144, 1968. Jacobowitz, D.: Application of histochemical techniques in clinical pharmacology. In, Animal and Clinical Pharmacologic Techniques in Drug Evaluation. Vol. 2. P. E. Spiegler and J. H. Moyer (Editors), pp. 291- 296, Year Book Medical Publishers, Inc., Chicago, 1967 • Jacobowitz, D. and Laties, A. M.: Direct adrenergic innervation of a Teleost melanophore. Anat. Rec., 162: 501-504, 1968. Jacquot, C., Rapin, J. R., Wepierre, J. and Cohen, Y.: Metabolisme et distribution tissulaire de l'ephedrine 14C chez le cobaye. Arch. Int. Pharmacodyn., 207: 298-309, 1974 . Janes, R. G. and Stiles, J. F.: The penetration of -Re­ labelled atropine into the eye. A.M.A. Arch. Ophthal., ' 62: 69-74, 1959 . Jervis, G. A.: Phenylpyruvic oligophrenia: introductory study of 50 cases of mental deficiency associated with excretion of phenylpyruvic acid. Arch. Neurol. Psychiat. , 38/ 944-963, 1937 .

Jimbow, K. and Kukita, A.: Fine structure of pigment granules in the human hair bulb: Ultrastructure of pigment granules. In, Biology of Normal and Abnormal Melanocytes. T. Kawamura, T. B. Fitzpatrick and M. Seiji (Editors), pp. 171-193, University Park Press, Baltimore, 1971. Jimbow, K. and Fitzpatrick, T. B.: Changes in distribution pattern of cytoplasmic filaments in human melanocytes during ultraviolet-mediated melanin pigmentation. J. Cell Biol., 65/. 481-488, 1975 .

Junod, A. F.: Accumulation of 14 C-imipramine in isolated perfused rat lungs. J. Pharmacol. Exp. Ther., 183: 182-187, 1972. Kalow, W.: Pharmacogenetics (Heredity and the Response to Drugs), pp. 52-68, W. B. Saunders, Co., Philadelphia, 1962 . 246

Karreman, G., Isenberg, I. and Szent-Gyorgyi, A.: On the mechanism of action of chlorpromazine. Science, 130: 1191-1192, 1959.

Kastin, A. J. and Schally, A. V.: Autoregulation of release of melanocyte stimulating hormone from the rat pituitary. Nature, 213: 1238-1240, 1967 . Legros, J., Rosner, I. and Berger, C.: Ocular effects of chlorpromazine and oxypertine in beagle dogs. Brit. J. Ophthal., 55_: 407-415, 1971. Legros, J., Rosner, I. and Berger, C.: Retinal toxicity of chlorpromazine in the rat. Toxicol. Appl. Pharmacol., 26: 459-465, 1973. Lerner, A. B. Effects of ions on melanin formation. J. Invest. Dermatol., 1£: 47-52, 1952. Lerner, A. B.: Neural control of pigment cells. In, Biology of Normal and Abnormal Melanocytes. T. Kawamura, T. B. Fitzpatrick and M. Seiji (Editors), pp. 3-16, University Park Press, Baltimore, 1971. Lerner, A. B. and Fitzpatrick, T. B.: Biochemistry of melanin formation. Physiol. Rev., 30^: 91-126, 1950. Lerner, A. B. and Fitzpatrick, T. B.: The control of melanogenesis in human pigment cells. In, Pigment Cell Growth, M. Gordon (Editor), pp. 319-333, Academic Press, New York, 1953 •

Leu, M. L.,Strickland, M. C.,Wang, C. C. and Chen, T. S. N.: Skin pigmentation in Wilson's disease. J. Amer. Med. Assn., 211: 1542-1543, 1970- Lindquist, N. G. : Accumulation ixi vitro of 35 S-chlorproma- zine in the neuromelanin of human substantia nigra and locus ceruleus. Arch. Int. Pharmacodyn., 200: 190-195, 1972 .

Lindquist, N. G.: Accumulation of drugs on melanin. Acta Radiol., Suppl. 325: 5-92, 1973.

Lindquist, N. G. and Ullberg, S.: The melanin affinity of chloroquine and chlorpromazine studied by whole body autoradiography. Acta Pharmacol. Toxicol., 31, Suppl. II: 1-32, 1972. 247

Lorincz, A. L.: Pigmentation. In, Physiology and Bio­ chemistry of the Skin, S. Rothman (Editor), pp. 515- 563, University of Chicago Press, Chicago, 1954.

Lowry, 0. H.: Rosebrough, N. J . , Farr, L. and Randall, R. J.: Protein measurement with the Folin phenol reagent. J. Biol. Chem., 193; 265-275, 1951. Maeda, T. and Wegman, R.: Infrared spectrometry of locus ceruleus and substantia nigra pigments in human brain. Brain Res., 14_: 673-681, 1969 * Malawista, S. E.: Reversible inhibition of melanin granule movement by cytochalasin B. Nature, London, 234: 354-355, 1971- Mann, D.M.A. and Yates, P. 0.: Lipoprotein pigments- Their relationship to aging in human nervous system. II. The melanin content of pigmented nerve cells. Brain, 97: 489-498, 1974. Marsden, C. D.: Pigmentation in the nucleus substantia nigrae of mammals. J. Anat. (London)., 9j>: 256-261, 1961 . Marsden, C. D.: Brain melanin. In, Pigments in Pathology, M. Wolman (Editor), pp. 395-420, Academic Press, New York, 1969• Mason, H. S.: The structure of melanins. In, Pigment Cell Growth, M. Gordon (Editor), pp. 277-303, Academic Press, New York, 1953 • Mason, H. S., Ingram, D.J.E. and Allen, B.: The free radical property of melanins. Arch. Biochem. Biophys., 86: 225-230, I960 . Meier-Ruge, W., Kalberer, F. und Cerletti, A.: Microhisto- autoradiographische untersunchungen iiber die vertei- lung von tritiummarkieten phenothiazinen im auge. Experientia, 22^: 153-155, 1966 . Meier-Ruge, W. and Cerletti, A.: The significance of the melanin-bearing choroid in the retina. In, Biochem­ istry of the Eye, Symposium, Tutzing Castle, pp. 521- 525, Karger, Basel, 1968. Mena, I., Marin, 0. and Fuenzalida, S.: Chronic manganese poisoning. Clinical picture and manganese turnover. Neurology, 17: 128-136, 1967. 248

Mintz, B.: Gene control of mammalian pigmentary differen­ tiation. I. Clonal origin of melanocytes. Proc. Nat. Acad. Sci., 58^: 344-351, 1967 .

Misuraca, G., Nicolaus, R. A., Prota, G. and Chiara, G.: A cytochemical study of phaemelanin formation in feather paillae of New Hampshire chick embryos. Experientia, 215: 920-922, 1969 • Molokhia, M. M. and Portnoy, B. : Trace elements and skin pigmentation. Brit. J. Dermatol., 89^: 207-209 , 1973 .

Moses, H. L., Ganote, C. E., Beaver, D. L. and Schuffman, S. S.: Light and electron microscopic studies of pigment in human and Rhesus monkey substantia nigra and locus ceruleus. Anat. Rec. 155; 167-183, 1966. Mottaz, J. H. and Zelickson, A. S.: Ultrastructure of hair pigment. In, Advances of Biology of the Skin, Vol. 9, Hair Growth, W. Montagna and R. L. Dobson (Editors) p. 471. Pergamon Press, Oxford, 1969. McGinnes, J. and Proctor, P.: The importance of the fact that melanin is black. J. Theor. Biol., 39_: 677-678, 1973 .

McGuire, J.: Adrenergic control of melanocytes. Arch. Dermatol., 101: 173-180, 1970. McGuire, J. and Moellman, G.: Cytochalasin B: effects on microfilaments and movement of melanin granules within melanocytes. Science, 175: 642-644, 1972 . Nagatsu, T.: Biochemistry of Catecholamines, The Biochem­ ical Method, p. 47, University Park Press, Baltimore, 1973 • Nagatsu, I., Sudo, Y. and Nagatsu, T.: Tyrosine hydroxyla- tion in the banana plant. Enzymologia, £3: 25-31, 1972 . Nicolaus, R. A.: Biogenesis of melanins. Rass. Med. Sper., 9_, Suppl. 1, 1962 (Cited by Blois, M. S., 1971)- Nordgren, L., Rorsman, H., Rosengren, A. M. and Rosengren, E.: Dopa and dopamine in the pigment of substantia nigra. Experientia, 27: 1178-1179, 1971. 249

Novales, R. R. and Davis, W. J.: Melanin dispersing effect of adenosine 3',5'-monophosphate in amphibian melano- phores. Endocrinology, 81_: 283-290, 1967 .

Obianwu, M. 0. and Rand, M. J.: The relationship between the mydriatic action of ephedrine and the colour of the iris. Brit. J. Ophthal., 4j?: 264-270, 1975. Pakkenberg, H. and Brody, H.: The number of nerve cells in the substantia nigra in paralysis agitans. Acta Neuropathol. , f5: 320-324, 1965 . Patil, P. N.: Cocaine-binding by the pigmented and non­ pigmented iris and its relevance to the mydriatic effect. Invest. Ophthal., ILL: 739-746, 1972. Patil, P. N. and Jacobowitz, D.: Unequal accumulation of adrenergic drugs by pigmented and nonpigmented iris. Amer. J. Ophthal., 7j8: 470-477, 1974. Patil, P. N., Shimada, K., Feller, D. R. and Malspeis, L.: Accumulation of (-)-l^c-ephedrine by the pig­ mented and nonpigmented iris. J. Pharmacol. Exp. Ther., 188: 342-352, 1974- Perry, T. L., Culling, C. F. A., Berry, K. and Hansen, S.: 7-hydroxychlorpromazine: potential drug metabolite in psychiatric patients. Science, 146: 81-82, 1964-

Poirier, L. J. and Sourkes, T. L.: Influence of the substantia nigra on the catecholamine content of the striatum. Brain, 88^ 181-192, 1965 •

Potts, A. M.: The concentration of phenothiazines in the eye of experimental animals. Invest. Ophthal., 1: 520-530, 1964 • Potts, A. M. : The reaction of uveal pigment iri vitro with polycyclic compounds. Invest. Ophthal., _3:4°f5-416, 1964 • Potts, A. M. and Au, P. C.: Thallous ion and the eye. Invest. Ophthal., _10: 925-931, 1971 . Pryor, W. A.: Free radicals in biological systems. Scient. Amer. 223: 70-83, 1970 . 250

Quevedo, W. C.: Genetic regulation of pigmentation in mammals. In, Biology of Normal and Abnormal Melano­ cytes, T. Kawamura, T. B., Fitzpatrick and M. Seiji (Editors), pp. 99-115, University Park Press, Baltimore, 1971. Rawles, M. E.: Origin of pigment cells from the neural crest in the mouse embryo. Physiol. Zool., 20: 248-266, 1947. Riley, P. A.: The biochemistry of pigment formation. In, The Physiology and Pathophysiology of the Skin, Vol. 3, A. Jarret (Editor), pp. 1149-1165, Academic Press, New York, 1974. Rodgers, A. D. and Curzon, G.: Melanin formation by human brain in vitro. J. Neurochem. , 2_4: 1123-1129, 1975 *

Roth, J. A. and Gillis, C. N.: Deamination of B-phenyl- ethylamine by monoamineoxidase - inhibition by imipramine. Biochem. Pharmacol., 23_: 2537-2545, 1974 -

Roy, S. and Wolman, L.: Ultrastructural observations in Parkinsonism. J. Pathol., 99^: 39-44, 1969 * Satanove, A.: Pigmentation due to phenothiazines in high and prolonged dosage. J. Amer. Med. Assn., 191: 263-268, 1965. Scott, A. W . : Retinal pigmentation in a patient receiving thioridazine. Arch. Ophthal., 70.: 775-778, 1963. Seidehamel, R. J., Tye, A. and Patil, P. N.: An analysis of ephedrine in relationship to iris pigmentation in the guinea-pig eye in vitro. J. Pharmacol. Exp. Ther., 171: 205-213, 1970. Seiji, M.: Melanogenesis. In, Ultrastructure of Normal and Abnormal Skin, A. Zelickson (Editor), pp. 183- 201, Lea and Febiger, Philadelphia, 1967. Seiji, M . , Fitzpatrick, T. B., Simpson, R. T. and Birbeck, M.S.C.: Chemical composition and terminology of specialized organelles (melanosomes and melanin granules) in mammalian melanocytes. Nature, London, 197: 1082-1084, 1963. 251 Seiji, M . , Yoshida, T . , Itakura, H. and Irimajiri, Y.: Inhibition of melanin formation by sulfhydryl compounds. J. Invest. Dermatol., 52_i 280-286, 1969. Sekeris, C. E. and Karlson, P.: Biosynthesis of catechola­ mines in insects. Pharmacol. Rev., 1£: 89-94, 1966. Sharpies, D.: Competition for plasma protein binding sites between phenothiazine tranquillizers and imino- dibenzyl antidepressants. J. Pharm. Pharmacol., 27: 279-281, 1975.

Shimada, K., Sokoloski, T., Baweja, R. and Patil, P. N.: Binding characteristics of drugs with L-dopa melanin. Fed. Proc., 32: 512, 1973.

Shimada, K., Sokoloski, T., Baweja, R. and Patil, P. N.: Binding characteristics of drugs to the synthetic L-dopa melanin. J. Pharm. Pharmacol., 1976, in press. Sidall, J. R.: Ocular toxic changes associated with chlorpromazine and thioridazine. Canad. J. Ophthal., 1: 190-198, 1966. Sidall, J. R.: Phenothiazines in ophthalmology. Int. Ophthal. Clin., 7: 207-215, 1967. Snyder, S. H., Banerjee, S. P., Yamamura, H. I. and Greenberg, D.: Drugs, neurotransmitters and schizo­ phrenia. Science, 184: 1243-1253, 1974. Stein, W. D.: Chemical composition of the melanin granule and its relation to the mitochondria. Nature, London, 175: 256-257, 1955. Stewart, T. W., Miles, D. W. and Earnshaw, E. R.: Pigmenta­ tion and retinopathy due to chloroquine. Acta Derm-veneorol., £8: 47-52, 1968.

Stupp, H., Kupper, K., Lagler, F., Sous, H. and Quanta, M.: Inner ear concentrations and ototoxicity of different antibiotics in local and systemic application. Audiology, 12: 350-363, 1973.

Swan, G. A . : Chemical structure of melanins. Ann. N.Y. Acad. Sci., 100: (Part II): 1005-1019, 1963. 252 Swan, G. A.: Current knowledge of melanin structure. In, Pigment Cell, Vol. 1, Mechanisms in Pigmentation, V. Riley (Editor), pp. 151-157, Karger, Basel, 1973.

Swan, G. A. and Wagott, A.: Studies related to the chemistry of melanins. X. Quantitative assessment of different types of units present in dopa-melanin. J. Chem. Soc. C,: 1409-1418, 1970. Swartz, H. M. : Cells and tissues. In, Biological Applica­ tions of Electron Spin Resonance. H. M. Swartz, J. R. Bolton and D. C. Borg (Editors), pp. 155-195, Wiley-Interscience, New York, 19 72. Takahashi, H. and Fitzpatrick, T. B.-: Large amounts of dihydroxyphenylalanine in hydrolisate of melanosomes from Harding-Passey mouse melanoma. Nature, London, 209: 888-890, 1966. Taylor, S.: Quantitative studies on acetylcholine receptors of the isolated iris sphincter muscle. J. Physiol. (London), 239: 6-7P, 1974 * Thathachari, Y. T.: Structure of melanins. In, The Pigment Cell, Vol. 1, Mechanisms in Pigmentation, V. Riley (Editor), pp. 158-164, Karger, Basel, 1973* Thomson, R. H.: The Pigments of reddish hairs and feathers. Angewandte Chemie., 13^: 305-312, 1974 * Toda, K. and Fitzpatrick, T. B.: The origin of melanosomes. In, Biology of Normal and Abnormal Melanocytes, T. Kawamura, T. B. Fitzpatrick and M. Seiji (Editors), pp. 265-278, University Park Press, Baltimore, 1971. Tousimis, A. J. and Barron, C. N.: Chlorpromazine and the eye of the dog. An electron microscopic study. Exp. Mol. Pathol., 1J3: 89-110, 1970. Underwood, E. J.: Trace Elements in Human and Animal Nutrition, 3d Edition, p. 80, Academic Press, New York, 1971. Van der Wende, C. and Sporlein, M. T.: Oxidation of dopa­ mine to melanin by an enzyme of rat brain. Life Sci., 6 : 386-392, 1963 . 253

Van Woert, M. H.: Reduced nicotinamide-adenine dinucleo­ tide oxidation by melanin: Inhibition by phenothiaz­ ines. Proc. Soc. Exp. Biol. Med., 129: 165-171, 1968.

Van Woert, M. H.: Effect of phenothiazines on melanoma tyrosinase activity, J. Pharmacol. Exp. Ther., 173: 256-264, 1970. Van Woert, M. H. and Ambani, L. M.: Biochemistry of neuromelanin. In, Advances in Neurology, Vol. 5, pp. 215-223, Raven Press, New York, 1974. Van Woert, M. H., Prasad, K. N. and Borg, D. C.: Spectro­ scopic studies of substantia nigra pigment in human subjects. J. Neurochem., 14_: 707-716, 1967.

Van Zutphen, L. F. M.: Qualitative and quantitative detection of atropinesterase and cocainesterase in two breeds of rabbits. Enzymologia, 42: 201-217, 1972. Walls, G. L.: The vertebrate eye and its adaptive radiation, p. 15, Hafner, 1967. Wasserman, H. P.: The circulation of melanin. Its clinical and physiological significance. S. Afr. Med. J., 39: 711-716, 1965. Wepierre, J., Cohen, Y. and Rapin, J.: Accumulation de l'ephedrine, de la norephedrine, de 1'amphetamine et de la tyramine marquees au carbone 14 dans les yeux pigmentes et non pigmentes de deux races de rats. Arch Int. Pharmacodyn., 218: 342-350, 1975. Werner, G.: Fermentdefekte als ursache unterschiedlicher mydriatischer wirksamkeit von atropin und cocain bei kaninchen. Naunyn-Schmiedeberg's Arch. exp. Path. Pharmak., 251: 320-334, 1965.

Whittaker, V. P., Michaelson, I. A. and Kirkland, R. J. A.: The separation of synaptic vesicles from nerve-ending particle (synaptosomes). Biochem J., £0: 293-303, 1964.

Wiskwo, M. A. and Szdbo, G.: Studies on the interaction between melanocytes and keratinocytes with special reference to the role of microfilaments. In, The Pigment Cell, Vol. 1, Mechanisms in Pigmentation, V. Riley (Editor), pp. 27-38, Karger, Basel, 1973. 254 Witkop, C. J.: Albinism. In, Advances in Human Genetics, Vol. 2, H. Harris and K. Hirschhorn (Editors), pp. 61- 142, Plenum Press, New York, 1971. Zelickson, A. S.: Skin pigmentation and chlorpromazine. J. Amer. Med. Assn., 194: 670-672, 1965.

Zelickson, A.: Melanocyte, Melanin granule and Langerhans cell. In, Ultrastructure of Normal and Abnormal Skin, A. Zelickson (Editor), pp. 163-182, Lea and Febiger, Philadelphia, 1967 .

Zidek, Z. and Janku, I.: Interaction of albinism with sensitivity of mice to lethal effects of repeatedly administered chlorpromazine and isoniazide. Pharmacol. Res. Comm., _3: 127-132, 1971 .