Textbook of Organic Medicinal and Pharmaceutical Chemistry

Because the chapters include a blend of chemical and pharmacological principles necessary for understanding structure—activity relationships and molecular mechanisms of drug action, the

Wilson and Gisvold's Textbook of

ANIC MEDICINAL AND PHARMAC

CHEMIS TRY

ICAL

Professor of Medicinal Chemistry

Department of Pharmaceutical Sciences College of Pharmacy

Oregon State University Corvallis Oregon John M Beale, Jr., Ph.D.

Associate Professor of Medicinal Chemistry and Director of Pharmaceutical Sciences

St Louis College of Pharmacy

Managing Lthsar: MatthewJ Hauber

MarkrsingMa,wger:Samantha S Smith

Production&Iitor:Bill

Designer:DougSmock

Compositor: Maryland Composition

Printer: QuehecorWOrld

Copyright 2(8)4 Lippincott Williams & Wilkins

351 Wcct Camden Street

Baltimore, MD 21201

530 Walnut Street

Philadelphia PA 19106

All rights reserved This book is protected by copyright No part of this book may be reproduced

in any lOon or by any means, including photocopying or by any infortnation storage and

retrieval system without written permission rrom the copyright owner

The publisher is not responsible (as amatterof product liability, negligence, or otherwise) Forany injury resulting from any material contained herein This publication contains infonttation

relating to general principles ui medical care that should not he construed a.s specific instructions

for individual patiettts Manuthcturers' prnduct information and package inserts should he viewed lOr current information, including contraindicationt dosages, and precautions

re-Prinle'rI inthe Uniteg! Stale.sof Anwrieu

Second Edition 1954 Sixth Edition 1971 Ninth Edition, 1991

Third Edition 1956 Seventh Edition, 1977 Tenth Edition, 1998

rswrtli Edition, (962

Llbrnry or Congrnas Cataloglng.In.Publkatloit Data

Wilson and Gisvold's textbook of organic medicinal and phartnaccutical chemistry.— 11th

ed / edited by John H Block John M Beale Jr

p cm,

Includes bibliographical references attd index.

ISBN 11-7817-34111-9

I Pharmaceutical chemistry 2 Chemistry Organic I Title: Textbook of organic medicinal

and pharmaceutical chemistry II Wilson Charles Owens 1911—2002 10 Gisvold Ole.

l904- IV Block John H V Ileak John Marlowe

IDNLM: I Chemistry Pharmaceutical 2 Chemistry Organic QV 744 W754 2ll(9JRS403 143 2111)4

6 IS' 19—dc2l

20031)48849

The puhlisher.s have tizade every effort to trace the c'opyri gut itolders for

borrowed material If they have inadvertently overlooked any they will bepleased to make the necessary arrangements a: the first opportunity

To purchase additional copies of this book, call our customer servicedepartmenl at (800) 638-3030 or fax orders to (301) 824-7390 lniernutionalcustomers should call (301) 714-2324

Visit Lippincols Williams & Wilkins on the Internet: htIp://

www.LWW.com LippincottWilliams & Wilkins customer service

representatives are available from 8:30 am to 6:00 pm, EST

05 06 (>7

2 3 4 5 6 7 8 9 10

l'he Fkrenth Edüion of Wilson and Gisvold's Texibook of Organic and Medicinal Pharmaceutical

Chem i stry is' (kYiica:ed Iv the q( Jaiine N !)elgado Charles 0 Wilson

A graduate of the University of Texas at Austin and the University of Minnesota Jaime Delgado

began his teaching career as an assistant professor at the University of Texas College of Pharmacy in

1959.He rose through the academic ranks to become professor andhead of the Division of Medicinal

Chemistry anda leader in research andgraduate education He essentially built both the graduateprogram and the Division from scratch,and his publication of research and scholarly works brought

national recognition to thedepartment.

AlthoughJaime Delgadobecameknown for his research and scholarship his first love and hisgreatest

legacy were in teaching and advising undergraduate and graduate students The University ofTexas at

Austin awarded him five major teaching awards and recognized him two times as one of its "best"professors In 1997 he was elected to the Academy of Distinguished Teachers at the university and

was honored as a Distinguished Teaching Professor, a permanent academic title Former dean JamesDoluisio described Dr Delgado's teaching style as "owning the classroom" because of his knowledge.communicationskills, and deep convictionthat pharmacy is a science-based profession His enthusiasm

andextemporaneoususe of the chalkboard were legendary In addition to his contributions to teaching

at the University of Texas, Dr Delgado traveled extensively in Mexico and South Americatopresentlectures on pharmaceutical education

Jaime Delgado's first contributions 10 the Textbook of Organic Medicinal (111(1 Phannaceutical istry were made as a chapter author in the seventh and eighth editions Much of the material he presented

Chem-came from his lecture notes Although he was proud of these contributions, which were expanded in

the ninth and tenth editions, he considered his role as coeditor in the latter editions one of the highlights

of his distinguished career Jaime was a true gentleman and a pleasure to have as a collaborator He

will he greatly missed by the editors, authors4 and professional staff for the Textbook

William A Reiners

Charles 0 Wilson

1911—2002

Asthe chapters for the eleventh edition were being sent to the publisher I was notified that mycolleague and friend Charles Wilson had died shortly Christmas I-Ic was a product of thePacific Northwest having received all of his degrees from the University of Washington His first

teaching job was at the now discontinued pharmacy school at George Washington Universityand then

he moved to the University of Minnesota Charles along with other medicinal chemistry faculty at theUniversity of Minnesota saw the need for textbooks that presented modern medicinal chemistry In

1949 he and Professor Ole Gisvold edited Organic chemistry in Pharmacy, which became the first

edition of the Textbook of Medicinal and Pharmaceutical che,nix:rv Continuing in this tradition Charles

and Professor Tailo Some assumed the authorship of Roger'.c Inorganic Pharmaceutical Chemistry,which included eight editions before its discontinuance Finally Charles and Professor Tony Jonesstarted the American Drug Index series Charles continued his publishing activities after moving to theUniversity of Texas and then assumed the position of Dean of Oregon State University's School olPharmacy, where he oversaw a major expansion of its faculty and physical plant

Although a medicinal chemist Charles devoted considerable time to his chosen pharmacy profession.students, and communily Charles was an active member of the American Pharmaceutical Association

as well as the pharmacy associations in each state where he lived In addition, he was a registeredpharmacist in each state where he taught: Washington Minnesota, Texas Oregon, and the District or

Columbia Charles chaired national committees and sections of the American Pharmaceutical

Associa-tion and the American AssociaAssocia-tion of Colleges of Pharmacy Related to these, his loyalty to students

included organizing student branches of the American Pharmaceutical Association al George

Washing-ton University the University of Minnesota and the University of Texas He was actively involved inthe local American Red Cross blood program and took the lead in developing the hugely successful

student centered blood drives at Oregon State University In 1960, Charles and his wife, Vaughn helped

launch the AFS (American Field Service) in Corvallis, an international high-school exchange program

He volunteered for Meals on Wheels for over 30 years after his retirement

We certainly miss this fine gentleman and leader of pharmacy education and the pharmacy profession

John H Block

For almost six decades, Wilson and Gisvo!d s Textbook of Organic Medicinal and Pharmaceuticalchemistry hasbeen a standard in the literature of medicinal chemistry Generations of students andfaculty have depended on this textbook not only for undergraduate courses in medicinal chemistry but

also as a supplement for graduate studies Moreover, students in other health sciences have found certainchapters useful at one time or another The current editors and authors worked on the eleventh editionwith the objective of continuing the tradition of a modem textbook for undergraduate studerns and alsofor graduate students who need a general review of medicinal chemistry Because the chapters include

a blend of chemical and pharmacological principles necessary for understanding structure—activity

relationships and molecular mechanisms of drug action, the book should be useful in supporting courses

in medicinal chemistry and in complementing pharmacology courses

II is our goal that the eleventh edition follow in the footsteps of the tenth edition and reflect the

dynamic changes occurring in medicinal chemistry Recognizing that the search for new drugs involvesboth synthesis and screening of large numbers of compounds, there is a new chapter on combinatorialchemistry that includes a discussion on how the process is automated The power of mainframe comput-ing now is on the medicinal chemist's desk A new chapter describes techniques of molecular modelingand computational chemistry With a significant percentage of the general population purchasing altema-tivc medicines, there is a new chapter on herbal medicines that describes the chemical content of many

erythro-on biotechnology describes these exciting applicatierythro-ons Recent advances in understanding the immune

system at the molecular level have led to new agents that suppress or modify the immune response,

producing new treatments for autoimmune diseases including rheumatoid arthritis, Crohn's disease, andmultiple sclerosis Techniques of genetic engineering now allow the preparation of pure surface antigens

as vaccines while totally eliminating the pathogenic organisms from which they are derived.The editors welcome the new contributors to the eleventh edition: Doug Henry Phillip Bowen,Stephen i Cutler 1 Kent Walsh, Philip Proteau and Michael J Deimling The editors extend thanks

to all of the authors who have cooperated in the preparation of the current edition Collectively, the

authors represent many years of teaching and research experience in medicinal chemistry Their chaptersinclude summaries of current research trends that lead the reader to the original literature Documentation

and references continue to be an important feature of the book

We continuc to be indebted to Professors Charles 0 Wilson and Ole Gisvold the originators ofthe book and editors of five editions Professor Robert Doerge who joined Professors Wilson and

Gisvold for the sixth and seventh editions and single-hundedly edited the eighth edition, and Professors

Jaime Dclgado and William Remers who edited the ninth and tenth editions They and the authors

have contributed significantly to the education of countless pharmacists, medicinal chemists, and otherpharmaceutical scientists

John H BlockJohn M Beale Jr

VI,

Associate Professorof Medicinal

Chemistry and Director of

Professorof Chemistry and

Director Center for Biomolecular

Structure and Dynamics

Computational Chemistry Building

Professorof Medicinal Chemistry

Department of Pharmacal Sciences

SeniorResearch Professor

Director of the Nutuml Products

MICHAEL J DEIMLING, R.PH., PH.D.

Professorof Pharmacology and Chair

Department of PharmaceuticalSciences

School of PharmacySouthwestern Oklahoma State

University

Weatherford, Oklahoma

JACK DERUITER, PH.D.

Professorof Medicinal Chemistry

Department of Pharmacal SciencesSchool of Pharmacy

Auburn UniversityAuburn Alabama

JACK N HALL, M.S., R.PH., BCNP

Clinical Lecturer

Department of Radiology/NuclearMedicine

College of Medicine University ofArizona

University of Arizona Health

Sciences Center

Tucson Arizona

DOUGLAS R HENRY

AdvisoryScientist

MDL Information Systems Inc

San Leandro, California

THOMAS J HOLMES, JR.,

PH.D.

Associate ProfessorSchool of Pharmacy

Campbell UniversityBuies Creek, North CarolinaTIM B HUNTER, M.D.

Vice-Chairmanand Professor

Department of RadiologyUniversity of ArizonaTucson Arizona

EUGENE I ISAACSON,

PH.D.

ProfessorEmeritus of Medicinal

ChemistryDepartment of PharmaceuticalSciences

DANIEL A KOECHEL,

PH.D.

ProfessorEmeritus—PharmacologyDepartment of Pharmacology

Medical College of OhioToledo Ohio

GUSTAVO R ORTEGA, R.PH., PH.D.

Professorof Medicinal Chemistry

Department of PharmaceuticalSciences

School of PharmacySouthwestern Oklahoma State

Oregon State UniversityCorvallis Oregon

ix

THOMAS N RILEY, PH.D.

Professorof Medicinal Chemistry

Department of Pharmacal Sciences

School of Pharmacy

Auburn University

Auburn Alabama

GARETH THOMAS, PH.D.

Associate Senior I.ecturer

TheSchool of Pharmacy and

System

Tucson, Arizona

Supports and Linkers

Solution-Phase Combinatorial Chemistry

Pooling Strategies

Detection, Purification, and Analysis

Encoding Combinatorial Libraries

High-Throughput Screening (HIS)

Virtual (in Silico) Screening

Chemical Diversity and Library Design

Report Card on Combinatorial Chemistry: Has It

Worked'

Resources for Combinatorial Chemistry

Combinatorial Chemistry Terminology

vu Role of Cytochrome P-450 Monooxygenases in

Oxidative BiotransformationsOxidative Reactions

Reductive Reactions

Hydrolytic ReactionsPhase II or Conjugation ReactionsFactors Affecting Drug Metabolism

CHAPTER 5

Prodrugs and Drug Latentiation 142

/-'orrest T.Smithand C Randall C/ask

Biotechnology and New Drug Development . 160

The Biotechnology of Recombinant DNA IrDNA) . 162

43 Some Types of Cloning 166

Expression of Cloned DNA 167

Manipulation of DNA Sequence Information . 168

New Biological Targets for Drug Development . 169

43 Novel Drug-Screening Strategies 170

46 Processing of the Recombinant Protein 172

48 Pharmaceutics of Recombinant DNA

111

126

3 3

9

17

2627

41

CHAPTER 4

Metabolic Changes of Drugs and Related

Organic compounds

Ste-ph:,: J C':i:ler and Jo!::: H Block

General Pathways of Drug Metabolism

Sites of Drug Biotransformation

6565

66

Xii Contents

CHAPTER 8

Anti-infective Agents 217

John M Beak Jr

Evaluation of the Effectiveness of a Sterilant . 219

Alcohols and Related Compounds 219

Phenols and Their Derivatives 221

Signal Transduction Inhibitors 438Immunotherapy 440Monoclonal Antibodies . 442

Radiotherapeutic Agents . 444Cytoprotective Agents 445Future Antineoplastic Agents 446Potential Future Developments 448CHAPTER 13

Agents for Diagnostic Imaging 454

Tin, Ii Hunter, T Kent Walsh, Jack N HallIntroduction to Radiation 454Characteristics of Decay 456Biological Effects of Radiation 457Radionuclides and Radiopharmaceuticals for

Organ Imaging 458

283 Radionuclide Production 461

285 Technetium RadiochemistryFluorine Radiochemistry 463

468Gallium Radiochemistry 468Iodine Radiochemistry 468Indium Radiochemistry 469Thallium Radiochemistry 472

Rot/tier L JohnsonAdrenergic Neurotransmitters 524

CHAPTER 17

Cholinergic Drugs and Related Agents 548

GeorgeII Combsand StephenJ Cutler

CholinergicReceptors 548

Cholinergic Neurochemistry 553

Cholinergic Agonists 553

Cholinergic Receptor Antagonists 558

Cholinergic Blocking Agents 572

Parasympathetic Postganglionic Blocking Agents. 573

Solanaceous Alkaloids and Analogues 574

Synthetic Cholinergic Blocking Agents 579

Ganglionic Blocking Agents 586

Neuromuscular Blocking Agents 589

CHAPTER 18

l.)anielit.At,i'chel

Anatomy and Physiology of the Nephron 596

Introduction to the Diuretics 601

Site 1 Diuretics: Carbonic Anhydrase Inhibitors . 603

Site 3 Diuretics: Thiazide and Thiazide-Like

Site 2 Diuretics High-Ceiling or Loop Diuretics . 610

Site 4 Diuretics: Potassium-Sparing Diuretics . 616

Emerging Developments in the Use of Diuretics

to Treat Hypertension and Congestive Heart

Stephen J ('iufrrandGeorge H Cocola.c

Antianginal Agents and Vasodilators 622

Rate of Onset and Duration of Anesthesia . 688

CHAPTER 21

Histamine and Antihistaminic Agents 696

Tliouius N Rilm.'vand JackI)eRuiter

Inhibition of Histamine Release Mast Cell

Stabilizers

Recent Antihistamine Developments: The Acting" Antihistamines

"Dual-Histamine H2 AntagonistsHistamine H3-Receptor LigandsCHAPTER 22

Chemical and Physical Properties of Steroids . 770

Changes to Modify Pharmacokinetic Properties of

CHAPTER 25 Proteins, Enzymes, and Peptide

StephenJ Cutler and Horace G Cutler

ProteinHydrolysates 830Amino Acid Solutions 830Proteins and Protein-Like Compounds 831

718

727

818818

822822823823

825

827

828828829

xiv Coiue,izs

and Commercial Production of Proteins and

Peptides as Pharmaceutical Products 858

Biotechnology-Derived Pharmaceutical Products . 860

C HAPTER 26

Vitamins and Related Compounds 866

Guslai,, R Oriega Michael J Dei,nling and Jaime N

Force Field Methods 923

Geometry Optimization 929Conformational Searching 930

Molecular Dynamics Simulations 933

CHAPTER 1

JOHN H BLOCK AND JOHN M BEALE, JR

The discipline of medicinal chemistry is devoted to the

dis-covery and development of new agents for treating diseases

MOSt ol thisactivity isdirected to new natural or synthetic

organic compounds Inorganic compounds continue to be

important in therapy e.g trace elements in nutritional

ther-apy antacids, and radiopharmaceuticals but organic

mole-with increasingly specific pharmacological activities

are clearly dominant Development of organic compounds

has grown beyond traditional synthetic methods It flow

in-eludes the exciting new held of biotechnology using the

cell' biochemistry to synthesii.e new compounds

Tech-niquesranging l'rom recombinant DNA and site-directed

mutugenesis to fusion of cell lines have greatly broadened

the possibilities for new entities that treat disease The

phar-macist now dispenses modified human insulins that provide

more convenient dosing schedules, cell-stimulating factors

that have changed the dosing regimens for chemotherapy

humaniicd monoclonal antibodies that target specific

tis-sues, and lused receptors that intercept immune

cell—gener-ated cytokines

This hook treats many aspects of organic niedicinals: how

they are discovered, how they act, and how they developed

into clinical agents The process of establishing a new

phar-maceutical is exceedingly complex and involves the talents

ut people from a variety of disciplines including chemistry

hiochetnistry molecular biology, physiology,

pharmacol-ogy pharmaceutics, and medicine Medicinal chemistry,

it-scif is concerned mainly with the organic, analytical, and

biochemical aspects of this process, hut the chemist must

interact productively with those in other disciplines Thus

medicinal chemistry occupies a strategic position at the

inter-face of chemistry and biology

To provide an understanding of the principles of medicinal

chemistry, it is necessary to consider the physicochemical

properties used to develop new pharmacologically active

compounds and their mechanisms of action, the drug's

jabolisni including possible biological activities of the

me-taholites the importance of stereochemistry in drug design,

and the methods used to determine what "space' a drug

occupies All of the principles discussed in this book are

based on fundamental organic chemistry physical

chemis-try' and biochemistry

The earliest drug discoveries were made by random

sam-pling of higher plants Some of this samsam-pling, although based

on anecdotal evidence, led to the use of such crude plant

drugs as opium belladonna, and ephedrine that have been

important for centuries With the accidental discovery of

penicillin came the screening of microorganisms and the

large number of antibiotics from bacterial and fungal

sources Many of these antibiotics provided the prototypical

structure that the medicinal chemist modified to obtain

anti-bacterial drugs with better therapeutic profiles With thechanges in federal legislation reducing the efficacy require-ment for "nutriceutical," the public increasingly is usingso-called nontraditional or alternative medicinals that are

sold over the counter, many outside of traditional pharmacy

distribution channels It is important for the pharmacist andthe public to understand the rigor that is required for pre-scription-only and FDA-approved nonprescription products

to be approved relative to the nontraditional products It also

is important for all people in the health care field and thepublic to realize that whether these nontraditional productsare effective as claimed or not, many of the alternate medi-cines contain pharmacologically active agents that can po-tentiate or interfere with physician-prescribed therapy

Hundreds of thousands of new organic chemicals arc pared annually throughout the world, and many of them areentered into pharmacological screens to determine whether

pre-they have useful biological activity This process of random

screening has been considered inefficient, but it has resulted

in the identification of new lead compounds whose structureshave been optimized to produce clinical agents Sometimes

a lead develops by careful observation of the cal behavior of an existing drug The discovery thaL amanta-dine protects and treats curly influenza A came from a gen-

pharmacologi-eral screen for antiviral agents The use of amantadine inlong-term care facilities showed that it also could he used

to treat parkinsonian disorders More recently automated

high-throughput screening systems utilizing cell culture tems with linked enzyme assays and receptor molecules de-rived from gene cloning have greatly increased the efficiency

sys-of random screening It is now practical to screen enormouslibraries of peptides and nucleic acids obtained from combi-natorial chemistry procedures

Rational design, the opposite approach to high-volumescreening, is also flourishing Significant advances in x-ray

crystallography and nuclear magnetic resonance have made

it possible to obtain detailed representations of enzymes andother drug receptors The techniques of molecular graphicsand computational chemistry have provided novel chemical

structures that have led to new drugs with potent medicinalactivities Development of HIV protease inhibitors and an-giotensin-convcrting enzyme (ACE) inhibitors came from

an understanding of the geometry and chemical character

of the respective enzyme's active site Even if the receptorstructure is not known in detail, rational approaches based

on the physicochemical properties of lead compounds can

provide new drugs For example, the development of

cimeti-dine as an antinuclear drug involved a careful study of thechanges in antagonism of H2-histamine receptors induced

by varying the physical properties of structures based on

1Introduction

2 IViIu,,, and Gi.o'ohlx Textbook of Orga,:ic Medicinal and PharmaceuticalChen,i.strv

histamine Statistical methods based on the correlation of

physicochcmical properties with biological potency are used

to explain and optimize biological activity

As you proceed through the chapters, think of what

prob-1cm the medicinal chemist is trying to solve Why were

cer-tain structures selected? What modilications were made to

produce more focused activity or reduce adverse

reactioo-or produce better pharmaceutical propenics? Was the typical molecule discovered from random screcns, or did themedicinal chemist have a structural concept of the

proto-or an understanding of the disease process that must be rupted?

inter-CHAPTER 2

JOHN H BLOCK

synthesize a new structure and see what happens—contin—

ucs to evolve rapidly as an approach to solving a drug design

problem The combination of increasing power and

decreas-ing cost of desktop computdecreas-ing has had a major impact on

solving drug design problems While drug design

increas-ingly is bawd on modern computational chemical

tech-niques. it also uses sophisticated knowledge of disease

mechanisms and receptor properties A good understanding

(if how the drug is transported into the body, distributed

throughout the body compartments, metabolically altered by

the liver and other organs and excreted from the patient

is required along with the structural characteristics of the

receptor Acid—base chemistry is used to aid in formulation

hiodistribution Structural attributes and substituent

pat-terns w.sponsiblc for optimum pharmacological activity can

he predicted by statistical techniques such as regression

analysis Computerized conformational analysis permits the

medicinal chemist to predict the drug's three-dimensional

shape that is seen by the receptor With the isolation and

structural determination of specific receptors and the

avail-ability of computer software that can estimate the

three-di-mensional shape of the receptor, it is possible to design

mole-cuks that will show an optimum lit to the receptor

OVERVIEW

A drug is a chemical molecule Following introduction into

lie body, a drug must pass through many barriers, survive

alternate sites of attachment and storage and avoid

signifi-cunt metabolic destruction before it reaches the site of action

usually a receptor on or in a cell (Fig 2-I) At the receptor

the following equilibrium (Rx 2-I) usually holds:

Drug + Receptor Drug-Receptor Complex

Pharmacologic Response

(Rx 2-I)

The ideal drug molecule will show favorable binding

char-acienstics to the receptor, and the equilibrium will lie to the

right At the same time, the drug will be expected to

disso-ciate (toni the receptor and reenter the systemic circulation

to he excreted Major exceptions include the alkylating

agents used itt cancer chemotherapy (see Chapter 12) a few

inhibitors of the enzyme acetylcholinesterase (see Chapter

17), suicide inhibitors of monoamine oxidase (see Chapter

14), and the aromatase inhibitors 4-hydroxyandrostenedione

and exemestane (see Chapter 23) These pharmacologicalagents form covalent bonds with the receptor, usually anenxyme's active site In these cases, the cell must destroythe receptor or enzynse, or in the case of the alkylatingagents, the cell would be replaced, ideally with a normal

cell In other words, the usual use of drugs in medical

treat-ment calls for the drug's effect to last for a finite period oftime Then, if it is to be repeated, the drug will be adminis-tered again, lithe patient does not tolerate the drug well, it

is even more important that the agent dissociate from thereceptor and be excreted from the body

DRUG DISTRIBUTION

Oral

An examination of the obstacle course (Fig 2-I) faced by

the drug will give a better understanding of what is involved

in developing a commercially feasible product Assume that

the drug is administered orally The drug must go into

solu-tion to pass through the gastrointestinal mucosa Even drugsadministered as true solutions may not remain in solution asthey enter the acidic stomach and then pass into the alkaline

intestinal tract (This is explained further in the discussion

on acid—base chemistry.) The ability of the drug to dissolve

is governed by several factors, including its chemical

struc-ture, variation in particle size and particle surface area, ture of the crystal form, type of tablet coating, and type

na-of tablet matrix By varying the dosage form and physicalcharacteristics of the drug, it is possible to have a drug dis-solve quickly or slowly, with the latter being the situationfor many of the sustained-action products An example isorally administered sodium phenytoin with which variation

of both the crystal form and tablet adjuvants can significantly

alter the bioavailability of this drug widely used in the ment of epilepsy

treat-Chemical modification is also used to a limited extent to

facilitate a drug reaching its desired target (see Chapter 5)

An example is olsalazine, used in the treatment of ulcerative

colitis This drug is a dimcr of the pharmacologically activemesalamine (5-aminosalicylic acid) The latter is not effec-tive orally because it is metabolized to inactive forms

3

Physicochemical Properties in

Relation to Biological Action

4 Wilson and GisvoldsTextbook of

Figure 2—1 • Summary of drug distribution

before reachingthe colon The dimeric form passes through

a significant portion of the intestinal tract before being

cleaved by the intestinal bacteria to two equivalents of

mesalamine

COOH

0I salaz no

Mesa lwnine

As illustrated by olsalazine any compound passing

through the gastrointestinal tract will encounter a large

num-ber and variety of digestive and bacterial enzymes, which

in theory, can degrade the drug molecule In practice, a new

drug entity under investigation will likely be dropped from

further consideration if it cannot survive in the intestinal

tract or its oral bioavailability is low, necessitating parenteral

dosage forms only An exception would be a drug for which

there is no effective alternative or which is more effective

than existing products and can be administered by an

In contrast, these same digestive enzymes can be usell.advantage Chloramphenicol is water soluble enoughmg/mL) to come in contact with the taste receptors authtongue, producing an unpalatable bitterness To mask ih;intense bitter taste, the palmitic acid moiety is added asester of chloramphenicol' s primary alcohol This reduce.' Ihiparent drug's water soluhility (1.05 mglmL) enough so iLl

it can be formulated as a suspension that passes overbitter taste receptors on the tongue Once in the inlectjit tract, the ester linkage is hydrolyzed by the digestiveases to the active antibiotic chloramphenicol and the set

common dietary fatty acid palmitic acid

ofprodrugs Most prodrugs are compounds that are inaLliir

in their native form but are easily metabolized to theagent Olsalazine and chloramphenicol palmitate are exan

pIes of prodrugs that are cleaved to smaller compounds 0th

of which is the active drug Others arc metabolic

to the active form An example of this ype of prodru;

Intramuscular

or

SubcutaneousInjection

TissueDepots

DRUG

IntravenousInjection

Drug administered directly Into systemic circulation

menadionc a simple naphthoquinone that is converted in

lie liver to phytonadione (vitamin

Menad lane

Phytonadions (Vitamin 1(2(20))

Occasionally, the prodrug approach is used to enhance

the absorption of a drug that is poorly absorbed from the

gastrointestinal tract Enalapril is the ethyl ester of enala

prilic acid, an active inhibitor of angiotensin-converting

en-zyme (ACE) The ester prodrug is much more readily

ab-sorbed orally than the pharmacologically active carboxylic

add.

CH3

Enalapril: R = C2H5

Enalaprilic Acid: R = H

Unless the drag is intended to act locally in the

gustroin-tcstinal tract, it will have to pass through the gastrointestinal

mucosal barrier into venous circulation to reach the site of

the receptor The drug's route involves distribution or

parti-honing between the aqueous environment of the

ga.strointes-tinal tract, the lipid bilayer cell membrane of the mucosal

cells possibly the aqueous interior of the mucosal cells, the

lipid bilayer membranes on the venous side of the

gastroin-(estinal tract, and the aqueous environment of venous

circu-lation Some very lipid-soluble drugs may follow the route

of dietary lipids by becoming part of the mixed micelles

incorporating into the chylomicrons in the mucosal cells into

the lymph ducts, servicing the intestines, and finally entering

venous circulation via the thoracic duct

The drug's passage through the mucosal cells can be

pa.s-sive or active As is discussed below in this chapter the

lipid membranes are very complex with a highly ordered

structure Part of this membrane is a series of channels or

tunnels that form, disappear and reform There are receptors

that move compounds into the cell by a process called

pino-niosis Drugs that resemble a normal metabolic precursor

or intermediate may be actively transported into the cell by

the same system that transports the endogenous compound

On the other hand, most drug molecules are too large to

enter the cell by an active transport mechanism through the

passages The latter, many times, pass into the patient's

cir-culatory system by passive diffusion

ill-oral administration is precluded But that does not mean that

the drug administered by injection is not confronted by

ob-stacles (Fig 2-I) Intravenous administration places the drug

directly into the circulatory system, where it will be rapidly

distributed throughout the body including tissue depots andthe liver, where most biotransformations occur (see below),

in addition to the receptors Subcutaneous and intramuscular

injections slow distribution of the drug because it must fuse from the site of injection into systemic circulation

dif-It is possible to inject the drug directly into specific organs

or areas of the body Intraspinal and intracerebral routes will

place the drug directly into the spinal fluid or brain, tively This bypasses a specialized epithelial tissue, the

respec-blood—brain barrier, which protects the brain from exposure

to a large number of metabolites and chemicals Theblood—brain barrier is composed of membranes of tightlyjoined epithelial cells lining the cerebral capillaries The netresult is that the brain is not exposed to the same variety

of compounds that other organs are Local anesthetics areexamples of administration of a drug directly onto the de-

sired nerve A spinal block is a form of anesthesia performed

by injecting a local anesthetic directly into the spinal cord

at a specific location to block transmission along specific

neurons

Most of the injections a patient will experience in a

life-time will be subcutaneous or intramuscular These parenteral

routes produce a depot in the tissues (Fig 2-I), from whichthe drug must reach the blood or lymph Once in systemiccirculation, the drug will undergo the same distributive phe-

nomena as orally and intravenously administered agents fore reaching the target receptor In general, the same factors

be-that control the drug's passage through the gastrointestinal

mucosa will also determine the rate of movement out of thetissue depot

The prodrug approach described above also can be used

to alter the solubility characteristics, which, in turn, can in.crease the flexibility in formulating dosage forms The solu-bility of methyiprednisolone can be altered from essentially

water-insoluble methylprednisolone acetate to slightly

water-insoluble methylprednisolone to water-soluble ylprednisolone sodium succinate The water-soluble sodiumhemisuccinate salt is used in oral, intravenous, and intramus-

mehh-cular dosage forms Methylprednisolone itself is normallyfound in tablets The acetate ester is found in topical oint-

ments and sterile aqueous suspensions for intramuscular jection Both the succinate and acetate esters are hydrolyzed

in-to the active methylprednisolone by the patient's own

sys-temic hydrolytic enzymes (esterases)

Pharmaceutical Chemi.sirv

Methyiprednisolone: R H

Meth)lprednisolone Acetate: R C(=O}CH3

Methyiprednisolone Sodium Succinate: R = C(0)CH2CH2COO' Na'

Anotherexampleof how prodrug design can significantly

alter biodistribution and biological half-life is illustr,tted by

two drugs based on the retinoic acid structure used

systemi-cally to treat psoriasis a nonmalignant hyperplasia

Etreti-nate has a 120-day "terminal" half-life after 6 months of

therapy In contrast, the active metabolite acitretin has a

33-to 96-hour "terminal" half-life Both drugs are potentially

teratogenic Female patients of childbearing age must sign

statements that they are aware of the risks and usually are

administered a pregnancy test before a prescription is issued

Acitretin, with its shorter half-life, is recommended for a

female patient who would like to become pregnant, because

it can clear her body within a reasonable time frame When

effective etretinate can keep a patient clear of psoriasis

le-sions for several months

Protein Binding

Once the drug enters the systemic circulation (Fig 2-I) it

can undergo several events, It may stay in solution, but many

drugs will be bound to the serum proteins, usually albumintRx 2-2) Thus a new equilibrium must be considered De-

pending on the equilibrium constant, the drug can remain in

systemic circulation bound to albumin for a considerable

period and riot be available to the sites ofthe pharmacological receptors, and excretion

Protein binding can have a profound effect on the drug'seffective soluhility biodistribution half-life in the body andinteraction with other drugs A drug with such poor water

solubility that therapeutic concentrations of the unbound

(ac-tive) drug normally cannot be maintained still can be a very

effective agent The albumin—drug complex acts as a

reser-voir by providing large enough concentrations of free drug

to cause a pharmacological response at the receptor.Protein binding may also limit access to certain body com-partments The placenta is able to block passage of proteins

from maternal to fetal circulation Thus, drugs that normally

would be expected to cross the placental harrier and possiblyharm the fetus are retained in the maternal circulation, bound

to the mother's serum proteins

Protein binding also can prolong the drug's duration of

through the renal glomerular membranes, preventing rapidexcretion of the drug Protein binding limits the amount of

drug available for biotransformation (see below and Chapter4) and for interaction with specific receptor sites For exam-ple, the large polar trypanocide suramin remains in the body

Chapter 2 Properties it, RiolugicalAction 7

in the protein-bound liwni Iiras long months (11,2 =

51) days) The maintenance dose tbr this drug is based on

weekly administration At first, this might seem to be an

advantage to the patient It can be but ii also means that,

¼hould the patient have serious adverse reactions, a

signifi-cam length of tune will be required before the concentration

of drug falls below toxic levels

The drug—protein binding phenomenon can lead to some

clinically significant drug—drag interactions resulting when

one drug displaces another from the binding site on albumin,

A large number of drugs can displace the anticoagulant

war-farm from its albumin-binding sites This increases the

effec-tive concentration of wurfarin at the receptor, leading to an

increased prothrombin time (increased time for clot

forma-tioll)and potential hemorrhage

Tissue Depots

Thedrug can also be stored in tissue depots Neutral fat

constitutes some 20 to 50% of body weight and constitutes

a depot of considerable importance The more lipophilic the

drug, the more likely it will concentrate in these

pharmaco-logically inert depots The ultra-short-acting, lipophilic

bar-biturate ihiopental's concentration rapidly decreases below

its effective concentration following administration It

"dis-appears" into tissue protein, redistributes into body fat, and

then slowly diffuses hack out of the tissue depots but in

concentrations too low for a pharmacological response

Thus, only the initially administered thiopental is present in

high enough concentrations to combine with its receptors

The remaining thiopenlal diffuses out of the tis.sue depots

into systemic circulation in concentrations too small to be

effective (Fig 2-I) is metabolized in the liver, and is

ex-creted.

In general structural changes in the barbiturate series (see

Chapter 14) that favor partitioning into the lipid tissue stores

decrease duration of action but increase central nervous

sys-tem (CNS) depression Conversely, the barbiturates with the

slowest onset of action and longest duration of action contain

the more polar side chains This latter group of barbiturates

both enters and leaves the CNS more slowly than the more

lipophilic thiopental

Drug Metabolism

All substances in the circulatory system, including drugs,

molecules absorbed from the gastrointestinal tract enter theportal vein and are initially transported to the liver A signifi-

cant proportion of a drug will partition or be transported

into the hepatocyte, where it may be metabolized by hepatic

enzymes to inactive chemicals during the initial trip through

the liver, by what is known as the first-pass effect (see

Chap-ter4).

Lidocaine is a classic example of the significance of thefirst-pass effect Over 60% of this local anesthetic antiar-rhythmic agent is metabolized during its initial passage

through the liver, resulting in it being impractical to ter orally When used for cardiac arrhythmia.s, it is adminis-

adminis-tered intravenously This rapid metabolism of lidocaine isused to advantage when stabilizing a patient with cardiac

arrhythmias Should too much lidocaine be administered

in-travenously, toxic responses will tend to decrease because

of rapid biotransformation to inactive metabolites An

under-standing of the metabolic labile site on lidocainc led to thedevelopment of the primary amine analogue tocainide Incontrast to lidocaine's half-life of less than 2 hours, tocai-nide's half-life is approximately IS hours, with 40% of the

drug excreted unchanged The development of orally activeantiarrhythmic agents is discussed in more detail in Chapter

A study of the metabolic fate of a drug is required for all

new drug products Often it is found that the metabolites arealso active Indeed, sometimes the metabolite is the pharma-cologically active molecule These drug metabolites can pro-

vide leads for additional investigations of potentially newproducts Examples of an inactive parent drug that is con-

Na

Sodium

8 Wilson and Giscolds Textbook of Organic Medicinal and Pharmaceutical Chemistry

inflammatory agent sulinduc being reduced to the active

sul-tide metabolite: the immunosuppressant azathioprine being

cleaved to the purinc antimetabolite 6-mercaptopunne; and

purine and pyrimidinc antimetabolites and antiviral agents

being conjugated to their nucleotide form (acyclovir

phos-phorylated to acyclovir triphosphate) Often both the parent

drug and its metabolite are active, which has ted to additional

commercial products, instead of just one being marketed

About 75 to 80% of phenacetin (now withdrawn from the

U S market) is converted to acetaminophen In the tricyclic

antidepressant series (see Chapter 14) imipramine and

ami-triptyline are N-deniethylated to desipramine and

nortripty-line, respectively All four compounds have been marketed

in the United States Drug metabolism is discussed more

Although a drug's metabolism can be a source of

frustra-tion for the medicinal chemist, pharmacist and physicianand lead to inconvenience and compliance problems withthe patient, it is fortunate that the body has the ability tometabolize foreign molecules (xenobiotics) Otherwise

many of these substances could remain in the body for years.This has been the complaint against certain lipophilic chemi-

cal pollutants, including the once very popular insecticide

DDT After entering the body, these chemicals reside in body

tissues, slowly diffusing out of the depots and potentiallyharming the individual on a chronic basis for several years.They can also reside in tissues of commercial food animalsthat have been slaughtered before the drug has "washedout" of the body

Themain route of excretion of a drug and its metabolites isthrough the kidney For some drugs enterohepatic circula-tion (Fig 2-I) in which the drug reenters the intestinal tract

from the liver through the bile duct, can be an important part

of the agent's distribution in the body and route of excretion.Either the drug or drug nietabolite can reenter systemic circu-lation by passing once again through the intestinal mucosa

A portion of either also may be excreted in the feces Nursingmothers must be concerned because drugs and their metabo-

lites can be excreted in human milk and be ingested by thenursing infant

One should keep a sense of perspective when learningabout drug metabolism As explained in Chapter 4 drug

metabolism can be conceptualized as occurring in two stages

or phases Intermediate metabolites that are

pharmacologi-cally active usually are produced by phase I reactions The

products from the phase I chemistry are converted into

inac-tive, usually water-soluble end products by phase II (ions The latter, commonly called conjugation reactions

reac-can be thought of as synthetic reactions that involve addition

of water-soluble substiiucnts In human drug metabolism

the main conjugation reactions add glucuronic acid, sulfate

or glutathione Obviously, drugs that are bound to serum

protein or show favorable partitioning into (issue depots are

going to be metabolized and excreted more slowly for the

reasons discussed above

This does not mean that drugs that remain in the body forlonger periods of time can be administered in lower doses or

be taken fewer times per day by the patient Several variablesdetermine dosing regimens, of which the affinity of the drugfor the receptor is crucial Reexamine Reaction 2-I and Fig-

ure 2-I If the equilibrium does not favor formation of thedrug—receptor complex, higher and usually more frequentdoses must be administered Further, if partitioning into tis-sue stores or metabolic degradation and/or excretion is fa-

vored, it will take more of the drug and usually more frequentadministration to maintain therapeutic concentrations at thereceptor

Receptor

With the possible exception of general anesthetics (seeChapter 14) the working model for a pharmacological re-sponse consists of a drug binding to a specific receptor

Many drug receptors are the same as those used by

endoge-nously produced ligands Cholincrgic agents interact with

Chapter 2 • 9the same receptors as the neurotransrnitter acetylcholine.

Synthetic corticosteroids bind to the same receptors as

corti-sone and hydrocorticorti-sone Often, receptors for the same

Ii-gand are found in a variety of tissues throughout the body

The nonsteroidal anti-inflammatory agents (see Chapter 22)

inhibit the prostaglandin-fomiing enzyme cyclooxygenuse

which is found in nearly every tissue This class of drugs

has a long list of side effects with many patient complaints

Note in Figure 2-I that, depending on which receptors

con-tain bound drug there may be desired or undesired effects

This is because a variety of receptors with similar structural

requirements are found in several organs and tissues Thus

the nonsteroidal anti-inflammatory drugs combine with the

desired cyclooxygenase receptors at the site of the

inflamma-tion and the undesired cyclooxygenase receptors in the

gastroiinestinal mucosa causing severe discomfort and

sometimes ulceration One of the "second-generation"

untihistamines is claimed to cause less

seda-tion because it does not readily penetrate the blood—brain

barrier The rationale is that less of this antihistamine is

available for the receptors in the CNS which are responsible

for the sedation response eharactenstic of anlihistamines In

contrast, some antihistamines are used for their CNS depres

sam activity because a significant proportion of the

adminis-tered dose is crossing the blood—brain barrier relative to

binding to the histamine H1 receptors in the periphery

Although ii is normal to think of side effects as

undesira-ble, they sometimes can he beneficial and lead to new

prod-ucts The successful development of oral hypoglycemic

agents used in the treatment of diabetes began when it was

found that certain sulfonamides had a hypoglycemic effect

Nevertheless, a real problem in drug therapy is patient

com-pliance in taking the drug as directed Drugs that cause

seri-ous problems and discomfort tend to be avoided by patients

Swnmary

One of the goals is to design drugs that will interact with

receptors at specific tissues There are several ways to do

this, including (a) altering the molecule, which, in turn, can

change the hiodistribution; (b) searching for structures that

show increased specificity for the target receptor that will

produce the desired pharmacological response while

de-creasing the affinity for undesired receptors that produce

adverse responses: and (c) the still experimental approach

of attaching the drug to a monoclonal antibody (see Chapter

7) that will bind to a specific tissue antigenic for the

anti-body Biodistribulion can be altered by changing the drug's

solubility enhancing its ability to resist being metabolized

usually in the liver), altering the fortnulation or physical

characteristics of the drug, and changing the route of

admin-istration If a drug molecule can be designed so that its

bind-ing to the desired receptor is enhanced relative to the

unde-sired receptor and biodistribution remains favorable, smaller

doses of the drug can be administered This, in turn, reduces

the amount of drug available for binding to those receptors

responsible for its adverse effects

The medicinal chemist is confronted with several

chal-lenges in designing a bioactive molecule A good fit to a

specific receptor is desirable, but the drug would normally

be expected to dissociate from the receptor eventually The

specificity for the receptor would minimize side effects The

time Its rate of metabolic degradation should allow ble dosing schedules and, ideally, oral administration Many

reasona-times, the drug chosen for commercial sales has been

se-lected from hundreds of compounds that have been screened

It usually is a compromise product that meets a medical needwhile demonstrating good patient acceptance

ACID—BASE PROPERTIES

Most drugs used today can be classified as acids or bases

As is noted shortly a large number of drugs can behave as

either acids or bases as they begin their journey into thepatient in different dosage forms and end up in systemic

circulation A drug's acid—base properties can greatly

inilu-ence its biodistribution and partitioning characteristics

Over the years at least four major definitions of acids andbases have been developed The model commonly used in

pharmacy and biochemistry was developed independently

by Lowry and Brønsted In their definition, an acid is defined

as a proton donor and a baseisdefined as a proton acceptor

Notice that for a base, there is no mention of the hydroxideion

in aqueous media, to cphedrine (reaction h), which is an

excellent proton acceptor

Notice the diversity in structure of these proton donors

They include the classical hydrochloric acid (reaction a) the

weakly acidic dihydrogen phosphate anion (reaction b), theammonium cation as is found in ammonium chloride (reac-tion c), the carboxylic acetic acid (reaction d) the enolicform of phenobarbital (reaction e), the carboxylic acidmoiety of indomethacin (reaction J), the imide of saccharin

(reaction g), and the protonated amine of ephedrine (reaction

It) Because all are proton donors, they must be treated as

acids when calculating the pH of a solution or percent

ioniza-tion of the drug At the same time, as noted below, thereare important differences in the pharmaceutical properties

of ephedrine hydrochloride (an acid salt of an amine) and

those of indomethacin phenobarbital or saccharin

Base-Conjugate Add

The Brønsted-Lowry theory defines a base as a molecule

that accepts a proton The product resulting from the addition

of a proton to the base is the c'onjugate acid cally important bases are listed in Table 2-2 Again, there

Pharmaceuti-are a variety of structures, including the easily recognizablebase sodium hydroxide (reaction a): the basic component of

an important physiological buffer, sodium monohydrogenphosphate (reaction b), which is also the conjugate base ofdihydrogen phosphate (reaction b in Table 2-I); ammonia

(reaction c), which is also the conjugate base of the

ammo-nium cation (reaction c in Table 2-I); sodium acetate

tion d), which is also the conjugate base of acetic acid

(reac-tion d in Table 2-I); the enolate form of phenobarbital

(a) Hy&ochlonc acid

The Midiwu muon and chlonde anion do not mike paul In

(reaction e), which is also the conjugate base of

phenobarbi-tal (reaction e in Table 2-I the carboxylate form of

indo-methacin (reaction ft which is also the conjugate base of

indomethacin (reaction fin Table 2-I); the imidate form of

saccharin (reaction g) which is also the conjugate base of

saccharin (reaction g in Table 2-I); and the amine ephedrine

(reaction I,), which is also the conjugate base of ephedrine

hydrochloride (reaction h in Table 2-I) Notice that the

con-jugate acid products in Table 2-2 are the reactant acids in

in Table 2.1 are the reactant bases in Table 2-2 Also, noticethat whereas phenobarbital, indomethacin, and saccharin are

un-ionized in the protonated form, the protonated (acidic)

forms of ammonia and ephedrine are ionized salts (Table

2-I ) The opposite is true for the basic (proton acceptors) forms

of these drugs The basic forms of phenobarbital

indometha-cm, and saccharin are anions, whereas ammonia and

ephed-rine are electronically neutral (Table 2-2) Remember thateach of the chemical examples in Tables 2-I and 2-2 can

Chapter 2 •Phv,ci,-oclu',,,ical Properties in Relation a, Biological Action 11

(a) Sodium hydroxide

(base) This can best be understood by emphasizing the

con-of conjugate acid—conjugate base pairing Complicated

as ii may seem at lit-st conjugate acids and conjugate bases

are nothing more than the products of an acid—base reaction

In other words, they appear to the right of the reaction

ar-rows Examples from Tables 2-I and 2-2 are rewritten in

Table 2-3 as complete acid—base reactions

careful study of Table 2-3 shows water functioning as a

proton acceptor (base) in reactions a c- e g i k and in and

a proton donor (base) in reactions I,, d.f 11.1, I and a, Hence

water is known as an ti,nphotes-,e substance Water can be either a weak base accepting a proton to form the strongly acidic hydrated proton or hydroniuni ion 1-1.10 (reactions

a c x' i,k and in) or a weak acid donating a proton to form the strongly basic (proton accepting) hydroxide anion OH- (reactions b, d I.,), I and a).

Acid Strength

Whileany acid—base reaction can be wrttten as an rium reaction, an attempt has been made in Table 2-3 to

Base + H' —. Conjugate Add

\

OH + H'

0

IiC

12 and of

TABLE 2—3 Examples of Acid—Base Reactions (WiththeException of Hydrochloric Acid, Whose Conjugate Base(C1) Has No Basic Properties in Water and Sodium Hydroxide Which Generates Hydroxide, the Reaction of theConjugate Base in Water is Shown for Each Acid)

Add + Base = Conjugate Acid + Conjugate BaseHydrochlonc acid

(a) HCI + H2O — H30 +

(h) H3O + CH3COO1NaIa _ CH3COOH +

Indornelhacan and its conjugate base Indomethacin sodium, show the Identical acid—base chemistry as aceticactd

!'ropcr,iesin Relation to Biological Action 13tndicatc which sequences are unidirectional or show only a

small reversal.For hydrochloric acid, the conjugate base

Cl is such a weak base that it essentially does not function

as a proton acceptor That is why the chloride anion was not

included as a base in Table 2-2 In a similar manner, water is

such a weak conjugate acid that there is little reverse reaction

involving water donating a proton to the hydroxide anion of

Two logical questions to ask at this point are how one

in which direction an acid—base reaction lies and

to what extent the reaction goes to completion The common

physical chemical measurement that contains this

informa-tion is known as the pK, The pK, is the negative logarithm

of the modified equilibrium constant K, for an acid—base

reaction written so that water is the base or proton acceptor

It can he derived u.s follows:

Assunie that a sveak acid HA reacts with water

Acid Base Acid Base

HA + U.O = H,0 + A

The equilibrium constant K01 for Reaction 2-3 is

— — lacidlihasel

In a dilute solution of a weak acid, the molar concentration

of water can be treated as a constant, 55.5 M This number

is based on the density of water equaling I Therefore I L

of water weighs 1000g With a molecular weight of 18, the

molar concentration of water inI L of water is

The modified equilibrium constant K, is customarily

converted to pK, (the negative logarithm) to use on the same

scale as pH Therefore, rewriting Equation 2-2 in logarithmic

fonn produces

Rearranging Equation 2-5 gives

Equation 2-7 ix more commonly called the

Henderson-Hasselbalch equation and is the basis for most calculations

involving weak acids and bases It is used to calculate the

pH of solutions of weak acids, weak bases, and buffers sisting of weak acids and their conjugate bases or weak bases

con-and their conjugate acids Because the pK, is a modifiedequilibrium constant, it corrects for the fact that weak acids

do not completely react with water

A very similar set of equations is obtained from the tion of a protonated amine BH in water The reaction is

Acid Base Acid Base

The equilibrium constant

K — — Iconi acid llconi bawl

Rearranging Equation 2-9 into logarithmic form and stituting the relationships expressed in Equations 2-3 and2-4 yields the same Henderson-Hasselbalch equation (Eq

Henderson-Hasselbalch equation for an HA or BH acid, it

is simpler to use the general form of the equation (Eq II) expressed in both Equations 2-7 and 2-10

Originally, a modified equilibrium constant, the pK5, was

derived following the same steps that produced Equation

2-2 it is now more common to express the basicity of a

chemi-cal in terms of the pK, using the relationship in Equation

2-12

Notice that Equation 2-8 is identical to Equation 2-I when

the general [conj acid Jlconj base] representation is used.IRs 23) Therefore, using the same simplifying assumption that svater

remains at a constant concentration of 55.5 M in dilute

solu-tions Equation 2-8 can be rewritten as

(Eq 2.1)

Thu.' with [H201 = 55.5, Equation 2-I can be simplified

= Iconi acid Jlconj bascl

lucidl (Eq 2-2)

pK., = —log K, (Eq 2-3) With this version of the equation, there is no need to

re-member whether the species in the numerator/denominator

con-(Eq 2-4) centration of the proton acceptor is the term in the numerator

and the molar concentration of the proton donor is the

de-nominator term

What about weak bases such as amnines? In aqueous

solu-tions water functions as the proton donor or acid (Rx 2-5).producing the familiar hydroxide anion (conjugate base)

= log H,0 I + log lconj busel — log lacidl

—log 1110' I = —log K,, + log lA1 — log (HAl (Eq 2-6)

= —log K, + log lconj hasel — log laeidl

Substituting Equations 2-3 and 2-4 into Equation 2-6

pro-duces

TABLE 2—4 Examples of Calculations Requiring the

1 What Is the ratio of eptiedulne to ephedrune l'lCt in the

Intestinal tract at pH 8.0? Lisa Equation 2-11

[ephedrune)

—1.6

80—9.6+tog

[pheliHOl[ephedrunel

0025

[ephedrine HCI]

The number whose tog is —1.6 Ia 0025, meaning that there are

25 parts ephedrine for every 1000 pails ephedrine Nd in the

intestinal tract whose environment is pH 8.0

2 What Is the pH ota buffer containing 0.t M acetic acid (p1<5 48)

end 0.08 M sodium acetate? Use Equation 2-11

0.08

3 What Is the pH of a 0.1 M acetic acid solution9 Use the following

equation for calculating the pt-f of a solution containing either an

I-tAor 8H acid

pH—pK5—log[acld]

=29

4 What Is the pH ot a (1.08 M sodium acetate solution? Remember

even though this is the conjugate base of acetic acid, the is

sf4 used The pK_ term In the following equation corrects tor

the fact that a proton acceptor (acetate anion) 'a present in the

solution The equation for calculating the pH of a solution

contain-ing either an or B base Is

+ pK5 + log (basel

5 What Is the pt-f of an enynonsim acetate sotution9 The pK5 01 the

amrnonltmi (NH4I cation is 9.3 Always bear in mund that the

refers to the ot the proton donor form to release the proton

Into water to form H304 Since this is the salt of a weak acid

(NH4) and the conlugate base of a weak acid (acetate anion),

the following equation is used Note that molar concentration is

rIot a vartable in this calculation

= 7 1

6 What Is the percentage Ionization of ephednne Nd (p1<5 9.6) In an

Intestinal tract buffered at 8.0 (see example 1)') Use Equation

2-14 because thus is a BH acid

1. is the percentage ionization of indocnethacin (p1<5 4.5) in an

intestinal tract buffered at pH 8.0? Use Equation 2-13 because

this is an HA acid,

tOt)

% ionization

— 1 + 1014 oi

For all practical purposes imidomethacin is present only as the

anionic conjugate base in that region of the intestine buttered at

p1-18.0.

Warning! It is imporkmttorecognize thai a pK, for a base

is in reality the of the conjugate acid (acid donor orprotonated form, BH I of the base The pK is listed in the

Appendix as 9.6 for ephedrine and as 9.3 for ammonia in

reality, this is the of the protonated form, such as rinc hydrochloride (reaction in in Table 2-31 and ammonium

ephed-chloride (reaction e in Table 2-3) respectively This is

con-fusing to students, pharmacists, clinicians, and experiencedscientists It is crucial that the chemistry of the drug be under-

stood when interpreting a pK value When reading tables

of pK5 values, such as those lound in the Appendix, one

must realize that the listed value is for the proton donor form

of the molecule, no matter what form is indicated by thename See Table 2-4 for several worked examples of how

the pK5 is used to calculate pHs of solutions, required ratios

of Iconjugate hasel/lacidi and percent ionization at specific

pHs

Just how strong or weak are the acids whose reactions in

water are illustrated in Table 2-3'! Remember that the

or are modified equilibrium constants that indicate the

extent to which the acid (proton donor) reacts with water to

form conjugate acid and conjugate base The equilibrium for

a strong acid (1(1W in water lies to the right, favoringthe formation of products (conjugate acid and conjugatebase) The equilibrium for a weak acid (high pK,; in waterlies to the left, meaning that the conjugate acid is a betterproton donor than the parent acid is or that the conjugate

base is a good proton acceptor

Table 2-5, substitute the term for each of the acids Forhydrochloric acid, a K, of 1.26 x I means that the product

of the molar concentrations of the conjugate acid [H10'and tile conjugate base Cli is huge relative to the denomi.nator term IHCII In other words, there essentially is nounreacted HC1 leli in an aqueous solution of hydrochloricacid At the other extreme is ephedrine HCI with a pK, of

9.6 or a of 2.51 x 10_Ia Here, tile denominator ing the concentration of ephedrine 1-ICI greatly predominatesover that of the products, which, in this example is ephedrine(conjugate base) and H10' (conjugate acid) In other words.the protonated form of ephedrine is avery poor proton donor

represent-It holds (Into the proton Indeed, free ephedrine (the

conju-gate base in this reaction) is an excellent proton acceptor

A general rule for determining whether a chemical isstrong or weak acid or base is

pK5 < 2: strong acid: conjugate base has no meaningful basicproperties in water

pK5 4—6: weak acid; weak conjugate base

pK, it— tO: very weak acid; conjugate hase getting srnmgerpK,, >12: essentially nui acidic properties in water: strong conju-gate base

This delineation is only approximate Other propertiesalso become important when considering cautions in han

dung acids and bases Phenol has a pK of 9.9 slightly lessthan that of ephedrine HCI Why is phenol considered corro-

sive to the skin, whereas ephedrine HCI or free ephedrine

is considered innocuous when applied to the skin'! Phenolhas the ability to partition through the normally protectivelipid layers of the skin Because of tilis property tilis ex-

Chapter 2 • PhvskochL'micalPropertiesinRelation to Riologiea/ Aelion IS

Reactions Listed in Table 2-3 (See the Appendix)

the simply tells a person the acid properties of the

pro-tonated form of the chemical It does not represent anything

else concerning other potential toxicities

Percent Ionization

Usingthe drug's pK0 the formulation or compounding

phar-macist can adjust the pH to ensure maximum water solubility

tionic form ol'the drug) or maximum solubility in nonpolar

media (nonionic form) This is where understanding the

drag's acid—base chemistry becomes important Note

Reac-tions 2-6 and 2-7:

Acid Base Acid Base

H20 + (Rx 2-7)

Acids can be divided into two types,HAand BH - on

the basis of the ionic form of the acid (or conjugate base)

HA acids go from un-ionized acids to ionized conjugate

bases IRs 2-6) In contrast, BH acids go from ionized

(polar) acids to un-ionized (nonpolar) conjugate bases (Rx

2-71 In general pharmaceutically important HA acids

in-dude the inorganic acids (e.g HCI, FI.S03), enols (e.g

barbiturates, hydantoins), carboxylic acids (e.g

low-molec-ular-weight organic acids, arylacetic acids, N-aryl

anthra-ailic acids salicylic acids), and amides and irnides (e.g

sulfonamides and saccharin, respectively) The chemistry is

simpler for the pharmaceutically importantBH acids: They

are aH protonated amines A poIyfunctional drug can have

several pK0s (e.g amoxicillin) The latter's ionic state is

based on amoxicillin's ionic state at physiological pH 7.4

(see the discussion below on percent ionization)

9,6

CH3/

ion-ized (or 50% un-ionion-ized) In other words, when the pK, isequal to the pH the molar concentration of the acid equals

the molar concentration of its conjugate ba.se In the

Hender-son-Hasselbalch equation pK = pH when log Iconj basel/tacidi = I An increase of I pH unit from the pK., (increase

in alkalinity) causes an HA acid (indomethacin) to become90.9% in the ionized conjugate base form but results in a

BH acid (ephedrine HCI) decreasing its percent ionization

to only 9.1% An increase of 2 pH units essentially shifts

an HA acid to complete ionization (99%) and a BK' acid

to the nonionic conjugate base form (0.99%)

Just the opposite is seen when the medium is made moreacidic relative to the drug's pK., value Increasing the hydro-gen ion concentration (decreasing the pH) will shift the equi-

librium to the left, thereby increasing the concentration of

the acid and decreasing the concentration of conjugate base

In the case of indomethacin a decrease of I pH unit belowthe pK will increase the concentration of un-ionized (pro-tonated) indomeihacin to 9.1% Similarly, a decrease of 2

pH units results itt only 0.99% of the indomethacin beingpresent in the ionized conjugate base lonu The opposite is

seen for the BH acids The percentage of ephedrine present

as the ionized (protonated) acid is 90.9% atI pH unit below

results are summarized in Table 2-6

With this knowledge in mind, return to the drawing ofamoxicillin At physiological pH the carboxylic acid (HA

acid: 2.4) will be in the ionized carhoxylate form, the

indonsethacin Episadrinc

90

80 40

FIgure 2—2 a Percent ionized versus pH for indomethacin (pK,

The percent ionization of a drug is calculated by using 4.5) and ephedrine

I 2 3 4 6 1 8 I 40 II 47 13 4

pH

50.0

9.1

0.99

primary amine (RH acid: pK.,2741will be 50% protonated

and 50% in the free amineform,and the phenol (HA acid:

9.6) will be in the un-ionized prolonated form A

knowledge of percent ioniiation makes it easier to explain

and predict why the use of some prcparations can cause

problems and discomibri as a result of pH extremes

Phenyt-oin (HA acid: pK, 8.3) injection mustbeadjusted to pH 12

with sodium hydroxide to ensure complete ionization and

maximize water soluhility In theory, a pH of 10.3 will result

in 99.0% of the drug being an anionic water-soluble

conju-gate base To lower the concentration of phenytoin in the

insoluble acid form even further and maintain excess

alkalin-ity, the pH is raised to 12 to obtain 99.98% of the drug in

the ionizcd form Even then, a cosolvent system of 40%

propylene glycol 10% ethyl alcohol, and 50% waler for

injection is used to ensure complete solution This highly

alkaline solution is irritating to the patient and generally

cannot be administered as an admixture with other

intrave-nousfluids that are buffered more closely at physiological

pH 7.4 This decrease in pH would result in the parent

un-ionized phenytoin precipitating out of solution

QI

NH-S0' Na

Phenytoin Sodiwn

Tropicarnide is an anticholinergic drug administered as

eye drops for its mydriatic response during eye

pH 4 toobtain more than 90% ionization The acidic eye

drops can sting Some optometrists and ophthalmologists

use local anesthetic eye drops to minimize the patient's

pyri-dine nitrogen The amide nitrogen has no acid—base

proper-ties in aqueousmedia

H-jO i=s H30'÷A

UBarrier

HA + H20 — H30 + A

Figure2—3 • Passageof HA acids through lipid barriers

Adjustments in pHtomaintain water solubility can times lead to chemical stability problems An example isindomethacin (HA acid: pK, 4.5), which is unstable in alka-line media Therefore, the preferred oral liquid dosage form

some-is a suspension buffered at pH 4 to 5 Because thsome-is some-is nearthe drug's pK, only 50% will be in the water-soluble form

There is a medical indication requiring intravenous

adminis-tration of indomethacin to premature infants The nous dosage form is the lyophilized (freeze-dried) sodiumsalt, which is reconstituted just prior to use

intrave-Drug Distribution and plc

The pK, can have a pronounced effect on the

pharmacoki-netics of the drug As discussed above, drugs are transported

in the aqueous environment of the blood Those drugs in anionized form will tend to distribute throughout the body more

exceptions. the drug must leave the polar environment of

the plasma to reach the site of action In general, drugs pass

through the nonpolar membranes of capillary walls, cellmembranes, and the blood—brain harrier in the un-ionized(nonpolar) tomi For HA acids, it is the parent acid that willreadily cross these membranes (Fig 2-3) The situation isjust the opposite for the acids The un-ionized conju-gate base (tree amine) is the species most readily crossingthe nonpolar membranes (Fig 2-4)

the drug molecule orally administered The drug first

en-counters the acidicstomach, where the pHcan range from

2 to 6 depending on the presence of food HA acids withpK,s of 4 to 5 will tend to be nonionic and be absorbed

partially through the gastric mucosa (The main reason mostacidic drugs are absorbed from the intestinal tract rather than

the stomach is that the inicrovilli of the intestinal mucosaprovide a huge surface area relative to that found in thegastric mucosa of the stomach.) In contrast amines (pK 9

to 10) will be protonated (BH acids) inthe acidic stomach

17and usually will not be absorbed until reaching the mildly

alkaline intestinal tract pH —8) Even here, only a portion

of the amine-containing drugs will be in their nonpolar

con-jugate base form (Fig 2-4) Remember that the reactions

shown in Figures 2-3 and 2-4 are equilibrium reactions with

K,, values Therefore, whenever the nonpolar form of either

an HA acid (as the acid) or a B base (the conjugate base of

the BH acid) passes the lipid harrier, the ratio of conjugate

base to acid (percent ionization) will be maintained Based

on Equations 2-13 and 2-14 this ratio depends on the

(a constant) and the pH of the medium

For example, once in systemic circulation, the plasma p11

oF 7.4 will he one of the determinants of whether the drug

will tend to remain in the aqueous environment of the blood

or partition across lipid membranes into hepatic tissue to be

metabolized, into the kidney for excretion, into tissue depots,

or to the receptor tissue A useful exercise is to calculate

either the Iconj base I/lucid! ratio using the

Henderson-Has-selbalch equation (Eq 2-Il) or percent ionization for ephed

tine (pK,,9.6; Eq 14) and indomethacin (pK,, 4.5: Eq

2-13) at pH 3.5 (stomach) pH 8.0 (intestine), and pH 7.4

(plasma) (see examples 1,6, and 7 in Table 2-4) Of course,

the effect of protein binding, discussed above, can greatly

alter any prediction of biodistrihution based solely on

STATISTICAL PREDICTION OF

PHARMACOLOGICAL ACTIVITY

Just as mathematical modeling is used to explain and model

many chemical processes, it has been the goal of medicinal

chemists to quantify the effect of a structural change on a

defined pharmacological response This would meet three

goals in drug design: (u) to predict biological activity in

untested compounds, (h) to define the structural

require-ments required for a good fit between the drug molecule and

the receptor, and (c) to design a test set of compounds to

maximize the amount of information concerning structural

requirements for activity from a minimum number of

com-pounds te.sted This aspect of medicinal chemistry is

coin-monly referred to as quantitative structure—activity

relation-ships (QSAR)

The goals of QSAR studies were tirst proposed about 1865

to (870 by Cram-Brown and Fraser who showed that the

grddual chemical modification in the molecular structure of

a series of poisons produced some important differences in

their action They postulated that the physiological action,

of t molecule is a function of its chemical constitution

C This can be expressed in Equation 2-IS:

'P = flC')Equation 2-IS states that a defined change in chemical

structure results in a predictable change in physiological

ac-tion The problem now becomes one of numerically defining

chemical structure It still is a fertile area of research What

has been found is that biological response can he predicted

front physical chemical properties such as vapor pressure

water solubility electronic parameters, steric descriptors

and partition coefficients (Eq 2-16) Today the partition

coefficient has beconte the single most important physical

chemical tneasurernettt for QSAR studies Note that

Equa-tiOn 2-lb is the equation for a straight line (Y = sn.s + I,)

where

BR = a defined pharmacological response usually expressed in

miltimoles such as the mnhibbory constant K, the cffecmivedose in St)'l of thc subjects the lethal dose in fit)% of

the subjects (LD51) or the minimum inhibitory concentration(MIC) It is common to express the biological response as a

reciprocal I/BR or 1/C

a = the regression coefficient or slope of the 'araight line

c = the intercept term on the yaxis (when tIme physical chemicalprtmpcny equals zero)

To understand the concepts in the next few paragraphs

it is necessary to know how to interpret defined

pharmaco-logical concepts such as the ED511 which is the amount of thedrug needed to obtain the defined pharmacological response

is 2 mmol Drug A is twice as potent u.s drug B In

etc.) the more potent the substance being tested

The logarithmic value of the dependent variable tration necessary to obtain a defined biological response) isused to linearize the data As shown below, QSARs are not

(concen-always linear Nevertheless, using logarithms is an able statistical technique (taking reciprocals obtained tnim

accept-a Michaccept-aelis-Menton study produces the lineaccept-ar Lineweaccept-aver-Burke plots found in any biochemistry textbook)

Lineweaver-Now, why is the biological response usually expressed u.s

a reciprocal? Sometimes, one obtains a statistically morevalid relationship More importantly expressing the biologi-

cal response as a reciprocal usually produces a positive slope(Fig 2-7) Let us examine the following published example

The mechanism of death is general depression of the CNS.Table 2-7 contains the pertinent data

The most lethal compound in this assay was zinc, with a BR (LDt151) of only 0.0(XX)063l mmol: and the

chlorproma-least active was ethanol, with a BR of 0.087096 nimol In

other words, it takes about 13,800 tintes as many milliniolc,s

of ethanol than of chlorpromatine to kill 100% of the test

subjects in this particular assay

Let's plot BR versus PC (partition coefficient) Figure

2-5 shows the scatter and an attempt at determining a linearfit for the relationship Note that compounds I and II lie at

a considerable distance from the remaining nine compounds

In addition to the 13.8(X) times difference in activity, there

is a 33,9(X) times difference in the octanol/watcr partitioncoefficient Also, the regression line whose equation is

(Eq 215) is meaningless statistically The slope isO, meaning that the

partition coefficient has no effect on biological activity, andyet from the plot and Table 2-7 it is obvious that the higher

the octanol/water partition coefficient, the more toxic thecompound The correlation coefficient is 0.05 meaning

that there is no significant statistical relationship betweenactivity and partition coefficient

Now, let's see if the damn 'an be linearized by using the

logarithms of the biological activity and partition coefticient.Notice that as logarithmic terms, the difference between the

LD1n, of ehlorpromazine and ethanol is only 4.14 mic units Similarly, the difference between chlorproma-

logarith-tog BR = a(physicat chemical property) c

lFiq 2-161

As cmphasizcd above, the drug will go through a series

of partitioning steps: (a) teaving the aqueous extracellular

fluids (b)passingthrough lipid membranes, and (e) entering

other aqueous environments before reaching the receptor

(Fig 2-I) In this sense, a drug is undergoing the same

parti-tioning phenomenon that happens to any chemical in a

sepa-ratory funnel containing water and a nonpolar solvent such

as hexane, chloroform, or ether The difference between the

separatory funnel model and what actually occurs in the body

is that the partitioning in the funnel will reach an equilibrium

at which the rate of chemical leaving the aqueous phase and

entering the organic phase will equal the rate of the chemical

moving from the organic phase to the aqueous phase This

is not the physiological situation Refer to Figure 2-1 and

note that dynamic changes are occurring to the drug, such

as it being metabolized bound to serum albumin, excreted

from the body and bound to receptors The environment for

the drug is not static Upon administration, the drug will be

pialtedthroughthe membranes because of the high

concen-tration of drug in the extracellular fluids relative to the

con-centration in the intracellular cornpartment.s In an attempt

to maintain equilibrium ratios, the flow of the drug will be

from systemic cireulation through the membranes onto the

receptors As the drug is metabolized and excreted from the

body it will be pulled backacross the membranes, and the

concentration of drug at the receptors will decrease

Because much of the time the drug's movement across

membranes is a partitioning process, the partition coefficient

has become the most common physicochemical property

The question that now must be asked is what initniscible

nonpolar solvent system best mimics the water/lipid

mem-brane barriers found in the body? It is now realised that

the n-ocianol/water system is an excellent estimator of drug

partitioning in biological systems Indeed, one could argue

that it was fortuitous that n-octanol was available in

reasona-ble purity for the early partition coefficient determinations

To appreciate why this is so, one must understand the

chemi-cal nature of the lipid menthranes

These membranes are not exclusively anhydrous fatly or

oily structures As a first approximation, they can be

consid-ered bilayers composed of lipids consisting of a polar cap

and large hydrophobic tail Phosphoglycerides arc major

components of lipid hilayers Other groups of hifunctional

lipids include the sphingoniyelins, galactocerebrosides, and

plasmalogens The hydrophobic portion is composed largely

of unsaturated fatty acids, mostly with ci,s double bonds In

addition, there are considerable amounts of cholesterol

es-ters, protein, and charged mucopolysaecharidcs in the lipid

membranes The final result is that these membranes are

highly organized structures composed of channels for

trans-pan of important molecules such us metaholites chemical

regulators (hormones) amino acids, glucose, and fatty acids

into the cell and removal of waste products and

biochemi-cally produced products out of the cell The cellular

tnem-branes are dynamic with the channels forming and

disap-pearing depending on the cell's and body's tteeds (Fig

2-8).

In addition, the tnenihrane.s on the surface of nucleated

cells have specific antigenic markers major

histocompatibil-ity complex (MHC), by which the immune system monitors

the cell's status There are receptors on the cell surface where

where the virus reproduces As newer instrumental

tech-niques are developed, and genetic cloning permits isolation

of the genetic material responsible for forming and

regulat-ing the structures on the cell surface, the image of a passive

lipid menthrutie has disappeared to be replaced by a verycomplex, highly organized dynamically functioning struc-

ture

For purposes of the partitioning phenomenon picture the

cellular membranes as two layers of lipids (Fig 2-9) The

two outer layers one facing the interior and the other facingthe exterior of the cell, consist of the polar ends of the bifunc-tional lipids Keep in mind that these surfaces are exposed to

an aqueous polar environment The polar ends of the charged

phospholipids and other bifunctional lipids are solvated bythe water tnolecules There are also considerable amounts

of charged proteins and mucopolysaccharides present on the

surface In contrast, the interior of the membrane is

popu-lated by the hydrophobic aliphatic chains from the fatly acid

esters

Coefficient

Withthis representation in mind apartial explanation can be

presetited as to why the n-octanol/water partitioning system

seems to mimic the lipid membranes/water systems found

in the body It turns out that n-octanol is not as nonpolar asinitially might be predicted Water-saturated octanol con-tains 2.3 M water because the small water molecule easily

clusters around octanol's hydroxy moiety ated water contains little of the organic phase because of the

n-Octanol—satur-large hydrophobic 8-carbon chain of octanol The water in

the n-octanol phase apparently approximates the polar

prop-PROTEIN ORMUCOPOLYSACCHARIDE-

PROTEIN LAYERS

\

20 WiLson (illS Gisiold's Texth,n,k ofOrganicMedicinaland l'lwr,nace,,:ical Cl,en,i.an'

erties of the lipid bilayer whereas the lack of uctanol in the

water phase mimics the physiological aqueous

conipart-ments which are relatively free of mmpolar components

In contrast, partitioning systems such as hexane/water and

chloroform/water contain so little water in the organic phase

that they are poor models for the lipid bilayer/water system

found in the body At the same time, remember that the

a-octanol/water system is only an approximation of the actual

environment found in the interface between the cellular

membranes and the extr.tcellular/intracellular fluids

The basic procedure for obtaining a partition coefficient

is to shake a weighed amount of chemical in a flask

contain-ing a measured amount of water-saturated octanol and

octa-nol-saturated water Many times, the aqueous phase will be

buffered with a phosphate at pH 7.4 to reflect

physio-logical pH l'his corrects for the ratio of conjugate ha.se)/

[acidi found in viva The amount of chemical in one or

both of the phases is delennined by an appropriate analytical

technique and the partition coefficient calculated from

Equa-tion 2-20 The octanol/water partiEqua-tion coefficient has been

determined for thousands of compounds, including drugs

agricultural chemicals, biochemical intermediates and

me-tabolites and common chemicals Many of these

determina-tions have been obtained in several other organic solventl

water systems such as ether chloroform triolein and

hex-ane Equations have been published relating the partition

coefficients determined in one solvent/water system to those

detenitined in another

chemical

I' = (h1 2-20)(chemical ç

The determination of partition coefficients is tedious and

time consuming Some chemicals arc too unstable and either

degrade during the procedure, which can take several hours

or cannot be obtained in sufficient purity for an accurate

determination This has led to attempts at approximating the

partition coefficient Perhaps the most popular approach has

been high-performance liquid chromatography (HPLC) or

thin.layer chromatography (TLC) In each case, the support

phase is nonpolar, either by pennanent bonding (usually

oc-tadecylsilane) or a coating of octanol mineral oils, or related

materials The mobile phase usually contains some

water-miscible organic solvent to hold enough of the chemical

whose partition coefficient is being determined in solution

Sometimes the partition coefficient is calculated from the

retention data by regression analysis using Equation 2-21

The a and c terms have the same uses as in Equalion 2-16

log P alog retention) + (Eq 2-21)

This model has at least two limitations First, to obtain

valid numerical values Ilir a and c in Equation 2-I (, partition

coefficients for a group of very closely related compounds

must be obtained initially by the classical shake flask

method The retention times for the sante group of

com-pounds are then obtained in the identical chrornatographie

system that will be used for the new compounds The values

for a and care obtained using Equation 2-21 using standard

linear regression The second limitation to the

chromato-graphic model is Ihat chromatochromato-graphic approximations of

the partition coefficients usually only work when one is

de-termining the retention times of chemicals of the same

chem-ical class and similar substitution patterns Because of these

limitations sometimes the medicinal chemist will use the

retention data directly in the prediction of biological sponse (Eq 2-221 A chemical's retention on a chromatu-

re-graphic support is the result of a combination of its

partition-ing steric and electronic properties Because these santephysical chemical properties are important variables in de-termining a drug's biological response, excellent correla-

tions have been obtained between chroniatographic retentionparameters and biological response While the model repre-

sented by Equation 2-22 is useful in predicting biological

response, it is not as definitive as the models presented belosv

(Eqs 2-23 to 2-25) because ihe precise physical chemical

properties are combined into one chromalographic retentionterm In other words, it is not possible to determine the rela-

tive importance of lipophilicity electronic effects, or stericinfluence on the biological response when using Eqtiation

2-22

log 8R = aclog retention) c Eq 2-221

Most recently, there has been a concentrated effort to

cal-culate the partition coefficient on the basis of the atomic

components of the molecule En-h Mont type is assumed to

contribute a lixed amount to the chemical's partition

coeffi-cient Because this assumption breaks down quickly several

correction (actors are used Cyclohexene will serve as an

example

lug F' = 6(carhon atoms) + I2(hydrogcn aloms)+ (ii — I (bonds 4- double bond correctionlog P 6)0.20) + 12(0.23) + S —009) ÷ (—0.55)

= 2.96

For purposes of comparison, the observed octanol/waterpartition coefficient (expressed as a logarithm is 2.86 Be-

cause of the correction factors, these calculations become

so complex that they must he done by a computer programthat analyzes the structure and identifies those structural at-

tributes requiring correction factors Convenient as the

cal-culation method may he its accuracy depends on first

deter-mining experimental partition coefficients of chemicalsexhibiting very similar chemistry The values for specific

atoms, groups of atoms and bond correction factors are

de-rived from these experimentally determined partition

coeffi-cients

There are several commercial drtig design software

pack-ages that contain modules that estintate a chemical's

parti-tion coefficient Sonic use the method described in the ous paragraph Others use quantum chemical parameters In

previ-all cases, the algorithm must he validated against test sets

of diverse chemical structures whose partition coefficients

have been determined by the classical shake flask method.There are simpler methods for estimating lipophilicity thatwill give reasonably correct results These are based on the

additive effect on the partition coefficient that is seen when

varying a series of substituents on the sante molecule Over

the years fairly extensive tables have been developed thatcontain the contribution fir) of a wide variety of substituents

to the partition coefficient The method can he illustrated

for chlorobenzene The log of P is 2.13 for bcnzene and 2.84

for chlorobenzcne The i,' value for the chlorine suhstituent

is obtained by subtracting the log of P values for benzene

and chlorohenzcne

log — log

Chapter 2 • Properiwain Re/win,,to I/join glen!fletin,, 21

While the irsubstituent method has its limitations,

partic-ularly when them are significant resonance and inductive

effects resulting from the presence of multiple substituents

it can work well for a series of compounds that have similar

substitution patterns

Other Physicochemical Parameters

There is a series of other constants that measure the

contribu-tion by substituents to themolecule's total physicochemicul

properties These include Hummett's Taft's

ste-nc parameter E,: Charton's steric parameter :'; Verloop's

multidimensional smeric parameters, L, B1 B5; and molar

refractivity, MR The latter has become the second most

useful physicochcmical parameter used in classical QSAR

modeling It is a complex term based on the molecule's

re-Inactive index, molecular weight, and density and can be

considered a measure of the molecule's bulk and electronic

character One reason for its popularity is that it is easy to

calculate from tables of atoms, using a minimum of

correc-lion factors Of the listed physicochemical parameters group

representative list can be found in Table 2-8

Table 2-8 illustrates several items that must be kept in

mind when selecting substituents to be evaluated in terms

of the type of factors that influence a biological response

For electronic parameters such as the location on an

am-matic ring is important because of resonance versus

induc-tive effects Notice the twofold difkrences seen between

and for the three aliphatic substituents and iodo,

and severalfold difference for methoxy amino fluoro, and

phenolic hydroxyl

Selection of substituents from a certain chemical class

may not really test the influence of a parameter on biological

activity There is little numerical difference among the

or P,,,,,,, values for the four aliphatic groups or the four

halo-TABLE 2—8 sampling of Physicochemical Parameters

Used In Quantitative Structure Activity Relationships

gens It isnotuncommon IC) go to the tables and find missing

parameters such as the L', values liir acetyl and N-acyl.Nevertheless, medicinal chemists can usc information

from extensive tables of physicocheniical parameters to

min-imize the number of substimuents required to find out if thebiological response is sensitive to electronic steric, and/orpartitioning effects.2 This is done by selecting substituents

in each of the numerical ranges for the different parameters

In Table 2-8 there are three ranges of B values (—1.23 to

—0.55 —0.28 to 0.56 and 0.71 to 1.55); three ranges of MR

values (0.92 to 2.85 5.02 to 8.88 and It).30 to 14.96): and

two main clusters of one for the aliphatic

substitu-ents and tile other for the halogens In the ideal situation

substituents arc selected from each of the clusters to

deter-mine the dependence of the biological response over the

largest possible variable space Depending (In the biological

responses obtained from testing the new compounds it ispossible to determine if lipophilicity (partitioning) stericbulk (molar refraction), or electron withdrawing/donating

properties are important determinants of the desired cal response

biologi-QSAR Models

Currently,there are three models or equations seen inQSARanalysis using physicochcniical parameters, represented by

Equation 2-23 2-24 and 2-25 These three equations are

illustrated in Figure 2-10 using tile logarithm of the partitioncoefficient (log P) as the physical chemical parameter First

there is the linear model (Eq 2-23) When plots of log If

BR or log BR indicated a nonlinear relationship between

biological response and the partition coefficient, a parabolic

model was tried (Eq 2-24) Examination of Figure 2-lUshows an optimum log P (log P.,) where maximum biologi-cal activity will he obtained before a decrease in activity isseen One explanation for this phenomenon is that hydro

philic drugs will tend to stay in the aqueous phase, whereas

lipophilic chemicals will prefer the lipid hilayer In both

cases, less drug is being transported to the receptor, resulting

in a decrease inthe actual concentration of receptor-bound

22 and

drug In other words, the equilibrium seen in Reaction 2-I

shifts to the left There will be a group of drugs whose log

I' places them ncar the top of the parabola: their

lipo-philic—hydrophilic balance will permit them to penetrate

both aqueous and lipid barriers and reach the reccptur

log I/BR = a kig I') + c tEq 2-23)

lug I/13R = aOog I') — (log ' t' (Eq 2-24)

tog I/BR = a (log P1 — b log(/3P + I I + c Eq 2-25)

The third QSAR equation in current usc is the bilinear

model (Eq 2-25) It consists of two straight lines, one

as-cending and one desas-cending The /3 term connects the two

lines There are several interpretations tbr the /3 term One

explanation is based on the ratio of the rate constant for

diffusion nut of the octanol layer into the aqueous

environ-ment being different front the rate of diffusion out of the

aqueous layer into the octanol layer In other words, what

may be simulated with the bilinear model is recognition that

the rate of diffusion from the cxtraeellular fluids into the

lipid bilayer differs from the rate of diffusion out of the lipid

bilayer into the intracellular environment Another

interpre-tation is recognition that the kinetics of partitioning through

the lipid bilayer differ from the kinetics of binding to the

receptor A third explanation takes into account the different

volumes of the aqueous and lipid bilaycrs in the biological

system

With this background in mind, three examples of QSAR

equations takett from the medicinal chemistry literature are

presented One shows a linear relationship (Eq 2-26), and

the others show parabolic (Eq 2-27) and bilinear (Eq

correlations A study of a group of griseofulvin analogues

showed a linear relationship (Eq 2-26) between the

biologi-cal response and both lipophilicity (log P) and electronic

character (ui.3 It was suggested that the antibiotic activity

maydependon the enone system fiieilitating the addition of

griscoflilvin to a nucleophilic group such as the SH moiety

in a fungal enzyme

log 13K = (0.Sôllog P + (2.19)tr, — (.32 Eq 2-2h)

A parabolic relationship (Eq 2-27) was reported for a

series of substituted acetylated salicylates (substituted

aspi-rins) tested for anti—inflammatory activity.' A nonlinear

rela-tionship exists between the biological response and

lipophil-icky, and a significant detrimental steric effect is seen with

substituenis at position 4 The two sterimol parameters used

in this equation were L defined as the length of the

substitu-en! along the axis of the bond between the first atom 01 the

substituent and the parent molecule, and B2 defined as a

width parameter Steric effects were not considered

statisti-cally significant at position 3 as shown by the sterimul

pa-rameters fur substituents at position 3 not being part of

Equa-tiOn 2-27 The optimal partition coefficient (log P,,) for the

increasing hulk, as meastired by the sterimol parameters

decreases activity

Aspirin: X V Hlog I/ED9, = l.t)3 log P — P)' (Eq

In addition to these QSAR models based on biological

responses QSAR is used to analyte pharmacokinetic ity One example of this(Eq 2-2S) is a simulation of barbitu-

activ-rate absorption which leads to the bilinear

0.949 log P — lost/il' + I) — 3.131 (Eq 2-2(t)

rate constant

At this point, it is appropriate to ask the question, are allthe determinations of partition coefficients and compilation

of physical chemical parameters useful only when a

statisti-cally valid QSAR model is obtained? l'he answer is a firm''no." One of the most useful spinoffs from the field of

QSARhas been the application of experimental design to

the selection of new compounds to be synthesized and tested

assume that a new series of drug molecules is to be

synthesized based on the fullowing structure The goal is totest the eftCct of the 16 substituents in Table 2-N at each 0)three positions on our new series The number of possibleanalogues is equal to 161 or 44)96 compounds assuming

that all three positions will always he substituted with one

of the substituetits from Table 2-7 If hydrogen is includedwhen a position is not substituted, there are 17" or 4913,different combinations The problem is to select a small

number of substituents that represent the different ranges oi

clusters of values for lipophilicity electronic influence anc

bulk An ittitial design set could include the methyl and

pro-pyl from the aliphatic cluster lluorine and chlorine from ththalogen cluster N-acetyl and phenol from the substitueni,showing hydrophilicity, and a range of electronic and bull

values Including hydrogen there will be 7" or 343 differcncombinations Obviously, that is too many for an initial evaluation Instead, certain rules have been devised to maximizl

the infurmation obtained from a minimum number of cornpounds These include the following:

I Each substiluent must occur more than once aL cacti positiol

on which it is found

2 The number at Limes that each substituent at a particular positiolappears should he approximately equal

3 No subscituents should he present in a constant combinaliot

4 When conthinations of substituents are a necessity, they shouti

aol occur more frequently than any oilier combination

R,

Chapter 2 • Propertiesin Relutin,, in Biolo'icul Action 23Following these guidelines, the initial test set can be re-

duced to 24 to 26 compounds Depending on the precision

of the biological tests, it will be possible to see if the data

will lit a QSAR model Even anapproximate model usually

will indicate the types of subscituenis to test further and what

positions on the molecules are sensitive to substitution and

if sensitive to what degree variation in lipophilic, electronic

or bulk character is important Just to ensure that the model

is valid, it is a good idea to synthesize a couple olcompounds

that the model predicts would be inactive As each group of

new compounds is tested, the QSAR model is refined until

the investigators have a pretty good idea what substitucnt

patterns are important tar the desired activity These same

techniques used to develop potent compounds with desired

activity also can be used to evaluate the influence of

substitu-ent patterns on undesired toxic effects and pharmacokinctic

properties

In their purelana. the rules listed above can be used to

select a minimum number of compounds for a test set, using

what are known as idenwv variables No physicochcmical

parameters are required In its simplest form, the equation

takes the form outlined in Equation 2-29 This approach has

been known as a Free-Wilson analysis.7

log BR = (substitucnt contributions)

+ contribution tram the ba.se molecule (Faq 2-29)

An example is a small set of phosphorus-containing

ace-tylcholinesterase inhibitors that were selected by using the

rules for designing a test set and evaluated as possible

insec-The result is a complex equation that produces a

coefficient for each substituent They are suniniariied in

Table 2-9

Examination of this table shows that ethyl and ethoxy at

R1 ethoxy and isopropoxy at R?, and oxo at R4 have minimal

influence on biological activity In contrast, methyl and

iso-propoxy at R1; methoxy, iso-propoxy and hutoxy at R2: all three

nitrophenoxy substituenis at and thio at R4 significantly

influence the biological response The predicted log I/BR

for the compound where R, = methyl R1 = propoxy R7

= 4-nitrophenoxy and R4 = thio would be calculated from

Equation 2-29:

log BR = R1 + R2 + R, + R4 + base niotecuk

= 5.353

One of the newer QSAR methods combines statistical

techniques with molecular modeling and has been referred

to as three-dimensional QSAR (3D-QSAR) because the

in-dependent variables, usually physicochernical parameters

take into account spatial distances among and between macophores and their location at specific distances from the

phar-molecule Each point has a location on x v and nates 3D-QSAR depends on the molecular modeling algo-rithms and is discussed in more detail in ChapterS Attempts

coordi-at including a variety of orientcoordi-ations in space and a variety

of biological responses have led to the use of terols such asfour.dwcenxu,nalandJive-dimensionalQSAR.'

Topological Descriptors

An alternate method of describing molecular structure isbased on graph theory in which the bonds connecting the

atoms is considered a path that is traversed from one atom

to another Consider Figure 2-Il containing t-phenylalanine

and its hydrogen-suppressed graph representation The

num-bering is arbitrary and not based on IUPAC or Chemical

,%h,.siraet.c nomenclature rules A connectivity table Table

2-10 is constructed

Table 2-1(1 is a two-dimensional connectivity table for

the hydrogen-suppressed phcnylalanine molecule No

three-dimensional representation is implied Further, this type of

connectivity table will be the same for molecules with metric atoms (a versus i.) or for those that can exist in more

asym-than one conformation (i.e., ''chair'' versus ''boat'' mation, antiversusgone/u' versus eclip.sedl

confor-Graph theory is not limited to the paths followed by

chem-ical bonds In its purest form, the atoms in the phenyl ring

of phenylalanine svould have paths connecting atom 7 with

atoms 9 10 II and 12: atom 8 with atoms 10 II and 12:atom 9 with atoms II and 12: and atom 10 with atom 12

Also, the graph itself might differentiate neither single ble, and triple bonds nor the type of atom (C 0, and N in the

dou-TABLE 2—9 Coefficients for Substituents In a Set of Acetyichollnesterase Inhibitors

(1.856

— 1.134 0.611

Osu Thur 0.052

1.673

In Such an SI cdt OSAR n l)enign of Molecule' Bat elonu, 1 5 Pious, 5K-I.

24 I*'slbook of Mcdieina! and Phar,nare,ujcal Chemistry

phenylalanine example) Connectivity tables can be coded

to indicate the type of bond

The most common application of graph theory used by

medicinal chemistry is called molecular comu.'c:ivitv.It

lim-its the paths to the molecule's actual chemical bonds, Table

2-Il shows several possible paths for phenylalanine

includ-ing linear paths and clusters or branchinclud-ing Numerical values

for each path or path-cluster arc based on the number of

nonhydrogen bonds to each atom Let's examine oxygen

atom I There is only one nonhydrogen bond, and it connects

oxygen atom I to carbon atom 2 The formula is the

recipro-cal square root of the number of bonds For oxygen I the

connectivity value is I For carbonyl oxygen 2 it is

or 0.707 Note that there is no dilference between oxygen

I and nitrogen 5 f3oth have only one nonhydrogen bond and

a connectivity value of I Similarly, there is no difference in

values for a carbonyl oxygen and a methylene carbon, each

having two nonhydrogen bonds The linai connectivity

val-ues br a path are the reciprocal square roots of the products

of each path For the second-order path 2C-4C-bC the

recip-rocal square root (3 x 3 x 21u/ is 4.243 The values for

each path order are calculated and summed

As noted above, the method as described so far cannot

distinguish between atoms that have the same number of

nunhydrogen bonds A method to distinguish heteroatoins

from each other and carbon atoms is based on the difference

between the number of valence electrons and possible

hydro-gen atoms twhich are suppressed in the graph) The

"va-Figure 2—11 • Hydrogen-suppressed graphicrepresentation of phenylalanine

lence" connectivity term for an alcohol oxygen would be 6valence electrons minus I hydrogen or 5 The "valence"connectivity term for a primary amine nitrogen would be 5

valence electrons minus 2 hydrogens or 3 There are a ety of additional modifications that are done to further differ-entiate atoms and define their environments svithin the mole-cule

vari-Excellent regression equations using topological indiceshave been obtained A problem is interpreting what theymean Is it lipophilicity steric hulk, or electronic terms thatdefine activity? The topological indice.s can be correlatedwith all of these common physicochemical descriptors An-other problem is that it is difficult to use the equation todecide what molecular modifications can be made to en-

hance activity further, again because of ambiguities in physi.cochemical interpretation Should the medicinal chemist in-

crease or decrease lipophilicity at a particular location onthe molecule? Should specific substituents be increased ordecreased? On the other hand, topological indices can bevery valuable in classification schemes that are describedbelow, They do describe the structure in terms of rings.branching, flexibility etc

Classification Methods

Besides regression analysis, there are other statistical niques used in drug design These fit under the classification

tech-of multivariate statistics and include discriminant analysis

TABLE 2-10 ConnectivIty Table for Hydrogen-Suppressed Phenylalanine

c-Il

I

x

X

Chapter 2 • Pfrxüoi/wmiial Properrir.s in Re/aiim, ,a item,,, 25

TABLE 2-11 Examples ofPathsFound in thePhenylalanine Molecule

tIC-I IC 7C-12C-t IC 7C-SC.SC-IOC 7C-SC.SIC-IOC-I IC bC-7C-HC-SC-IOC-I IC

IIC-12C 8C-9C-IOC

SC-IIJC-IIC

IOC-IIC-12C

7C-12C-IIC-IOC8C-9C-It)C-IICSC-IOC.IIC.12C

7C-12C-IIC-IOC-9C8C-9C-IOC-IIC.12C

6C-7C-12C-IIC.IOC7C-MC-SC-IOC-IIC-12('7C-1!C-JIC.II)C-9C-SC

principal component analysis and pattern recognition The

latter can consist of a mixture of statistical and nonstatistical

methodologies The goal usually is to try to ascertain what

physicocheinical parameters and structural attributes

con-tribute to a class or type of biological activity Then the

chemicals are classified into groupings such as CarCinogcnic/

noncarcinogenic sweet/bitter, active/inactive, and

depres-sanllstimulant

Theterm ,nulgi,'ariugeis used because of the wide variety

andnumberofindependentordescriptorvariables that may

be used The same physicochemical parameters seen in

QSAR analyses are used, but in addition, the software in

the computer programs 'breaks" the molecule down into

substructures These structural fragments also become

vari-ables Esamples of the typical substructures used include

carbonylt enones conjugation, rings of different sizes and

types N-substitution patterns, and aliphatic substitution

pat-lerns such as 1.3- or 1,2-disubstituted The end result is that

for even a moderate-size molecule typical of most drugs.

there can he 50 to 100 variables

The technique is to develop a large set of chemicals well

characterized in terms of the biological activity that is going

ma be predicted This is known as the training set Ideally

it should contain hundreds, if not thousand.s, of compounds

divided into active and inactive types In reality, sets smaller

than 1(10 are studied Most of these investigations are

retro-spective ones in which the investigator locates large data

sets from sevenil sources This means that the biological

te.sting likely followed different protocols That is why

clas-sification techniques tend to avoid using continuous

vari-ables such as ED51,.LD50.and MIC Instead, arbitrary

end-points such as active or inactive, stimulant or depressant.

sweet or sour are used

Once the training set is established, the multivuriate

tech-nique is curried out The algorithms are designed to group

the underlying commonalitics and select the variables that

have the greatest influence on biological activity The

predic-tive ability is then tested with a test set of compounds that

have been put through the same biological tests used for the

training set For the classification model to be valid, the

investigator must select data sets whose results are not tively obvious and couldnot beclassified by a trained medic-inal chemist Properly done, classification methods can iden-

intui-tify structural and physicochemical descriptors that can be powerful predictors and determinants of biological activity There are several examples of successful applications of this technique.'' One study consisted of a diverse group of 14() tranquilizers and 79 sedatives subjected to a two-way

classification study (tranquilizers versus sedatives) The ring

types included phenothiazines indoles benzodiazcpines barbiturates diphenylmethanes and a variety of hemerocy- dies Sixty-nine descriptors were used initially to character- ize the molecules Eleven of these descriptors were crucial

to the classification, 54 had interntediatc use and depended

on the composition of the training set, and 4 were of little

use The overall range of prediction accuracy was 88 to 92%

The results with the 54 descriptors indicate an important limitation when large numbers of descriptors are used The inclusion or exclusion of descriptors and parameters can de- pend on the composition of the training set The training set must be representative of the population of chemicals that are going to be evaluated Indeed, repeating the study on different randomly selected training sets is important Classification techniques lend themselves to studies lack- ing quantitative data An interesting classification problem

involved olfactory stimulants in which the goal was to seleci

chemicals that had a musk odor A group of 300 unique compounds was selected from a group of odorants that in- cluded 60 musk odorants plus 49 camphor 44 floral 32 ethereal 41 mint SI pungent, and 23 putrid odorunts Ini- tially 68 descriptors were evaluated l)epending on the ap- proach, the number of descriptors was reduced to II to 16 consisting mostly of bond types Using this small number the 60 musk odorants could be selected from the remaining

240 compounds with an accuracy of 95 to 97%.

The use of classification techniques in medicinal try has matured over years of general use The types of de- scriptors have expanded to spatial measurements in three-

dimensional space similar to those used in 3D-QSAR (see

below) Increasingly, databases of existing compounds are

scanned for molecules that possess what appear to be the

desired parameters If the scan is successful, compounds that

arc predicted to be active provide the starting point for

syn-thesizing new compounds for testing One can see parallels

between the search of chemical databases and screening

plant, animal, and microbial sources for new compounds

Although the statistical and pattern recognition

methodolo-gies have been in use for a very long time, there still needs

to be considerable research into their proper use, and further

testing of their predictive power is needed The goal of

scan-ning databases of already-synthesized compounds to select

compounds for pharmacological evaluation will require

con-siderable additional development of the various multivariate

techniques

COMBINATORIAL CHEMISTRY

Elegant as the statistical techniques described above are, the

goal remains to synthesize large numbers of compounds so

promising marketable products are not missed At the same

time, traditional synthetic and biological testing arc very

costly This has led to the technique called co,nbinatorial

chenu.ctrv The latter uses libraries of chemical moieties that

react with a parent or base molecule in a small number of

defined synthetic steps Return to the two examples

pre-sented with the discussion of the Free-Wilson analysis

above As the number of different substituents is considered,

literally more than 10.000 compounds are possible

Remem-ber that the medicinal chemist can select subsets of

substitu-ems that vary in lipophilicity steric bulk, induction, and

resonance effects and use the four rules for placing and use

of the substituents If this process is properly done, a

rela-tively small number of compounds will be obtained that

show the dual importance of each of the physicochemical

parameters being evaluated at each position on the molecule

and the effect of specific moieties at each position This

"rational" approach to drug design assumes that there is

some understanding of the target receptor and that there is

a lead molecule, commonly called the prototype molecule

A classic example is the dihydrofolate reductase inhibitor

methotrexate which has been one of the prototypes that

lab-oratories have used to synthesize and test new inhibitors

Another example is benzodia2epine, which has a defined

structure whose activity varies with the substituents

What about the situation in which little is known about

the mechanisms causing the disease process? Until recently,

this has been the normal situation when searching for new

molecules with the desired pharmacological response With

the discovery of penicillin came the realization that

micro-bial organisms produced "antibiotics." This started

screen-ings of microbial products looking for new antibiotics In

a similar manner, thousands of synthetic compounds and

plant extracts have been screened for anticancer activity

Some have called this "irrational" drug design, but it has

produced most of the drugs currently prescribed This

ap-proach also is very expensive, particularly when one realizes

the cost to synthesize, isolate, and test each new compound

pound through efficacy and safety testing before its release

to the general public is approved by government regulatoryagencies

Combinatorial chemistry is one method of reducing the

cost of drug discovery in which the goal is to find new leads

or prototype compounds or to optimize and refine the ture—activity relationships)2'5 Libraries of ''reactive"

struc-chemical moieties provide the struc-chemical diversity olproducts

that will be screened for activity The chemistry is elegantbut relatively simple in that the same few reactions are re-quired to malte thousands of compounds in a particular se-

ries The reactions most be clean and reproducible and have

high yields Often, solid-state synthetic methods are used

in which compounds are "grown" onto polymer support

Robotics can be used to reduce further the cost of synthesis.Biological testing can also be automated in a process calledlzigh'zhroughpur screening, which can test tens to hundreds

of structures at a time Many times it is possible to take

advantage of gene cloning techniques, clone the desired

re-ceptor, and measure the binding of the newly synthesized

compounds to the cloned receptor

To maintain some locus on the needed structures muation theory has been used to construct the libraries ofsubstituents These libraries tend to maximize chemical di-versity in terms of physicochemical parameters Many ofthese libraries are sold commercially by firms specializing

infor-in this technique The synthetic methodologies cover the

spectrum from producing thousands of relatively pure pounds to producing mixtures of compounds that are tested

com-as mixtures Of course, mixtures can be difficult to clcom-assify,

and it can be difficult to determine which products in the

mixture are active and which are inactive Elegant methodshave been developed that chemically "tag" each compound

with a small peptide nucleotide or other small moleculethat is pharmacologically inert When mixtures of products

are obtained, they are screened for activity Only those

lures that are biologically active are retained In a process

called deconvolution the synthesis is repeated in an iterativemanner, producing smaller and sometimes overlapping mix-tures The screening is repeated until the active compounds

are identified Examine Table 2-12 This simplified outlineshows how four steps will identify the three active compo-nents in a 20-compound investigation (Keep in mind thatthe actual combinatorial process will produce hundreds orthousands of compounds for testing.)

Assume that the project calls for synthesizing 20 comrn

pounds, A to T Rather than carry out 20 distinct synthesesfollowed by 20 separate screening experiments, all of whichcan take weeks, four combinatorial syntheses are carried outsuch that four mixtures containing live compounds each are

obtained Only the three mixtures that test positive in the

screening assay are retained The synthesis is repeated ducing live mixtures of three components each, and the test-ing is repeated Six more syntheses are carried out this time.producing overlapping two-component mixtures, and the as-says are repeated It is now possible to determine that com-pounds B H and N are active Instead of 21) syntheses and

pro-20 assays, only 15 were required Further, time-consumingpurification of each mixture was not required This process

is very sinmilar to that carried out by natural-product chem'ists The microbial, plant, or animal tissue is extracted with

Chapter 2 • Phvsiroche,uiea! Prapei-tiev in Relation to &oluA'scalAcria,, 27

A! Sal C D E F 0 I Ha! I J K L Ml NOJ 0 I 01 R S

Carry out the synthesis producing tour fIve-component mixtures

Screen the mixtures

Retain only the three mixtures containing active components

Repeat the synthesis producing three-component mixtures and repeat the screening

Discard the Inactive mixtures

Repeat the synthesis producing overlapping two-component products and repeat the screening

ABa

Only compounds B H, and N need to be chemically characterized

and ending with an alcohol or waler, and the fractions are

screened for activity Only the active fractions are retained

The latter are more carefully fractionated, using biological

assays to follow the puritication In either combinatorial

syn-thesis or natural product isolation, once active compounds

are identified, larger-scale more focused syntheses can be

done, using QSAR-derived experimental design and/or

mo-lecular modeling (see below) to yield compounds different

from those produced from the combinatorial library of

chem-ical fragments

Other methods that are used commonly in combinatorial

chemistry include attaching structures of known

composi-tion to polystyrene beads (one compound per bead) or

syn-thesizing structures onto a microchip-sized matrix where a

compound's location gives its identity The latter is called

spatially addressable synthesis This topic is covered in

more detail in Chapter 3

MOLECULAR MODELING

(COMPUTER-AIDED DRUG DESIGN)

The low cost of powerful desktop computers gives the

me-dicinal chemist the ability to "design" the molecule on the

basis of an estimated lit onto a receptor or have similar

spa-tial characteristics found in the prototypical lead compound

Of course, this assumes that the molecular structure of the

receptor is known in enough detail for a reasonable

estima-tion of its three-dimensional shape When a good

under-standing of the geometry of the active site is known,

data-bases containing the three-dimensional coordinates of the

chemicals in the database can be searched rapidly by

com-puter programs that select candidates likely to fit in the active

site As showis below, there have been some dramatic

suc-cesses with use of this approach but first one must have

an understanding of ligand (drug)—receptor interactions and

conformational analysis

Drug-Receptor Interactions

At this point, let us assume that the drug has entered the

riers and is now going to make contact with the receptor

As illustrated in Reaction 2-I this is an equilibrium process

A good ability to lit the receptor favors binding and the

desired pharmacological response In contrast, a poor tit vors the reverse reaction With only a small amount of drugbound It) the receptor, there will he a much smaller pharma-

fa-cological effect Indeed, if the amount of drug hound to the

receptor is too small, there may be no discernible response.Many variables contribute to a drug's binding to the receptor

These include the structural class, the three-dimensionalshape of the molecule, and the types of chemical bondinginvolved in the binding of the drug to the receptor

Most drugs that belong to the same pharmacological classhave certain structural features in common The barbiturates

act on specific CNS receptors causing depressant effects:

hydantoins act on CNS receptors, producing an sant response: benzodiazepines combine with the y-aunino-

anticonvul-butyric acid (GABA) receptors, with resulting anxiolytic

activity: steroids can be divided into such classes as

cortico-steroids, anabolic steroids progestogens and estrogens.each acting on specific receptors: nonsteroidal anti-inflam-

matory agents inhibit enzymes required for the prostaglandincascade: penicillins and cephalosporins inhibit enzymes re-

quired to construct the bacterial cell svall: and tetracyclines

act on bacterial ribosomes

Receptor

Withthe isolation and characterization of receptors

becom-ing a common occurrence, it is hard to realize that the

con-cept of recon-ceptors began as a postulate It had been realized

early that molecules with certain structural features would

elucidate a specific biological response Very slight changes

in structure could cause significant changes in biologicalactivity These structural variations could increase or de-

crease activity or change an agonist into an antagonist This

early and fundamentally correct interpretation called fur the

drug (ligand) to fit onto some surface (the receptor) that had

luirly strict structural requirements for proper binding of thedrug The initial receptor model was bused on a rigid lock-and-key concept, with the drug (key) fitting into a receptor(lock) It has been used to explain why certain structural

28 Wilxo,i cii:d leoh,u;kof Medicinaland Phur,naceuiieal C'hemis,rv

attributes produce a predictable pharmacological action

This model still is useful, although one musi realize that both

the drug and the receptor can have considerable flexibility

Molecular graphics using programs that calculate the

prc-terred conformations of drug and receptor, show that the

receptor can undergo an adjustment in three-dimensional

structure when the drug makes contact Using space-age

lan-guage the drug "docks" with the receptor

More complex receptors now are being isolated,

charac-terized and cloned The first receptors to be isolated and

characterized were the reactive and regulatory sites on

en-zymes Acetylcholinesterasc dihydrofolate reductase

an-giotensin and I-IIV protease.converting enzyme are

exam-ples or enzymes whose active sites (the receptors) have been

modeled Most drug receptors probably are receptors for

natural ligands used to regulate cellular biochemistry and

function and to communicate between cells Receptors

in-clude a relatively small region of a macromolecule which

may be an isofatable enzyme a structural and functional

component of a cell membrane, or a specific intracellular

substance such u.s a protein or nucleic acid Specific regions

of these macromolecules are visualized us being oriented in

space in a manner that permits their functional groups to

interact with the complementary functional groups of the

drug This interaction initiates changes in structure and

func-tion of the niacromolccule which lead ultimately to the

ob-servable biological response The concept of spatially

ori-ented functional areas forming a receptor leads directly to

specific structural requirements for functional groups of a

drug, which must complement the receptor

It now is possible to isolate membrane-bound receptors

although it still is difficult to elucidate their structural

chem-istry because once separated from the cell membranes, these

receptors may lose their native shape This is because the

membrane is required to hold the receptor in ifs correct

ter-tiary structure One method of receptor isolation is affinity

chromatography In this technique, a ligand often an altered

drug molecule known to combine with the receptor is

at-tached to a chrornatographic support phase A solution

con-taining the desired receptor is passed over this column The

receptor will combine with the ligand It is common to add

a chemically reactive grouping to the drug, resulting in the

receptor and drug covalently binding with each other The

drug—receptor complex is washed from the column and then

characterized further

A more recent technique uses recombinant l)NA The

gene for the receptor is located and cloned It is transferred

into a bacterium, yeast or animal, which then produces the

receptor in large enough quantities to permit further study

Sometimes it is possible to determine the DNA sequence of

the cloned gene By using the genetic code for amino acids

the amino acid sequence of the protein component of the

receptor can be determined, and the receptor then modeled

producing an estimated three-dimensional shape The model

for the receptor becomes the template for designing new

ligands Genome mapping has greatly increased the

infomia-lion on receptors Besides the human genome the genetic

composition of viruses, bacteria, fungi, and parasites has

increased the possible sites for drugs to act The new field

of proteolnics studies the proteins produced by structural

genes

cell membrane is a highly organized, dynamic structure that

interacts with small molecules in specific ways; its focus is

on the lipid bilayer component of this complex structure.The receptor components of the membranes appear to bemainly protein They constitute a highly organized region

of the cell membrane The same type of molecular speciliciy

seen in such proteins as enzymes and antibodies is also aproperty of drug receptors The nature of the amide link inproteins provides a unique opportunity for the formation of

multiple internal hydrogen bonds, as well as internal

forma-lion of hydrophobic van der Waals' and ionic bonds byside chain groups leading to such organized structures asthe a helix, which contains about four amino acid residuesfor each turn of the helix An organized protein structurewould hold the amino acid side chains at relatively fixed

positions in space and available for specific interactions with

a small molecule

Proteins can potentially adopt many different tions in space svithout breaking their covalent amide link-ages They may shift from highly coiled structures to par-tially disorganized structures, with pans of the moleculeexisting in "random chain" or "folded sheet" structures.contingent on the environment, In the monolayer of a cellmembrane, the interaction of a small foreign molecule with

conforma-an orgconforma-anized protein may lead to a significconforma-ant chconforma-ange in the

structural and physical properties of the membrane Suchchanges could well he the initiating events in the tissue ororgan response to a drug, such as the ion-translocating ci-

lects produced by interaction 01 acetylcholune and the ergic receptor

cholin-The large body of information now available on ships between chemical structure and biological activitystrongly supports the concept of flexible receptors The fit

relation-of drugs onto or into macromolecules is rarely an

all-or-none process as pictured by the earlier lock-and-key concept

of a receptor Rather, the binding or partial insertion of

groups of moderate size onto or into a macromolecular pouchappears to be a continuous process, at least over a limited

range, as indicated by the frequently occurring regular

inbiological activity as one ascends ahomologous series of drugs A range of productive associa-

tions between drug and receptor may be pictured which

leads to agonist responses, such as those produced by

choliri-ergic and adrencholiri-ergic drugs Similarly, strong associationsmay lead to unproductive changes in the configuration ofthe macromolecule, leading to an antagonistic or blocking

response, such as that produced by anticholinergic agents

and HIV protease inhibitors Although the fundamentalstructural unit of the drug receptor is generally considered

to be protein, it may be supplemented by its associations

with other units, such as mucopolysaecharides and nucleicacids

Humans (and mammals in general) very complex ganisms that have developed specialized organ systems It

or-is not surpror-ising that receptors are not dor-istributed equallythroughout the body It now is realized that, depending on

the organ in which it is located, the same receptor class may

behave differently This can he advantageous by focusing

drug therapy on a specific organ system but it can also cause

adverse drug responses because the drug is exerting two

different responses based on the locution of the receptors