Citation
Prodrug Strategies Aimed at Improving Topical Delivery of Drugs Using the N-Alkyl-N-Alkyloxycarbonylaminomethyl (NANAOCAM) Promoiety

Material Information

Title:
Prodrug Strategies Aimed at Improving Topical Delivery of Drugs Using the N-Alkyl-N-Alkyloxycarbonylaminomethyl (NANAOCAM) Promoiety
Creator:
MAJUMDAR, SUSRUTA
Copyright Date:
2008

Subjects

Subjects / Keywords:
Chlorides ( jstor )
Esters ( jstor )
Hexanes ( jstor )
Hydrolysis ( jstor )
Lipids ( jstor )
pH ( jstor )
Phenols ( jstor )
Prodrugs ( jstor )
Skin ( jstor )
Solubility ( jstor )
City of Gainesville ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Susruta Majumdar. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
8/31/2006

Downloads

This item has the following downloads:


Full Text












PRODRUG STRATEGIES AIMED AT IMPROVING TOPICAL DELIVERY OF
DRUGS USING THE N-ALKYL-N-ALKYLOXYCARBONYLAMINOMETHYL
(NANAOCAM) PROMOIETY















By

SUSRUTA MAJUMDAR


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2006

































Copyright 2006

by

Susruta Majumdar




























This document is dedicated to my wife, PUJA and my father, BABA















ACKNOWLEDGMENTS

I would like to thank my advisor, Dr. Ken Sloan, wholeheartedly for allowing me

to pursue his and my ideas independently in his lab. His constant support, patience,

ability to keep me focused, thorough understanding of literature, and stress on basic

principles proved vital in my finishing this manuscript. The rigorous training imparted to

me in and outside the lab was responsible in my doing well in the lab, for the publications

that have or will come out of this dissertation and talking to people in various

conferences and interviews I attended during my stay at Florida. I greatly appreciate his

ability to communicate with me on a daily basis the answers to my questions in an

articulate manner. I feel confident enough to take myself to the next level in search of

something new again.

There are two people who inspire me the most: one of them is my father whom I

call baba and second is my wife, Puja. I would like to dedicate 4 years of my work to my

father and my wife.

My father and my advisor have two things in command. They both believe in

keeping things simple and have great faith in basic principles and honestly I thought I had

those under control till I joined the Sloan lab. My father introduced me to science at an

early age. He along with my mother made it possible for me to have the best education. I

went to the best private schools back home just because they valued education a lot in

spite of facing financial hardships sometimes. I did not pick my college major, chemistry;









he did. He said, "Just trust me on this one" and I am so glad I did. My dad being a

pharmacist influenced me to do what he did. I got my chance at Florida and I loved it.

I dated my present wife for 6 years before we got married. She keeps me going

everyday day after day with those constant words of encouragement. If my father was

responsible for me being in science, my wife ensured that his dream of my being a

scientist gets fulfilled. I admire the strength in her character to stay from me to finish her

education.

I would like to thank my mother and sister for their love and affection. My mother

was responsible for me being independent, tolerant and disciplined. I admire my sister's

ability to fight 'time' against all odds.

I would like to thank my graduate colleagues Dr. Scott Wasdo, Dr. Joshua Thomas

and Maren Muellar Spaeth. Scott actually was my second advisor; sometimes his

explanations for a physical chemistry concept were just too easy to follow. I hope he

decides to teach one day. I do not have words to describe Josh. He is too nice, too good

and focused. I would like to thank him for the critical discussions we had in the lab on

science, religion, politics and everything on earth. I admire him as a wonderful scientist; I

do not know of too many people who will run an experiment for 36 hrs non stop week

after week. Such dedication to work amazes me. Maren was an undergraduate exchange

student who worked with me. Together we finished couple projects in the limited time

she had in her disposal. I appreciate her adding insight and thoughtfulness into our work.

I would like to thank my committee members Dr. M.O.James, Dr. Raymond G.

Booth and Dr. William Dolbier. I would like to thank Dr. James for her constant

encouragement and her writing good recommendation letters for me. Dr. Dolbier was









largely responsible for a project I did with Maren; I thank him for adding insight into the

project and for being a part of my committee.

Finally I would like to thank Dr. Geeti Bansal (MIT), Dr. Sushma Chauhan

(Ranbaxy) and Dr. Gagan Kukreja (Ranbaxy) who taught me basically how to operate in

an organic chemistry lab like running TLC's, column chromatography and doing

crystallizations during my stay as an undergrad student at the University of Delhi.
















TABLE OF CONTENTS

page

ACKNOW LEDGM ENTS ........................................ iv

LIST OF TABLES ................. ........ ..... ......................x

LIST OF FIGURES ........... ........... .............. .......... xii

ABSTRAC T ............. ............ ................. ........ ........ xv

CHAPTER

1 INTRODUCTION ................... ................ ... ............. ...... ....... .........

The Skin........................................................ 1
Topical Delivery .............. .... ...... ..... ......... ...... ... ......... 8
Approaches to Increase Permeation across the Skin ..................................................11
Theory of Percutaneous Diffusion................... ...................... 13
Mathematical Modeling of Flux through Human Skin. Derivation of Roberts-
Sloan equation from Fick's law of Diffusion................. ...............17
Prodrugs ................... ............. .. ....... ......... ......... 21
Prodrugs for Dermal and Transdermal delivery ................................. ...........33
Research Objectives................ .......... ................. 41

2 DESIGN, SYNTHESIS, HYDROLYSIS OF N-ALKYL-N-
ALKYLOXYCARBONYL-AMINOMETHYL DRUG DERIVATIVES AND
ITS IMPLICATIONS ON PRODRUG DESIGN .................................................43

Introduction ................... ................. ............ .. ...................... 43
Synthesis of N-methyl-N-methyloxycarbonylaminomethyl and N-aryl-N-
methyloxy-carbonylaminomethyl Prodrugs........................45
N-Methyl-N-methyloxycarbonylaminomethyl chloride synthesis............48
N-Aryl-N-methyloxycarbonylaminomethyl chloride synthesis ...................49
Alkylation of Phenols, 6-Mercaptopurine, Dimethylaminobenzoic acid and
Naproxen with N-methyl-N-methyloxycarbonylaminomethyl chloride or
N-aryl-N-methyloxycarbonylaminomethyl chloride ....................................51
H ydrolysis Studies. ..................................... .. ... ..... ..............59
Implications of NANAOCAM-phenol Hydrolysis on Prodrug Design.......67
Implications of NArNAOCAM-carboxylic acid Hydrolysis on Prodrug
D e sig n .................. .......... ........................................... 7 0









3 SYNTHESIS AND TOPICAL DELIVERY OF N-ALKYL-N-
ALKYLOXYCARBONYLAMINOMETHYL PRODRUGS OF A MODEL
PHENOLIC DRUG:ACETAMINOPHEN................................74

Experim ental Procedure.................................................. 75
M materials and M ethods ......................................................75
Synthesis of Prodrug Derivatives................................................................75
Synthesis of NANAOCAM-Cl..........................................76
Determination of Solubilities and Partition Coefficients ..........................82
Determination of Flux through Hairless Mice Skins ...................................84
Determination of Prodrug Hydrolysis by UV Spectroscopy ........................87
Calculation of Maximum Flux .......... .................. ................ 87
Physicochemical Properties of NANAOCAM Prodrugs of APAP ............................88
Solubilities............................ ................88
D iffusion C ell Experim ents................................................................... 91
Prodrug Bioconversion to Parent Drug .............................. .................92
Permeability Coefficients and Solubility Parameter Values........................92
Residual Amounts in Skin.....................................93
Second A application Fluxes............................... .. ........................... 94
Modelling the Flux of NANAOCAM prodrugs of APAP through Hairless Mouse
Skin from IPM using the R S equation. ........................................ ............... 95
Conclusions............................. .......... ......... 97

4 SYNTHESIS AND TOPICAL DELIVERY OF N-ALKYL-N-
ALKYLOXYCARBONYLAMINOMETHYL PRODRUGS OF AN IMIDE
CONTAINING DRUG: THEOPHYLLINE........................................... 99

Experim ental Procedure.................................................. 101
M materials and M ethods ..................................... ........ ............ 101
Synthesis of Prodrug D erivatives................ ......................... .................102
Synthesis of NANAOCAM-Cl..............................102
Determination of Solubilities and Partition Coefficients ......................... 108
Determination of Flux through Hairless Mice Skins ............................110
Determination of Prodrug Hydrolysis by UV Spectroscopy .................113
Calculation of Maximum Flux ........... .......... ........................13
Physicochemical Properties of NANAOCAM Prodrugs of Theophylline............1...14
Melting Point Behaviour of NANAOCAM Prodrugs of Theophylline .....114
S olu b ilities ........................................... 1 14
Diffusion Cell Experiments......... ....................................116
Prodrug Bioconversion to Parent Drug. .................. .. .....................118
Permeability Coefficients and Solubility Parameter Values......................120
R esidual A m ounts in Skin .................................................................... 121
Second Application Fluxes..................................................... ................121
Modelling the Flux of NANAOCAM Prodrugs of Acetaminophen and
Theophylline through Hairless Mouse Skin from IPM using the RS equation. ...122
Conclusions........................................ ........ 125









5 SUMMARY OF RESULTS OBTAINED AND FUTURE WORK ......................126

L IST O F R E F E R E N C E S .............. ..... ............ ......................................................... 129

B IO G R A PH IC A L SK E T C H .............. .... ............. ................................................... 143
















LIST OF TABLES


Table page

1-1 x, y, z, r2 and Average Residual Errors for Various Databases Fit to RS. ...............21

1-2 Physicochemical Characterization of 1-Alkyloxycarbonyl esters of 5-FU..............35

1-3 Physicochemical Characterization of Alkyloxycarbonyl esters of
Acetaminophen. .................. ........ ...................38

1-4 Physicochemical Characterization of PEG esters of Indomethacin....................40

2-1 N-methyl-N-methyloxycarbonylaminomethyl-phenol Conjugates Synthesized .....46

2-2 N-methyl-N-methyloxycarbonylaminomethyl and NArNAOCAM Conjugates of
Naproxen. ........................................................46

2-3 Correlation of Rates of Hydrolysis of NANAOCAM-Y with pKa and Sigma
Values of the Leaving Group (Y). ........................................................... 61

2-4 Effect of pH of Buffer on Rates of Hydrolysis of 7.......... ......................61

2-5 N-aryl-N-alkyloxycarbonylaminomethyl Derivatives of p-nitrophenol (PNP). ......70

2-6 Hydrolysis of N-aryl-N-alkyloxycarbonylaminomethyl Derivatives of Naproxen. 72

2-7 Effect of pH of Buffer on Rates of Hydrolysis of 14............. ..... ...........72

3-1 NANAOCAM Prodrugs of APAP. ...................................... ....................80

3-2 Molecular Weights, Melting Points, Log Solubilities in Isopropyl Myristate ,
Log Solubilities in Water and Log Solubilities in pH 4.0 Buffer. ........................90

3-3 Molar Absortivities in Acetonitrile and Buffer, Log Solubility Ratios between
IPM and Water, the Differences Between Log SRIPM:AQ, the Log of Partition
coefficients Between IPM and pH 4.0 Buffer, and the Differences Between Log
KIPM:4.0. ......................................... ............................. 90

3-4 Solubilities in IPM, Solubilities in Water and Flux through in vitro Hairless
M house Skins from IPM ............................................... ............... 92









3-5 Log Permeability Values for the APAP from IPM through Hairless Mouse Skins
and Solubility Parameter values. ............................... .... .......................... 93

3-6 Residual Skin Concentrations of Total APAP and Ratios of Dermal versus
Transdermal Fluxes ........ ........ ......... ..................94

3-7 Second Application Theophylline Flux data for Flux of Theophylline from
Propylene Glycol ................ ............ .... ...... .. ...................... 94

3-8 Experimental Flux, Calculated Flux and Error in Predicting Flux for Compounds
1, 17-26 through Hairless Mouse Skins from IPM. .....................................96

4-1 NANAOCAM Prodrugs of Theophylline. ....................... ............... 107

4-2 Melting Point Comparisons of NANAOCAM-Th with AOC-Th and ACOM-Th. 115

4-3 Molecular Weights, Melting Points, Log Solubilities in Isopropyl Myristate, Log
Solubilities in Water and Estimated Log Solubilities in pH 4.0 Buffer...............117

4-4 Molar Absortivities in Acetonitrile and Buffer, Log Solubility Ratios between
IPM and Water, the Differences Between Log SRIPM:AQ, the Log of Partition
Coefficients Between IPM and pH 4.0 Buffer, and the Differences Between Log
KIPM:4.0. ........................................ ............. ................ 118

4-5 Solubilities in IPM, Solubilities in Water and Flux through in vitro Hairless
Mouse Skins from IPM. ................ ....... ......... ..... .........119

4-6 Log Permeability values for Theophylline Prodrugs from IPM through Hairless
Mouse Skins and Solubility Parameter Values. ..................................... 120

4-7 Residual Skin Concentrations of Total Theophylline Species ...............................121

4-8 Second Application Theophylline Flux data for Flux of Theophylline from
Propylene Glycol ............... ............ ............ .................. 22

4-9 Experimental Flux, Calculated Flux and Error in Predicting Flux for Compounds
1, 17-26 through Hairless Mouse Skins from IPM ......................... ...............124
















LIST OF FIGURES


Figure page

1-1 Cross sectional view of hum an skin ........................................................ 2

1-2 Brick and mortar model of drug absorption through the skin..............................6

1-3 Structures of ceramides seen in the lipid bilayers of the lamellar bodies. ..............7

1-4 Illustration of a typical flux profile. .......................... ............14

1-5 Two compartment diffusion model. ............................... ............... 15

1-6 Esterase m ediated hydrolysis of prodrugs.............................................................23

1-7 Soft alkyl prodrug hydrolysis. ........................................ ................. 24

1-8 Hydrolysis pathways for Saccharin based O-imidomethyl derivative of estradiol..26

1-9 Hydrolysis of phosphoryloxymethyl prodrug mediated by phosphatases. ............27

1-10 Hydrolysis of PEG conjugated to drug using a hydroxyl-benzylalcohol linker in
vivo. .................................................... ......... 3 0

1-11 P E G -A la-C am pothecan ...................................................................................... 30

1-12 Ionic complexes formed by oligoarginine of Tat with the phosphate groups in
cell membranes....................................... ........ 32

1-13 Chemical hydrolysis of oligoarginine conjugates of cyclosporine and taxol...........33

1-14 Structure of testosterone and indomethacin esters with polarisable side chains......39

2-1 Design of NANAOCAM promoiety ...........................................44

2-2 Structures of N-methyl-N-methyloxycarbonylaminomethyl prodrugs of
dimethylaminobenzoic acid and 6MP. .................................. .........47

2-3 Synthesis of NANAOCAM prodrugs ........................................ ............... 47

2-4 Synthesis of NArNAOCAM prodrugs.......................................48









2-5 Hydrolysis of NANAOCAM-Y in aqueous buffers..............................................60

2-6 Pseudo-first order rate constants (sec-1) versus a- of parent phenol ..................62

2-7 Pseudo-first order rate constants (sec-1) versus acidic pKa of parent. ....... .........62

2-8 pH rate profiles of compds 7 and 14 ........................................... 63

2-9 SN2 type of hydrolysis of NANAOCAM-Y with YV or carbamate as the leaving
group. ................................................64

2-10 SNI type of hydrolysis of NANAOCAM-Y.................. ................. ............64

2-11 Hydrolysis of N-alkylamidomethylcarboxylic acid esters in aqueous buffers. .......65

2-12 Hydrolysis of N-imidomethyl derivatives of phenols in aqueous buffers.............66

2-13 Attempts at the synthesis of NANAOCaminoethylidene chloride and
NANAOCaminobenzylidene chloride. .............................................. .....68

2-14 Hydrolysis of NArNAOCAM-carboxylicacid conjugates at pH 9.2. ........ .........73

3-1 Synthesis of NANAOCAM-Cl from 1, 3, 5-trialkylhexahydrotriazine.................76

3-2 Preperation of alkyl chloroformate in situ from alcohol and synthesis of
NANAOCAM -Cl from alkyl amine........................ ..................................... 77

3-3 A Franz diffusion cell ................... ............................ ....... .. ...... ...... 86

3-4 Plot of solubility parameter versus log PMIPM for 1, 17-20. ..................................93

3-5 Experimental versus calculated log maximum flux values through hairless
mouse skin from IPM using equation 3.15. ........................................96

3-6 Experimental versus calculated log maximum flux values through hairless
mouse skin from IPM using equation 3.16. .......................................97

4-1 Carbocations formed as intermediates by hydrolysis of Mannich bases of
theophylline and NANAOCAM-Th.............. ............. ...... .......... 100

4-2 Synthesis of NANAOCAM-Cl from 1, 3, 5-trialkylhexahydrotriazine.................102

4-3 Synthesis of NANAOCAM-Cl from alkyl amine and preparation of alkyl
chloroform ate in situ from alcohol........................ ...................................... 104

4-4 Alkylation of theophylline with NANAOCAM-Cl................................................106

4-5 Plot of solubility parameter versus log PMIPM for 21-26. ........... ...............120









4-6 Experimental versus calculated log maximum flux values through hairless
mouse skin from IPM using equation 4.13 for n = 69............... .. ............ 124

4-7 Experimental versus calculated log maximum flux values through hairless
mouse skin from IPM using equation 4.14 for n = 74.................................. 125
















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

PRODRUG STRATEGIES AIMED AT IMPROVING TOPICAL DELIVERY OF
DRUGS USING THE N-ALKYL-N-ALKYLOXYCARBONYLAMINOMETHYL
(NANAOCAM) PROMOIETY

By

Susruta Majumdar

August 2006

Chair: Kenneth. B. Sloan
Major Department: Medicinal Chemistry

Topical delivery is an attractive route to deliver drugs into systemic circulation;

however the poor biphasic solubility of drug molecules limits delivery across the skin. N-

Alkyl-N-alkyloxycarbonylaminomethyl (NANAOCAM)-drug conjugates were designed

as prodrugs with good biphasic solubilities to increase the topical delivery of phenol and

imide containing drugs.

Prodrugs are biological inactive derivatives of a drug which hydrolyse in vivo to the

parent drug molecule. These derivatives mask polar functional groups present in a drug

molecule which, in this case, leads to increased solubility of the drug in the skin.

Prodrugs must hydrolyse to exhibit their pharmacological activity. To elucidate the

mechanism of chemical hydrolysis, a series of NANAOCAM conjugates of phenols,

thiols and carboxylic acids were synthesized and their rates of hydrolysis determined in

aqueous buffers. The hydrolysis followed pseudo-unimolecular first order kinetics and

was dependent on the nucleofugacity of the leaving group. The hydrolysis was also









independent of the pH of buffers. To further elucidate the mechanism of hydrolysis, N-

aryl-N-alkyloxycarbonylaminomethyl (NArNAOCAM) conjugates of phenols and

carboxylic acids were also synthesized and their rates of hydrolysis determined. Since the

hydrolysis of NArNAOCAM conjugates were slower than NANOCAM conjugates, the

NANAOCAM conjugates are proposed to hydrolyse by a SNI type of pathway with the

lone pair on the nitrogen stabilizing the carbocation formed as an intermediate.

To investigate if the NANAOCAM promoiety increases the dermal delivery of

phenolic drugs and imide containing drugs, a homologous series of NANAOCAM-

acetaminophen and NANAOCAM-theophylline were synthesized. These derivatives

were characterized by determination of their solubilities in IPM and water, partition

coefficients between IPM and pH 4.0 buffer and flux through hairless mouse skins from

IPM.

Only two prodrugs of acetaminophen, Cl alkyloxy and C2 alkyloxy, increased flux

through the skin. These derivatives were the most water soluble prodrugs in the series of

more lipophilic prodrugs. In the theophylline series, only one derivative, C2 alkyloxy,

increased delivery through the skin. This derivative was the most lipid and most water

soluble member of the series. The flux of NANAOCAM prodrugs from IPM was

accurately predicted by the Roberts-Sloan equation.














CHAPTER 1
INTRODUCTION

The Skin

Skin is the largest organ in the human body, accounts for 10% of the total body

weight and has a surface area of 2m2 (Schaefer and Redelmeir, 1996). It acts as a physical

and chemical barrier, protecting the body from the surrounding environment. Besides

this, it enables the body to control fluid loss, regulate body temperature and intercept

external stimuli. The skin can be divided into three distinct layers, the hypodermis, the

dermis and the epidermis (Block, 2000). A microscopic view of skin is shown in Figure

1-1.

The hypodermis is made up of a network of connective fibers and adipocytes which

play an important role in energy storage and metabolism as well as providing insulation

and protection against injury. The intracellular fat droplets may also act as a reservoir of

hydrophobic compounds which may have penetrated the stratum corneum (the outermost

layer).The hypodermis is 1-2 mm thick. It houses blood vessels whose prime function is

to deliver nutrients to the skin and remove waste products, metabolites and xenobiotics

from the skin.

The dermis is the thickest layer of the skin and is 2 mm thick. It is responsible for

maintaining the structural integrity of the skin. The dermis consists of a network of

collagen and elastin fibers which provide flexibility and tensile strength to the dermis.

Proteoglycans fill up most of the intracellular space and these macromolecules are

responsible for the water-retaining properties of the dermis. These are polysaccharides









covalently bound to a polypeptide backbone. The polysaccharide side chains are

frequently sulfated. Examples of proteoglycans frequently seen in the dermis are heparin,


Hair
Shaft



Corneum
Epidermis Sebaccous
Gland


Apcrine
Sweat Gland
Demiis





Subcutaneous -W-030 11ily
pat


Hair Smooth Eccrine Lymphalic
Folliole Muscle Swea Gland Vessel
Figure 1-1 Cross sectional view of human skin (Reproduced with permission from K.G.
Siver, Ph.D. Dissertation, University of Florida, 1987).

chondroitin sulfate and dermatin sulfate. The anionic charge associated with these

proteoglycans also allows a lot of water to be associated with the carboxylate or the

sulfate groups and impedes the ability of lipophillic compounds to cross the dermis. The

dermis has an extensive vascular network which participates in various processes like

nutrition, exchange of gases, repair of tissues, immune responses and thermoregulation.

Nerve endings sensitive to vibration, pressure, pain and temperature are seen in dermis

along with mast cells, macrophages and t-cells which are present for protection against

foreign antigens. A variety of appendages which permeate the stratum corneum like

sweat glands, hair follicles and sebaceous glands are derived from this region. The hair









follicles and exocrine sweat glands have pores that extend to the skin surface and provide

openings to the epidermis which contains the stratum corneum: the main permeability

barrier of the skin. Thus xenobiotics can use this route, also called the shunt route, to

penetrate the dermis; this is especially true for highly polar molecules whose permeability

is limited by the stratum corneum. However these pores constitute only 0.1 % of the total

skin surface area and are responsible for early diffusion processes (Scheuplein and Blank,

1971; Stuttgen, 1982). Sebaceous glands are located within the hair follicles and secrete

sebum which makes its way on to the skin surface. Sebum is a complex mixture of fatty

acids, triglycerides, squalene and waxes. Eccrine sweat glands are also present in the

dermis and their primary purpose is thermoregulation. Since the epidermis is devoid of

any vasculature, it depends on the dermis for supply of nutrients, oxygen and removal of

waste metabolites. The presence of papillae from each layer interlocked with one another

at the border of the dermis and epidermis facilitates this process of exchange between the

two layers by increasing the surface area. The papillary layer is 50 nm thick and has

capillaries which feed the epidermis with the required nutrients (Bisset, 1987).

The epidermis, contains the stratum corneum, provides the main barrier function to

the skin and prevents the entry of foreign particles and loss of water. Keratinocytes

(keratin producing cells) make up most of the viable epidermis which is 50-100 [tm thick.

This protein is very polar with every third amino acid in the backbone of keratin

containing an ionizable side group (-COOH, -NH2 and -SH) (Flynn, 1990). Besides the

keratinocytes the viable epidermis also has melanocytes which are responsible for skin

pigmentation, langerhans cells for immune response and merkel cells responsible for

sensory reception. The epidermis also acts as a store for antioxidants like ascorbic acid,









tocopherols and tocotrienols, glutathiones, rates, ubiquinone-10 and enzymatic

antioxidants like catalases, superoxide dismutases, glutathione peroxidases. The primary

function of these antioxidants is to offer protection to the skin against the ultra violet

radiation of the sun (Thiele et al., 2002). Division and differentiation of the keratinocytes

lead to the formation of five distincts layers within the epidermis: stratum basale, stratum

spinoum, stratum granulosum, stratum lucidum and finally stratum comeum. The cells

migrate from the basal cell layer to the skin surface while gradually changing their

structure and function resulting in the formation of physically and chemically resistant

cell remnants called comeocytes which make up the stratum corneum. Once on the

surface of the skin the cells are shed off as a part of desquamation. This cycle of cell

division followed by differentiation, proliferation and migration from basal to the surface

and finally desquamation repeats itself over and over again with the new cells rapidly

replacing the old ones. The transition from a terminally differentiated keratinocyte to

corneocyte takes 24 h in the epidermis. The process of cornification and desquamation is

intimately linked; synthesis of stratum corneum occurs at the same rate as loss.

The outermost layer of the epidermis is called the stratum comeum, and it is the

main rate limiting barrier to drug delivery (Stuttgen, 1982; Berti and Lipsky, 1995). It is

10-20 [tm thick consisting of polyhedral, flat and nonnucleated cells called the

corneocytes (Berti and Lipsky, 1995).The pH of the stratum corneum lipids is 4-6.5. The

stratum comeum has 5-20 layers of comeocytes. These cells are remnants of the

terminally differentiated cells called keratinocytes which are found in the viable

epidermis. Their cell organelles and cytoplasm have disappeared during the process of

cornification. Corneocytes are composed of insoluble keratins surrounded by a cell









envelope stabilized by cross linked proteins and covalently bound lipids. These

corneocytes are further interconnected by lipids. The presence of these intracellular lipids

is required for a competent skin barrier and follows a tortuous path. These lipids are

synthesized and assembled into lamellar structures which surround the comeocytes.

These lamellar bodies which envelope the stratum corneum store a complex mixture of

lipids and lipid-like materials like free fatty acids, cholesterol and hydroxyamide

derivatives called ceramides. Interconnecting corneocytes of the stratum comeum are

polar structures such as corneodesmosomes which contribute to stratum comeum

adhesion. The stratum comeum is not uniformly homogeneous and the layers represent

various stages of comeocyte and lipid maturation. In short, the stratum corneum can be

visualized as a brick and a mortar model with the flattened cornified cells being the offset

layers of bricks lying one over another and the lipid rich intracellular matrix as the mortar

separating the bricks or corneocytes (Figure 1-2). By physically stripping the outermost

layer of the skin an increase in permeability of various chemicals and water can be seen

proving stratum comeum to be the main rate limiting barrier. Thus topical approaches at

increasing flux through the skin have relied on increasing the solubility of compounds in

the stratum corneum (Sloan et al.1989, 1992 and 2003). It is primarily believed that drugs

cannot penetrate the protein rich corneocytes cells but need to take a tortuous path

alongside the corneocytes through the interconnecting lipids (Michaels et al., 1975). Thus

the drug enters systemic circulation through this narrow channel of lipids and need to be

hydrophobic so that they can permeate these semi-fluid matrix lipid barriers.













Corneocyte


S_ kI


Lipid
matrix -


9
I-


Epidermis



Dermis


Figure 1-2 Brick and mortar model of drug absorption through the skin.
Unlike cell membranes in most parts of the body which are composed of

phospholipids, the skin is devoid of any phospholipids. However, the presence of

functional groups which are capable of hydrogen bonding in ceramides (Figure 1-3), free

fatty acids, and cholesterol sulfate in the lamellar bodies which surround the corneocytes

makes the stratum corneum hydrophilic as well as lipophilic. The stratum corneum is

about 30% water by weight and most of it is due to the association of water with these

polar groups seen in the lamellar lipids.

Thus, drugs with good biphasic solubility can permeate the stratum corneum better

than drugs which are merely hydrophobic.














0 0

HN
'HNOH
OH

Ceramide 1




HN


Ceramrde2 O


OH


Ceramide 2
HN

























CeramCde 6
/\_ tOH
OH









OH










Ceramde eramide 3
HN
OH

























CeramdeCeramide 4
OH

-'O




Ceramide 5C
























Figure 1-3 Structures of ceramides seen in the lipid bilayers of the lamellar bodies.


The heterogenous composition of the lipids and disorder in packaging of these



lipids leadS to the formation of lipid microdomains. Ruthenium tetroxide staining of


normal skin shows the presence of alternate hydrophilic regions connecting lipophilic


bands in the bilayer structure (Hou et al., 1991). The structure and arrangement of the


barrier is such that drugs need to pass through alternate layers of hydrophobic and
barrier is such that drugs need to pass through alternate layers of hydrophobic and









hydrophilic regions, formed as a result of the packing disorder of the lipids present, to

reach the systemic circulation and thereby must possess adequate lipid as well as water

solubilities (van Hal et al., 1996). The experimental evidence for this concept of biphasic

solubility being important in terms of permeability through the skin was first provided by

Scheuplein and Blank (1973) for the diffusion of a homologous series of alcohols from

water through human skin in vitro.The alcohol which gave maximum flux was not the

most lipid soluble but was the most water soluble. Sloan et al. (1997) confirmed these

results in in vitro hairless mice skin.

Topical Delivery

Most drugs developed today are delivered through the oral route. This is primarily

because it is economical for the company developing the drug. The delivery of the drug

does not require a visit to a physician or a trained medical representative as would be the

case for an intravenous or parental formulation. Drugs given orally also have the ability

to be given in large doses because of the high permeability of the enterocyte cells present

in the GI tract. However, the pH of the stomach is acidic, so the stability of molecules is

compromised. Oral bioavailability of drugs is sometimes limited by efflux of active drugs

by Mrp or P-gp transporters which prevent drugs from reaching intestinal vasculature

(Wacher et al., 1996; Watkins, 1997; Schinkel, 1997 and De Mario et al., 1998) and CYP

450 catalysed metabolism of drugs (Watkins, 1992). The portal vein carries the drug from

the intestine to the liver by a process known as "enterohepatic cycling" before the drug

finally enters systemic circulation. The liver is the largest drug metabolizing organ in the

body and has enzymes like CYP 450, GSH/GST and UDP/UGT which do not have strict

substrate structure specificity. These enzymes transform the drug molecule to an inactive









metabolite or lead to the formation of toxic by-products and hence limit the therapeutic

effectiveness of the drug.

To circumvent these problems associated with oral drug delivery, an alternate and

useful site of delivery can be the skin. Dermal delivery evades first pass metabolism of

drugs because once the drug permeates the skin it reaches the systemic circulation and

hence the bioavailability is not limited by metabolism, and side effects arising out of

toxic metabolites are minimized. Dermal drugs are more likely to reach systemic

circulation intact because of the absence of high concentrations of drug-metabolizing

enzymes like CYP 450 enzymes, GSH/GST (concentration in nmol amounts, Shindo et

al., 1994) or P-gp efflux transporters. CYP isozymes have been shown to be induced in

skin (Ahmad et al., 1996, 2004 and Swanson, 2004) and P-gp transporters are seen in

epidermal cells but their primary function seems to be limited to clearing out endogenous

compounds (Laupeze et al., 2001). The concentration of the enzymes is low and there is

no systematic arrangement of the enzymes and transporters on the surface of the skin so

they cannot clear the drugs efficiently or affect the therapeutic efficacy and availability of

drugs. Topical delivery can be used for delivery of potent molecules whose dose

requirements are low or require slow and sustained serum concentration levels of drug for

a prolonged period of time. It is possible to deliver about 30 mg/4.9 cm2 of an

appropriately soluble drug through the skin over a period of 24 h (Beall et al., 1996).

Dermal delivery means delivery into the skin, i.e., the epidermal and dermal skin cells.

Such delivery is particularly useful for skin cancers. Transdermal delivery means delivery

through the skin into systemic circulation. Drugs with high transdermal delivery

invariably show high dermal delivery too, though it may be possible to increase dermal









absorption by making a hydrolytically labile prodrug which hydrolyses to give a

lipophilic promoiety whose permeation through the dermis is limited and hence

selectively accumulates in the dermis.

Skin is a useful target organ for skin disorders and local action. Attempts are

currently being made to deliver soft estrogens locally because estrogen receptors are

expressed in the skin (Labaree et al., 2001 and 2003). The idea is to increase uptake

locally and prevent systemic circulation and hence the side effects that may result from it.

These locally active estrogens can be used to treat vaginal dyspareunia without the risk

involved with systemic administration. These soft drugs are inactivated by esterases so

that their action is geographically limited to the site of action.

Delivery of 5-FU through the skin enables treatment of actinic keratosis (Dillaha et

al., 1965) and may be useful for psoriarsis (Tsuji and Sugai, 1972) and yet spare the body

from most of its systemic effects. Nicotine, fentanyl, estradiol and nitroglycerine are used

in some transdermal patches available in the market today. Some attractive targets for

transdermal delivery include alcohol cessation agents like naltrexone, testosterone, and

antioxidants like vitamin C and vitamin E. Currently, the annual US market for

transdermal patches is greater than $3 billion dollars; many companies are thus

recognizing the value of transdermal delivery (Prausnitz et al., 2004).

The problems associated with topical delivery include local irritation or allergic

reactions on the skin because of formulation contents and poor penetration of drugs

through the stratum corneal barrier due to poor physicochemical properties of the

molecule.











Approaches to Increase Permeation across the Skin

The three approaches commonly used in increasing transdermal delivery of drugs are.

1. Mechanical or electromechanical methods like ultrasound, iontophoresis and
microneedles.

2. Penetration enhancers and hydration.

3. Prodrugs.

Ultrasound can be used to increase the transport of both low molecular weight

compounds and macromolecules like insulin and interferon. It relies on usage of low

frequency waves which causes disorganization of lipid bilayers and increased

permeability (Mitragotri and Kost, 2004).

Iontophoresis is the use of an electric current applied across the skin to drive drugs

through the epithelium. The flow of electric current increases the permeability of the

skin. The current does not pass through the skin uniformly but transappendageally

through pores or sometimes through lipid channels. The electrical potential provides an

electromotive force that is capable of driving charged molecules through the stratum

corneum (Riviere and Heit, 1997). The formation of HCl and NaOH by the electrolysis of

NaCl leads to a lowering of skin pH and are responsible for the irritation associated with

iontophoresis (Mitragotri et al., 1996).

Microneedles have also been used for transdermal delivery of macromolecules.

Micron sized needles in the patch pierce the skin into 10-15 lm of the upper stratum

corneal layer so that they do not stimulate the nerves found in the lower layers and hence

cause no pain while increasing the permeability of the drug used (Henry et al., 1998).









The second approach to increase permeation is by using penetration enhancers and

hydration. The two distinct mechanisms by which penetration enhancers work are, the

'push' and the 'pull' mechanisms. The 'push' mechanism relies on the use of volatile

components in a formulation to drive a drug into the skin. Evaporation of the volatile

components leads to a supersaturated solution of the drug with thermodyanamic activity

greater than one and results in increased flux through the membrane (Kadir, 1987). The

'pull'mechanism on the other hand increases topical absorption by the use of vehicles

which interact with the skin and decrease the diffusional resistance by disrupting the

barrier, i.e., increase the degree of hydration, leach the lipid components in the stratum

corneal barrier or increases the solubilizing capacity of the skin, so that flux of drug can

increase (Barry, 1987). It is possible to increase the flux of a compound by about 10 fold

by merely altering the vehicle properties before damage to the skin occurs (Sloan, 1992).

Commonly used penetration enhancers include azone, DMSO, propylene glycol and

decylmethyl sulphoxide. Keratolytic agents like salicyclic acid also increase penetration

through the skin but cause damage.

Hydration of skin leads to an increase in elasticity and permeability of the stratum

corneal barrier. Occlusion is the most command way to increase the hydration of the skin

as it prevents loss of moisture from the surface of skin by evaporation. It also leads to an

increase in temperature and humidity on the surface and thereby to increased

permeability of the drug. Topical formulations with fats and oils like glycols, glycerols,

paraffins, waxes, silicones and clay also increase hydration (Block, 2000).

Prodrugs represent yet another way in increasing topical absorption. They are

biologically inactive and labile derivatives of a drug which revert to the active drug









molecule by chemical or enzymatic hydrolysis. In this the case the drug in the form of a

prodrug has better solubility in the skin and hence gets pulled into the skin. An advantage

of prodrugs over penetration enhancers is that prodrugs are a 1:1 molecular combination

of drug and the promoiety compared to penetration enhancers where excess amounts are

used which may cause allergic reactions or local irritation. Prodrugs thus increase

solubility of the drug in the skin while penetration enhancers increase the solubalization

capacity of the skin for a drug molecule. Combination of prodrug with a penetration

enhancer in the same formulation may prove to be more useful (Waranis and Sloan,

1987) but hasn't been fully utilized yet.

Theory of Percutaneous Diffusion

The mechanism of permeation across human skin relies on passive diffusion which

is driven by a difference in the concentration gradient across the membrane. Any solute

diffusing through the skin shows an initial, nonlinear increase in drug concentration

which represents a build up of the drug in barrier region. The emergence of drug from the

barrier region is different from drug to drug because of the time required to saturate the

membranes. This is also true for most controlled release systems like transdermal

patches. This is called a lag-time effect. Once steady state is reached a linear increase of

drug concentration is seen which represents a condition where the exodus of solute from

the dermal side is equal to the mass of material entering the epidermal side (Poulsen,

1971). By simply extrapolating the steady state line to the abscissa it is possible to

determine the time to reach steady state. It also gives an estimate of how much drug

permeates through the transfollicular shunt route. This is very well illustrated by Figure

1-4.














o < 45














TIME (hours)
Figure 1-4 Illustration of a typical flux profile.

The rate of mass transfer across a membrane or flux (J) is proportional to the

concentration gradient expressed across the membrane. Imagine two solution filled

compartments containing different concentrations of some compound (CD and CR) and

separated by a permeable membrane of thickness h (Figure 1-5).The rate at which the

drug diffuses across the membrane from one compartment to the other is described by

Fick's first law. Equation 1-1 shows the relationship in a mathematical form and forms

the basis of Fick's first law (Fick, 1855).

J = (dM/dt) unit area = -D (dC/dx) (1-1)

where dM is the amount in mass or moles of solute passing through the membrane in

time dt, dC/dx is the concentration gradient within the membrane over infinitely small

distances and D is called the diffusion coefficient diffusivityy): its units are area per unit

time (cm2/s). The negative sign indicates that the concentration decreases with distance.






















Figure 1-5 Two compartment diffusion model.

To utilize Fick's law experimentally, the concentration in both compartments (CD

and CR) are maintained constant so as to have a constant concentration gradient (Ci-C2)

across the membrane once equilibrium is established and sink conditions are maintained

i.e., CR--O. The partial differential can be replaced by Ci, C2 and h. The concentration

gradient can further be simplified as

dC/dt = (C1-C2)/h (1-2)

where Ci is concentration of the solute in the donor side of the membrane, C2 is the

concentration of the solute on the receptor side of the membrane and h is the thickness of

the membrane as shown in Figurel-5. Combining (1-1) and (1-2) we get

J = (dM/dt) = -D (Ci C2)/h (1-3)

Both C1 and C2 are difficult to measure being inside the membrane.

However under sink conditions, C1-C2-C1, C1 = SSkin (solubility in the skin), J = JMV

(maximum flux from a saturated solution of drug in a vehicle). To maintain sink

conditions it is necessary that the concentration of drug in the donor compartment doesn't

change with time and hence a suspension in equilibrium with a saturated solution is

generally used for in vitro diffusion experiments and excess drug is used in a transdermal

patch formulation. This type of transport process follows zero order kinetics.


CR
C, Receptor
compartment
(C2 CR)
CD
Donor
compartment
(CD C1)









JMV = D SSkin (1-4)

h

SSkin = Sv. K Skin: V (1-5)

Inserting (1-5) in (1-4)

JMV = (D K Skin: V Sv) (1-6)

h

where Sv is the solubility of solute in the vehicle, JMV/ (Sv) = P:

P = (K Skin: VD) (1.7)

h

P is known as the permeability coefficient and can be determined experimentally. Since

DA/h is constant, permeability coefficient is proportional to vehicle/membrane partition

coefficient of the drug. The more soluble the drug is in the vehicle, the lower the

permeability coefficient for the delivery of the drug from that vehicle (Sloan, 1992).

The partition coefficient between drug in vehicle and in the membrane can be calculated

from theory using equation (1-8).

In K = [(6i -6) 2 Dv2- (6i -_6) 2 s2]Vi (1-8)

RT

where 6i is the solubility parameter of the drug, 6, of the vehicle (8.5 (cal/cm3)1/2) for

IPM isopropyll myristate) and 6, is the solubility parameter of skin (10 (cal/cm3)1/2), Vi is

the molar volume of the drug (cm3/mol), R is the gas constant (1.98 cal/K) and T is the

temperature (305 K) (Sloan et al., 1986; Sherertz et al., 1987; Sloan et al.,1986a). The

solubility parameter of the drug can be determined experimentally or from individual









group contribution methods (Sloan et al., 1986; Sherertz et al., 1987; Sloan et al., 1986a;

Fedors et al., 1974; Martin et al., 1985; Martin et al., 1985a).

Mathematical Modeling of Flux through Human Skin. Derivation of Roberts-Sloan
equation from Fick's law of Diffusion (Roberts and Sloan, 1999).

When the vehicle is water, equation (1-7) which relates permeability P with

partition coefficient between skin and the vehicle can replaced by equation (1-9).

P = (K Skin: AQ D)/h (1-9)

The diffusivity or diffusion coefficient can be estimated from molecular volume through

the following relationship (Cohen and Turnbull, 1959).

D = Doe-z MV (1-10)

The partition coefficient between skin and the vehicle is difficult to measure and thus can

be estimated from the partition coefficient between OCT (octanol) or IPM isopropyll

myristate) and water (AQ) where a lipid like vehicle (OCT or IPM) replaces skin, which

is considered to be lipid-like.

K Skin: AQ= (KOCT: AQ) (1-11)

Here y represents the difference between the partitioning domain of the skin with respect

to the solvent that replaces it, in this case OCT. The closer y is to 1, the closer the solvent

is to being is a good surrogate of stratum corneum lipids, e.g., ether has a y value of 0.53

while OCT has a y value of 0.7. Thus, OCT mimics the skin more closely than ether

does.

By substituting equations (1-10) and (1-11) in equation (1-9) we get equation (1-12)

P= [(KOCT: AQ) Do e-Z MV]/h (1-12)









By collecting the constants, taking the log of both sides and replacing MV (molecular

volume) by MW (molecular weight), the Potts-Guy equation (Potts and Guy, 1992) can

be obtained.

log P = y log KOCT: AQ- z MW +x (1-13)

where x = Do/h

The Roberts-Sloan equation previously known as the transformed Potts-Guy model

can be derived from the Potts-Guy model. The Potts-Guy is useful in predicting

permeability through the skin but its prime limitation is that it correlates permeability

with partition coefficient between a lipid vehicle and polar vehicle and hence is inversely

proportional to solubility of the drug in a polar vehicle like water. This has serious

implications in drug design and misleads medicinal chemists to synthesize more lipid

soluble derivatives in order to optimize topical delivery of a drug. We will talk about the

importance of biphasic solubility on topical delivery in the chapters that follow. A model

for flux (J) instead of permeability (P) is more clinically relevant because it actually tells

us how much solute in moles (amount) is going into the skin compared to permeability.

Since:

JMAQ = (P) (SAQ) (1-14)

Inserting equation (1-14) in equation (1-13) we get

log (J/SAQ) = y log KOCT: AQ z MW + x (1-15)

log JMAQ log SAQ = y log KOCT: AQ z MW + x

log JMAQ log SAQ = y log SOCT y log S AQ z MW + x

log JMAQ = y log SOCT y log SAQ -z MW + log SAQ + X

log JMAQ = x + y log SOCT + (1-y) log SAQ z MW (1-16)









Equation (1-16) is commonly referred as the Roberts-Sloan equation (RS) and in it flux is

proportional to both the lipid and the water solubilities of a drug and inversely

proportional to molecular weight.

If IPM instead of water was used as the vehicle in diffusion cell studies, because

some prodrugs were unstable, a different derivation is required.

Starting at equation (1-6), which was previously derived, but using IPM as the vehicle

and D = Do exp (-z MW) as in equation (1-10) we get

JMIPM =Do exp (-z MW) K Skin: IPM.SIPM

h

Taking the log of both sides and combining constants (x = Do/h) we get

log JMIPM= X + log K Skin: IPM + log SIPM Z MW (1-17)

Since,

K Skin: IPM = K Skin: AQ/ K IPM: AQ (1-18)

And since equation 1-11 relates the partition coefficient between skin and the vehicle to

the partition coefficient between OCT and AQ; where the vehicle is IPM, we can replace

OCT with IPM to give equation (1-19).

K Skin: AQ = (K IPM: AQ) (1-19)

Taking the log of both sides of equation (1-18) and substituting (1-19) for KSkin:AQ we get

equation (1-20)

log KSkin: IPM = y log K IPM: AQ log K IPM:AQ

log KSkin:IPM = y log SIPM y log SAQ log SIPM + log SAQ (1-20)

Substituting equation (1-20) into equation (1-17)

log JMIPM = x + y log SIPM y log SAQ log SIPM + log SAQ + log SIPM z MW









Collecting the terms we arrive at the RS equation.

log JMIPM = x + y log SIPM + (1-y) log SAQ z MW (1-21)

or

log JMV = x + y log SLIPID + (1-y) log SPOLAR z MW (1-22)

Equation 1-22 is the general form of the RS equation.

From equation 1-4

JMV = D SSkin (1-4)

h

Inserting equation 1-10 (D = Do exp (-P MV)) into equation 1-6 we arrive at:

JMV = Do exp (-P MV) SSkin (1-23)

h

Taking the log of both sides and collecting the constants we get:

log JMV = x + log SSkin Z MV (1-24)

where x = D/h and z = P

Correlating equation 1-24 with equation 1-22 (RS) we get

log SSkin = y log SLIPID + (1-y) log SPOLAR (1-25)

Equation 1-25 illustrates that solubility in skin can be modeled by solubility in two

phases: solubility in a lipid phase and solubility in a polar phase. Biphasic solubility thus

becomes an important determinant in optimizing flux through the skin.

The RS equation has been used to predict the topical delivery of homologous series

of prodrugs across hairless mouse skin from their suspensions in isopropyl myristate

(Wasdo and Sloan, 2004) and water in vitro (Sloan et al., 2003) and permeation of

nonsteroidal anti-inflammatory drugs through human skin from mineral oil in vivo










(Wenkers and Lippold, 1999 and Roberts and Sloan, 2001), permeation of solutes,

chemicals and drugs from water in vitro (Flynn, 1990 and Majumdar et al., 2004) and

penetration of sunscreens and antimicrobials in in vivo human skin from PG/water

(Leweke and Lippold, 1995 and Majumdar and Sloan, 2005). In all cases a significant

dependence of flux on water solubility was observed regardless of the vehicle used. Table

1-1 shows the coefficients of x, y, z and r2 of these databases. The RS equation can used

to predict flux of any drug molecule across skin from any vehicle if three

physicochemical parameters SLIPID (solubility in a lipid like vehicle like OCT or IPM),

SPOLAR (solubility in a polar vehicle like water or propylene glycol) and MW are known.

Table 1-1 x, y, z, r2 and Average Residual Errors for Various Databases Fit to RS.
Fit to RS (see text)
Model Database Vehicle n x y z r2 AlogJM
in vivo Wenkers Mineral 10 -1.459 0.722 0.00013 0.93 0.133
Human and oil
Skin Lippold
(1999)
in vitro Sloan and Water 18 -1.497 0.66 0.00469 0.77 0.193
Hairless coworkers
Mouse (2003)
Skin
in vitro Wasdo Isopropyl 61 -0.491 0.52 0.00271 0.91 0.15
Hairless and Myristate
Mouse Sloan
Skin (, 111 i 4
in vivo Flynn Water 103 -2.571 0.56 0.00444 0.9 0.44
Human (1990)
Skin
in vivo Leweke 30% 10 -2.116 0.45 0.00048 0.97 0.11
Human and PG/ water
Skin Lippold
(1995)

Prodrugs

The word prodrug was coined by Adrien Albert to describe compounds that

undergo biotransformation prior to eliciting their pharmacological effects (Albert, 1958).

It is impossible to review prodrugs of all classes; we will therefore cover only the basic









principles involved in prodrug design and their applications in clinical practice (Ettmayer

et al., 2004).

Prodrugs have been primarily designed to increase oral bioavailability, solubility,

enhance chemical stability, prevent premature metabolism, decrease toxicity and improve

taste (Stella, 1985). The primary function of a prodrug is to mask a polar functional group

(-XH) where X can be -OH (phenolic or alcoholic), -COOH, -SH, -NH amidee, imide or

amine) in a transient manner so that once the prodrug is in the target site (which may be a

tissue, cell or membrane) it hydrolyses to release the active drug molecule (Bundgaard,

1991). Since the polar functional group is masked, the tendency of the drug molecule to

form intramolecular hydrogen bonding is also blocked. This leads to improved solubility

properties of the prodrug in both lipid and aqueous phases with respect to the drug

(Bundgaard, 1991). By merely increasing biphasic solubility of drugs it is possible to

improve absorption and permeability across biological membrane barriers like the

enterocyte cells of the GI tract or stratum corneum of skin (Sloan and Wasdo, 2003). An

increase in absorption and permeability results in increased bioavailability. Prodrugs

containing charged promoieties like phosphates, amino acids, hemisuccinates and amines

have often been used in the promoiety to increase the aqueous solubility of drugs which

leads to increased dissolution rates and enhanced oral (Fleisher et al., 1996) and

transdermal absorption of the drug (Sloan and Wasdo, 2003). Masking of a functional

group transiently also may allow drugs to evade P-gp mediated efflux across the GI tract.

(Jain et al., 2004).

To make a hydrolytically labile prodrug it is necessary to attach the drug to a

promoiety which is stable enough to reach the target site but labile enough to release the









drug efficiently. Most prodrugs utilize an ester promoiety because of the presence of

esterases which are largely nonspecific and found in most tissues, biological fluids,

organs and blood etc, and which hydrolyse the prodrug to give the corresponding drug

(Beamount et al., 2003). A schematic representation is shown below (Figure 1-6).

0 ESTERASES

DRUG OR DRUG-COOH+ ROH


0
ESTERASES
DRUG-X R' ESTERASES DRUG-XH + R'COOH


X = N, 0, S; R = ALKYL; R' = OALKYL, N-ALKYL
Figure 1-6 Esterase mediated hydrolysis of prodrugs.

Carbonates of drugs have been made where the esters were too labile. Insertion of

an oxygen atom in the promoiety led to greater stability of the corresponding prodrug in

vivo because of the +1 effect imparted by the oxygen atom which decreases the

electrophilicity of the carbonyl group towards esterases. Similarly a nitrogen atom can be

inserted in the promoiety if a more stable derivative is desired.

Since a prodrug needs to hydrolyse to be effective, the inability of esterases to

efficiently transform penicillin esters to penicillin led to the design of soft alkylated

derivatives which are alkylcarbonyloxymethyl (ACOM) prodrugs. The ACOM promoiety

relies on a 'OCH2' spacer between the drug functional group and the labile ester

functionality. Penicillin esters are sterically hindered towards enzymatic hydrolyses

however insertion of a 'OCH2' spacer made it easier for the esterases to freely access the

ester functionality and hydrolyze it efficiently (Jansen and Russell, 1965).These soft

alkylated derivatives rely on esterase hydrolysis to yield a hydroxymethyl derivative of









the parent drug which is chemically labile and yields the parent molecule and

formaldehyde (Figure 1-7). These types of promoieties rely on both enzymatic and

chemical hydrolyses to release the drug molecule at the target site. In cases where the

ACOM moiety was not sufficiently stable, the alkyloxycarbonyloxymethyl (AOCOM)

promoiety have been used where the ester is replaced by a carbonate functional group.

The use of soft alkylation has been extended to other drugs containing various functional

groups like imides (Bodor and Sloan, 1977; Buur and Bundgaard, 1985 and Taylor and

Sloan, 1998), amines, phenols (Sloan et al., 1983a and 1983b; Bundgaard, et al., 1986

and Seki et al., 1988); alcohols (Beamount et al.), thiols (Sloan, 1983) etc.

0 0 0
ESTERASES -HCHO
DRUG O O R ASES RUG 0 OH -HCHO DRUG-COOH

0
ESTERASES -HCHO
DRUG-X 0 R DRUG-X OH DRUG-XH

X = N, O, S; R = ALKYL or OALKYL
Figure 1-7 Soft alkyl prodrug hydrolysis.

One of the critical issues involving prodrugs is that they should be labile but if they

are too labile they won't reach the target site and their purpose will be lost. Thus it is

important to design derivatives accordingly and use esters or carbonates (ACOMs or

AOCOMs) depending on the problem in hand and fine tune stability to meet design

directives. Sometimes it is necessary to release drugs slowly and in such cases prodrug

derivatives need to be more stable so that the residence time in the body is enhanced. We

will look at some examples of such prodrugs later on.

One of the prime reasons for esters being so widely used as prodrug promoieties is

that they are easy to make and are economical. They release non-toxic side products on









hydrolysis which is an important factor in drug development. Soft alkylated prodrugs

release formaldehyde which is relatively innocuous and biocompatible since it is released

as a byproduct during 0 or N-dealkylation of drugs by CYP 450 catalyzed oxidative

reactions in liver and intestines (Beamount et al., 2003).

Another type of soft alkylating promoiety which has been useful in increasing oral

bioavailability of phenolic drugs like estradiol is 0-imidomethyl saccharin. The oral

bioavailability of estradiol is drastically reduced by sulfation and glucuronidation of the

phenolic functional group which leads to rapid elimination of the drug from the body

before or after it is absorbed. Such premature metabolism can be prevented by transiently

masking the phenolic group with an imidomethyl group such as O-imidomethyl-saccharin

(Figure 1-8). The potency of the drug delivered as a prodrug was increased by about 8

times compared to an equimolar dose of estradiol alone (Patel et al., 1991). The prodrug

has a half life of 100 min under physiological conditions and the mechanism of

hydrolysis is believed to be SN2 (Getz and Sloan, 1993), although others suggest an

addition to the carbonyl group followed by a vinylogous elimination (Iley et al., 1998).

Both mechanisms may be operating. This type of prodrug approach is independent of

enzymatic hydrolysis and can be useful in cases where enzymatic variability becomes an

issue; the ester promoiety is enzymatically too labile or efficient hydrolysis to the active

principle is limited by strict structure specificity of the enzymes as in case of blood

(Beamount et al., 2003).

Although pharmaceutical scientists have often utilized esterases to biotransform a

prodrug to its active parent drug, human alkaline phosphatases are another class of











-OH



O=S-N\/- SN2
S0


0'O
HO

-HCHO


HOI


-OH


Figure 1-8 Hydrolysis pathways for Saccharin based 0-imidomethyl derivative of
estradiol.

enzymes which are found throughout various tissues (Strigbrand and Fishman, 1984) and

which have been used by chemists to dephosphorylate phosphoryloxymethyl prodrugs.

Fosphenytoin (Stella, 1996) is one such drug which is activated by phosphatases to give

hydroxymethyl phenytoin which spontaneously loses formaldehyde to generate phenytoin

(Figure 1-9). Delivery of phenytoin was a problem because of its poor aqueous solubility


S NH









(20-25 [g/mL) and was formulated at pH 12 when given parentally; crystallization of

drug at the injection site was seen. The aqueous solubility was enhanced by using the

phosphoryloxymethyl promoiety by about 4410 times. The water solubility of the

prodrug increased to 142 mg/mL (88.2 mg/mL equivalent of phenytoin).




HO
N HO
0 O
O N N
H PHOSPHATASES
0 N
H

FOSPHENYTOIN

I -HCHO

HN

O N
H


PHENYTOIN
Figure 1-9 Hydrolysis of phosphoryloxymethyl prodrug mediated by phosphatases.

This promoiety has also been used to increase the solubility of amines and alcohols

(Safadi et al., 1993 and Krise et al., 1999). Thus phosphoryloxymethyl is another

example of soft alkylating promoiety. The generic chemical structure of soft alkylating

promoities is shown below. Phosphoryloxymethyl has two X"'R's, i.e.,

X'-P=X"(X"'R')2.










R X"

DRUG-X X' 'X"'R'

X = N, O, S; R = H or ALKYL; X' = N or O; Y= C or P; X" = O; X"' = ALKYL, N or O; R' ALKYLL and H

The following discussion about the use of prodrugs to increase delivery covers some

miscellanenious examples which include the use of a novel enzyme system

(amidoreductases), use of a water soluble polymer (polyethylene glycol) and a

polypeptide (Tat protein).

Amidoximes are bioconverted to amidines by reductases expressed in the kidneys,

liver, brain, lungs and GI tract (Clement, 2002). The double prodrug Ximelagatran relies

on amidoreductases and esterases to yield the active drug melagatran.

H2N H2N
+/H /TOH
NH N


\ [H H 0 N-
H N N H
0 N N
0 0



MELAGATRAN
XIMELAGATRAN


Melagatran is a charged molecule at physiological pH with the carboxylate anion

(acidic pKa 2), amidine (basic pKa 11.5) and secondary amine (basic pKa 7)

functional groups making it very hydrophilic. As a result absorption across the GI tract is

decreased and its oral availability compromised. In the prodrug, the carboxylate ion is

blocked as an ester. Since the ester is electron withdrawing the pKa of the secondary

amine is reduced to 4.5 and it is no longer charged at physiological pH. The amidine is

protected as an amidoxime whose basic pKa was reduced to 5.2 so it is no longer charged









either. Thus, the resultant molecule is neutral at physiological pH, 170 times more

lipophilic based on K values and about 3-6 fold more bioavailable as a result of increased

absorption (Gustafsson et al., 2001 and Clement and Lopian, 2003).

In the prodrugs we have talked about till now, the drug was attached to the labile

promoiety directly or through a linker. In the following example, a drug is conjugated to

the linker (which should be a labile and bifunctional group), through a bridging

functional groups, the linker in turn is attached to a promoiety (which may be something

that increases the solubility of the drug or increases its uptake). This type conjugate is a

new class of prodrugs called tripartate prodrugs. Amino acids, aminoethoxyalcohols,

hydroxybenzylalcohols, aminobenzylalcohols and o-hydroxyphenylpropionic acids have

all been used as bifunctional linkers through which PEGs can be transiently attached to

drugs (Greenwald et al., 1999 and 2000). The active drug molecule is released by

hydrolysis in two steps. An example utilizing hydroxyl-benzylalcohol as the linker and

CO2 as the bridging functional group is shown below for an amine drug like

Daunorubicin (Figure 1-10).These polyethyleneglycol ether (PEGs) prodrugs have been

used to increase the systemic circulation time. Greenwald and coworkers (1996, 2001,

2003) have evaluated a series of PEGylated conjugates of anti cancer drugs like

Campothecan and Taxol with PEGs of various molecular weights. PEG40000-Ala-

Campothecan (Prothecan) is presently in phase II clinical trials for treatment of lung,

pancreatic and gastric cancers. Here the molecule has the aminoacid alanine acting as a

linker between the Campothecan and PEG (Figure 1-11).










0 ESTI
PEG 0 / 0ST


0



HOOH
HO\ OH


ERASES O -D
0 0 NDRUG
H





OH 2+ CO

+


DRUG-NH2
Figure 1-10 Hydrolysis of PEG conjugated to drug using a hydroxyl-benzylalcohol linker
in vivo.


SOH


Figure 1-11 PEG-Ala-Campothecan

These conjugates have higher water solubility than Campothecan which allows

increased dissolution of the drug. Also, since the molecular weight of the conjugates is

close to 45KDa (which is the renal threshold limit), these drugs are retained in the body

longer because of a phenomenon known as 'enhanced permeation retention' (EPR)

(Maeda et al., 1992). These prodrugs need to circulate in the body for longer periods and

release the drug slowly in a controlled manner; they have half lives >10 h compared to

other traditional prodrugs with half-lives of 20 min. These molecules evade glomerular

filteration of kidneys which excretes molecules whose size is 30KDa (Yamaoka et al.,

1994). Thus the drug selectively accumulates in the cancer cells taking advantage of the









leaky tissues and reduced drainage seen in cancerous tissues. This selectively towards

cancerous cells leads to reduced toxicity and better targeting of anti cancer agents.

Drugs have also been conjugated to oligoarginine residues to increase their uptake

across cell membranes and skin. Certain proteins contain subunits like Tat 49-57

(RKKRRQRRR) of HIV-1, that enable the translocation of proteins across the plasma

membrane into the cells (Green and Loevenstein, 1988; Anderson et al., 1993 and

Lindgren et al., 2000). The intact Tat protein has been conjugated to proteins to enhance

their delivery (Kim, 1997 and Nagahara, 1998). The chirality of the arginine residue (D

or L) didn't affect cellular uptake. The Tat protein efficiently crosses membranes of cells

in an energy dependent fashion after endocytosis (Mann and Frankel, 1991; Vives, 2003

and Fuchs and Raines, 2004). There is some evidence for guanidium rich molecules

undergoing receptor mediated endocytosis due their big size i.e. MW >3000. An

alternative mechanism of uptake is believed to be electrostatic interaction between the

positively charged guanidine (pKa ~ 12-13) and the carboxylate ion of the lipid bilayer or

phosphate group of the phospholipids bilayer to form a transient ion pair complex (Figure

1-12) which is less polar than the drug-Arg conjugate (Luedtke et al., 2003 and Rothbarg

et al., 2002). The driving force for the passage through the membrane is voltage potential

across most cell membranes. The positively charged complex formed due to incomplete

ion pairing with anionic cell components moves along the direction of transmembrane

potential intracellularr K levels extracellularlar K+ levels). Uptake is decreased when

membrane potential decreases extracellularr and intracellular K+ levels equal) while it

increases when a peptide antibiotic like valinomycin which selectively shuttles K+ across

the membrane is used in the assay. The complexes dissociate on the inner leaf of the









membrane. Truncation studies and synthesis of a series of analogues of Tat 49-57 revealed

that short chain oligomers (7-9 residues) of arginines are more efficient in translocation

than the entire protein (Wender et al., 2000). Wender and coworkers analyzed a series of

Cyclosporin-Arg7 and Paclitaxel-Args conjugates for uptake across human, mice skins

and various in vitro cell line surrogates of plasma membrane. They found that it is

necessary to have the guanidine moiety; oligolysine, oligohistidine or citruline residues

were not as successful as oligomers of arginine (Mitchell et al., 2000). When the

guanidine residues were alkylated the complex formation was hindered and uptake

diminished. A bidentate ligand like guanidine is able to form stronger ionic complexes

with a phosphate than the regular amine of lysine: as a result uptake is higher for arginine

rich residues than other charged amino acid residues (Rothbard et al., 2005).

H NHI2 -O
O N / /O
N---- 0
H
HN

Figure 1-12 Ionic complexes formed by oligoarginine of Tat with the phosphate groups in
cell membranes.

Too few arginine residues diminishes cell surface adherence while too many positively

charged guanidines lead to reduced escape from the inner leaf of the membrane. The

active drug is released from the conjugate hydrolytically with half lives depending on the

pH sensitive linker being used (Figure 1-13). Cyclosporine- Arg7 conjugates release the

active drug at pH 7.4 and 370C with a half life of 90 min. The mechanism of hydrolysis is

intramolecular nucleophilic attack by the secondary amine on the ester carbonyl

functional group (Rothbard et al., 2000). Similarly in case of Taxol- Args conjugates, the

hydrolysis rates vary depending on the R group attached to the amine from 1 min for R =









H to 107 min for R = Boc. Thus it is possible to control the rates of hydrolysis just by

varying the R group from an electron donating to an electron withdrawing group

(Kirschberg et al., 2003).

0 0
N NHArg7-COOH pH 7.4 N N NNHArg7-COOH


O N OCSA O + CSA

PhPh 0



0 pH 7.4 TAXOL + RN S

S 0 H2NCO-8grA-HN-

RHN NH-Arg,-CONH 0
R = H, Ac, Piv, Boc; CSA Cyclosporin.
Figure 1-13 Chemical hydrolysis of oligoarginine conjugates of cyclosporine and taxol.

Prodrugs for Dermal and Transdermal delivery

Although the previous paradigm (Guy and Hadgraft, 1988; Flynn, 1990) for

increasing dermal delivery focused on increasing KOCT:AQ, most effective prodrug

approaches aimed at improving the delivery of drugs across the skin have relied on

increasing the solubility in a lipid vehicle (SLIPID) like octanol or IPM and the solubility

in a polar vehicle like water (SAQ) of the prodrug over the parent drug. It is believed that

an increase in biphasic solubility of the drug leads to an increase in permeability of drugs

across the stratum corneum (Sloan et al., 1984; Sloan, 1989 and 1992; Sloan and Wasdo,

2003). In a homologous series of lipophilic prodrugs, the more water soluble of the more

lipid soluble member of the series gave the highest flux (Sloan et al., 1984, 1984, 1992,

2003 and 2004). Thus the balance between solubility in lipid and aqueous phases









becomes paramount and flux is better modeled by a combination of these two parameters

rather than K (partition coefficient). Previously it was shown that by merely masking

polar functional groups transiently it is possible to increase biphasic solubility and

thereby flux. While it is easy to see how addition of carbon atoms to the promoiety

conjugated to the drug increases SLIPID, it is difficult to visualize why a simultaneous

increase in SAQ occurs. For example, one may compare caffeine, theophylline and

theobromine to illustrate this increase in SAQ (Windholz et al., 1983). Theobromine has

two intramolecular hydrogen bonding sites; theophylline has one hydrogen bonding site

while caffeine has none. A look at the melting points and SAQ Of these three compounds

reiterates the fundamental idea behind prodrug synthesis. The masking of polar functional

groups leads to decreased intramolecular hydrogen bonding and in turn leads to a

decrease in lattice energy, melting points and increase in biphasic solubilities.

0 0 0

HN N ON N ON N
N N


H


THEOBROMINE THEOPHYLLINE CAFFEINE
mp: 357 270-274 238
Saq: 3 mmol/l 46.55 mmol/l 112.64 mmol/l
Codeine, the ether derivative of morphine, is 40 times more water soluble than

morphine and 2400 times more lipid soluble. Thus masking of a phenolic functional

group increases solubility in both phases.

There are two types of prodrug promoieties that have been used in dermal delivery;

the acyl based approach (N-acyl, O-acyl and S-acyl) and soft alkyl based approach

(ACOM or AOCOM). In the discussion that follows we will look at the importance of









solubilities in optimizing flux across the skin with examples. Consider the table below for

series of 1-alkyloxycarbonyl-5-FU prodrugs and 5-FU with their physicochemical

properties (Table 1-2). Melting points (mp), molecular weight (MW), log solubility in

isopropyl myristate (log SIPM), log solubility in water (log SAQ), log partition coefficients

between IPM and pH 4 buffer (log KIPM: AQ) and log maximum flux of total species

delivered by the prodrug (and 5-FU) through hairless mice skins in vitro from an IPM

donor phase (log JMIPM) (Beall et al., 1993) are given.

Table 1-2 Physicochemical Characterization of 1-Alkyloxycarbonyl esters of 5-FU.
Compound mp MW log SIPM log SAQ log KIPM:AQ log J MIPM
Cl 160 188 0.328 2.05 -1.72 0.42
C2 128 202 1.117 2.24 -1.12 0.77
C3 126 216 1.182 1.63 -0.45 0.36
C4 98 230 1.529 1.37 0.16 0.35
C6 67 258 2.186 0.7 1.48 0.19
C8 98 286 1.561 -0.89 2.46 -0.53
5-FU 284 130 -1.308 1.93 -3.24 -0.62

The melting points of the prodrugs decrease with the addition of each CH2 unit

from Cl to C6 while the melting point of C8 prodrug is higher than the C6 probably

because of van der Waals interaction between the alkyl side chains (Yalkowsky, 1977).

This increase in melting point also leads to a decrease in SIPM. All members of the series

are more lipid soluble than 5-FU. The C6 prodrug derivative is the most lipid soluble of

the series, has the highest K and SIPM. So, if SIPM were the most important determinant,

then its log JMIPM should be the highest instead it is the next to lowest in the series. On the

other hand, the C2 prodrug has a higher lipid solubility than 5-FU (not the highest) and

highest SAQ in the series (1.8 times more water soluble than 5-FU) and gives the highest

JMIPM (25 fold higher than 5-FU) across the skin. Sloan and coworkers (1983, 1984 and

2003) have reported such dependence on biphasic solubility for a series of 6-









mercaptopurine (6MP), ThH and 5-FU prodrugs. It was first observed for 7-ACOM

prodrugs of ThH (Sloan et al., 1982) but not recognized as a new paradigm until later

(Sloan et al., 1984).

There are two schools of thought for optimizing flux across the skin. The first relies

on increasing the lipid solubility without increasing aqueous solubility of the prodrug and

uses increased KOCT: AQ as an indicator of increased lipophicity. Since skin is primarily a

lipophilic barrier it relies on making a prodrug with higher K than the parent drug so that

penetration across the barrier can be enhanced. In this school of thought, any dependence

of flux on aqueous solubility is attributed to the use of a polar vehicle, the nature of

mouse skins (which are more hydrophilic than human skin), in vitro experimental

conditions and metabolic rates of conversion. In in vitro cell experiments, the skin is

abnormally hydrated particularly when water is used as a vehicle. Thus some argue that

the dependence of flux on SAQ could be due to the necessity for increased solubility in the

vehicle. This school of thought also argues that lipid soluble prodrugs undergoing fast

rates of hydrolysis to the more water soluble parent drug bypasses the hydrophilic region

of the skin dermis easily and permeates the skin better. However in case of lipid soluble

prodrugs undergoing slower hydrolysis, the dermis becomes a rate limiting barrier and

the prodrug gets trapped in the dermis, compromising delivery across the skin. Thus

water solubility becomes important only if the designed prodrugs are stable to hydrolysis

during their passage through the skin (Stinchcomb and coworkers, 2002 and 2005).

The second school of thought relies on the importance of biphasic solubility in

order to increase flux. It relies on increasing both lipid and water solubility at the same

time to optimize delivery. It is important to note that flux is independent of metabolic









conversion rates of prodrug to the drug because flux is dependent only on the solubility in

the membrane (Equation 1.4). The observations that flux depends on SAQ and SLIPID holds

true for prodrugs with a half-life of 1-5 min (7-AC-Th, Sloan et al., 2000) to prodrugs

which permeate intact in diffusion cells experiments (1-ACOM-FU, >50% intact prodrug

obtained, Taylor and Sloan, 1998) run on hairless mice skins. The argument against the

dependence of water solubility due to the presence of stagnant water layers in presence of

a polar vehicle or in in vitro conditions when the skin is heavily hydrated comes from

experiments carried out with isopropyl myristate as a vehicle (a non polar vehicle in

which the longer chain prodrugs were more soluble). The same dependence on SAQ was

again observed where IPM was the vehicle, reemphasizing the fact that this dependence

is a result of the inherent properties of the skin. In Vivo human skin (human skin is less

hydrophilic than mouse) experiments using mineral oil also show dependence on water

solubility. Irrespective of the vehicle used, the best performing prodrug of a series from a

vehicle like IPM is also the best performing prodrug from a different vehicle like water.

The best prodrug was the one with the best biphasic solubility. The increase in flux

though varies from vehicle to vehicle depending on the interaction of the skin with the

vehicle. IPM causes more damage than water does; it leaches out the lipids from the

stratum corneum hence compromises the barrier and leads to increased permeability of

prodrugs compared to water (Sloan et al., 2003a). Water solubility is important in

improving flux across the skin whether it is human skin or mouse skin, polar or non polar

vehicle, in vitro or in vivo, and prodrugs which are labile or stable to hydrolysis.

In the 5-FU prodrug series we considered above, the prodrug exhibiting the best

flux had a higher SAQ and SLIPID than 5-FU. It is not always possible to increase SAQ of the









prodrug compared to the parent drug. In such cases, the best performing prodrug is the

one which shows minimum decrease in SAQ with respect to the parent drug, e.g.

methyloxycarbonyl-APAP prodrug shows the highest flux through skin in the

homologous series of alkyloxycarbonyl-APAP prodrugs investigated by Wasdo and

Sloan (Table 1-3). This particular derivative was 6.3 times more lipid soluble and about

3.5 times less water soluble than APAP, yet was the most water soluble prodrug of the

series being investigated.

Table 1-3 Physicochemical Characterization of Alkyloxycarbonyl esters of
Acetaminophen.
Compound mp MW log SIPM log SAQ log KIPM:AQ log J MIPM
Cl 112 209 1.08 1.31 -0.16 0.0
C2 120 223 0.97 0.58 0.32 -0.76
C3 104 237 1.37 0.43 0.9 -0.45
C4 118 251 1.14 -0.43 1.5 -1.01
C6 108 279 1.22 -0.37 2.71 -1.49
APAP 171 151 0.28 2.0 -1.72 -0.29

Similarly in the 7-alkyloxycarbonyl-Th series (Sloan, 2000), C3-Th was most water

soluble prodrug member of the lipid soluble series. This prodrug derivative was 1.3 times

less water soluble than ThH and still gave higher flux than ThH.

Ever since SAQ of a drug was recognized as being important to optimize dermal

delivery, attempts have been made by Sloan and coworkers (Sloan et al., 1984; Sloan et

al., 1988; Saab et al., 1989 and 1990) to incorporate basic amino groups into the

promoiety. The rationale behind introducing a polar functional group like amine into a

promoiety is to increase the water solubility of poorly soluble drugs. Insertion of simple

alkyl groups into the promoiety does increase lipid solubility but increases in water

solubility are modest and are restricted to the shorter alkyl chain members. Mannich

bases (DrugX-CH2-NR2, X = S, N, 0; R = alkyl) of 6MP, 5-FU and ThH were thus









designed to improve SAQ of the prodrugs without decreasing SLIPID compared to the parent

drug and hence to increase flux. Hussain and coworkers (Milosovich et al., 1993 and Jona

et al., 1995) added polarisable amino groups to the alcohol portion of testosterone and

indomethacin ester prodrugs. When the 4-dimethylaminobutyrate hydrochloride ester of

testosterone was examined in vitro using human skin, a 35 fold flux enhancement was

observed (Figure 1-14). Similarly when 2-diethylaminoethyl group was built into the

ester promoiety of indomethacin the SAQ was enhanced compared to the parent drug and

flux through skin increased by about six fold (Figure 1-14).

0
0
0 N(CHCH,C)2
0 ^ ^N(CH,).HC1 MeO







Figure 1-14 Structure of testosterone and indomethacin esters with polarisable side
chains.

Also, attempts have been made by Bonina and coworkers (1995) to increase SAQ

and SLIPID by incorporating short chain polyethylene glycolic ethers into the promoiety.

Thus NSAIDS esters with small PEGs have been evaluated for their flux across human

skin. These polyethers did increase both SAQ and SLIPID, e.g. for the indomethacin esters

addition of a -OCH2CH2 unit increased SAQ by 1.3 fold and SLIPID by 1.02 times and as a

result flux also increased. Although all the derivatives were more lipid soluble than

indomethacin, the prodrug containing n = 5 -OCH2CH2 units, which was 1.77 times more

water soluble and 9 fold more lipid soluble than indomethacin, gave the highest flux,

(Table 1-4).





















PEG-esters of Indomethacin
Prodrug promoieties containing short PEGs can thus increase the biphasic solubility of

lipid soluble drugs with very poor aqueous solubility. It is interesting to note even in this

series of 'ethylene oxy' homologous series it is the more water soluble derivative that

gave the highest flux just as in case of simple alkyl derivatives where a 'CH2'unit is

inserted for 'O'along the series.

Thus, the importance of the balance between SLIPID and SAQ of the prodrugs is

important in optimizing delivery of drug molecules topically.

Table 1-4 Physicochemical Characterization of PEG esters of Indomethacin.
COMPOUNDS SOCT (mmol mL-x 102) SAQ JM
([tmol mL-' x102) ([mol cm-2x 102)
Indomethacin 27.7 22 0.84
n =l1 221.9 6.7 0.4
n= 2 225.8 9.2 0.65
n =3 240.1 17 0.73
n =4 244.3 25 2.83
n =5 251.8 39 3.12









Research Objectives

N-alkyl-N-alkyloxycarbonylaminomethyl (NANAOCAM) promoiety has been used to

make prodrugs of 6-mercaptopurine for topical delivery (Siver, 1990). The mechanism of

hydrolysis however wasn't clearly established. The purpose of this research project was

to explore the use of NANAOCAM promoiety as prodrugs of other functional groups like

phenols, carboxylic acids, imides and evaluate its potential use for enhancing the skin

penetration of model phenolic and imide containing drugs.

1. Design, synthesize, and investigate rates of hydrolysis and mechanism of
hydrolysis of N-alkyl-N-alkyloxycarbonylaminomethyl (NANAOCAM) and N-
aryl-N-alkyloxycarbonylaminomethyl (NArNAOCAM) prodrug derivatives of
phenols and carboxylic acid.

2. Synthesis of homologous series of NANAOCAM prodrugs of acetaminophen
(APAP, model phenolic drug) and their physicochemical characterization which
includes: measuring solubilities in isopropyl myristate (IPM) and water, partition
coefficients of prodrugs between IPM and pH 4.0 buffer, investigation of flux of
compounds through hairless mice skins from IPM.

3. Synthesis of homologous series of NANAOCAM prodrugs of theophylline (ThH,
imide containing drug) and their physicochemical characterization which includes:
measuring solubilities in isopropyl myristate (IPM) and water, partition coefficients
of prodrugs between IPM and pH 4.0 buffer, investigation of flux of compounds
through hairless mice skins from IPM.

4. Mathematical modeling of flux

These goals are discussed in three separate chapters. Chapter 2, talks about the

design of NANAOCAM promoiety, synthesis of NANAOCAM-drug derivatives by

alkylation of drug with N-alkyl-N-alkyloxycarbonylaminomethyl chlorides,

determination of rates of hydrolysis in aqueous buffers, its implication on prodrug design

leading to the synthesis and hydrolysis of N-aryl-N-alkyloxycarbonyl-aminomethyl

(NArNAOCAM) prodrugs of phenols and carboxylic acids.









Chapter deals with alkylation of acetaminophen with N-alkyl-N-

alkyloxycarbonylaminomethyl chlorides to synthesize a homologous series of

NANAOCAM prodrugs, evaluation of physicochemical properties and transdermal and

dermal penetration of these prodrugs and prediction of flux of APAP prodrugs using the

RS equation.

Chapter4 deals with alkylation of theophylline with N-alkyl-N-alkyloxycarbonyl-

aminomethyl chlorides to synthesize a homologous series of NANAOCAM prodrugs,

evaluation of physicochemical properties and transdermal and dermal penetration of these

prodrugs and prediction of flux of ThH prodrugs using the RS equation, and all

NANAOCAM prodrugs when fitted to the existing n = 63 database.














CHAPTER 2
DESIGN, SYNTHESIS, HYDROLYSIS OF N-ALKYL-N-ALKYLOXYCARBONYL-
AMINOMETHYL DRUG DERIVATIVES AND ITS IMPLICATIONS ON PRODRUG
DESIGN.

Introduction

Drugs containing polar functional groups pose problems of membrane

permeability, solubility and premature metabolism which limit their oral and dermal

delivery. The prodrug approach, which involves masking these polar functional groups as

labile derivatives which then hydrolyze to the native drug either enzymatically or

chemically, has proved useful in numerous cases (Bundgard, 1991 and Sloan and Wasdo,

2003). In most cases drug functional groups are masked as simple esters. Acyloxymethyl

(ACOM, R'COOCH2-) and alkyloxycarbonyloxymethyl (AOCOM, R'OCOOCH2-)

promoieties have been used to derivatize carboxylic acids in cases where the simple ester

approach wasn't useful because, although simple esters were reasonably stable

chemically so that they could be conveniently formulated, they were not sufficiently

labile enzymatically (Jansen and Russell, 1965; Bundgard et al., 1986; Seki et al., 1988

and Beaumont et al., 2003). Similarly, simple alkylation of phenolic groups with alkyl

halides gives derivatives that are not efficiently reversible in vivo, while ACOM or

AOCOM gives derivatives that are too unstable during the oral absorption process to

protect the phenolic drug from premature metabolism. Replacing the oxygen atom in

OCH2 with nitrogen (N-R, R = alkyl) in R'OCOOCH2- to give a N-alkyl-N-

alkyloxycarbonylaminomethyl (NANAOCAM, R'OCONRCH2-) promoiety could give

medicinal chemists an additional handle and flexibility to improve solubility (better











balance between solubilities in lipid and water) and stability (enzymatic versus chemical)


of prodrugs. The evolution of NANAOCAM promoiety from simple esters to


ACOM/AOCOM to NANAOCAM is shown in Figure 2-1.



DRUG-X R Insert ACOM DRUG CONJUGATE
DRUG-X DRUG-X O R

O 'CH20' Spacer
Insert R- ALKYL




DRUG-X O
R 'CH pacer DRUG-X O olR AOCOM DRUG CONJUGATE
O X O, S, N
ESTER OR CARBONATE-DRUG CONJUGATE
R- ALKYL
Replace 'O' with 'N'

H
DRUG-X ,-N fO R

0

alkylate N

R'

DRUG-X N O



N-ALKYL-N-ALKOXYCARBONYLAMINOMETHYL DRUG CONJUGATE (NANAOCAM)
X O, S, N
R, R' ALKYL
Figure 2-1 Design of NANAOCAM promoiety.


Only one example of the use of a NANAOCAM promoiety has been reported: 6-


mercaptopurine (Siver and Sloan, 1989). The use of a close analogue of NANAOCAM


(R'CONRCH2-) in which the promoiety is an amide instead of a carbamate has been


reported for carboxylic acids (Bundgard et al., 1991; Moreira et al., 1996 and Iley et al.,


1997). Here we extend the use of the NANAOCAM promoiety to other drug functional


groups and investigate the mechanism of chemical hydrolysis and its implications in









prodrug design of these soft alkylated derivatives with the goal of being able to tailor the

promoiety to give the appropriate rates of conversion to the parent drug based on the

route of administration.

In the discussion that follows we will talk about the synthesis ofN-methyl-N-

methyloxycarbonylaminomethyl derivatives of phenols, carboxylic acids and 6-

mercaptopurine (6MP) and N-aryl-N-alkyloxycarbonylaminomethyl (NArNAOCAM)

derivatives of phenols and carboxylic acids. We will then investigate the mechanism of

hydrolysis of N-methyl-N-methyloxycarbonylaminomethyl and NArNAOCAM

derivatives of phenols, carboxylic acids and 6MP.

Synthesis of N-methyl-N-methyloxycarbonylaminomethyl and N-aryl-N-methyloxy-
carbonylaminomethyl Prodrugs.

Ten N-methyl-N-methyloxycarbonylaminomethyl and six N-aryl-N-

methyloxycarbonyl aminomethyl derivatives were synthesized (Tables 2-1 and 2-2).

Alkylation of the parent compound with N-methyl-N-methyloxycarbonylaminomethyl

chloride or N-aryl-N-methyloxycarbonylaminomethyl chloride was accomplished in the

presence of a base like triethylamine with CH2Cl2 or DMSO as the solvent. In every case

it was necessary to synthesize the corresponding alkylating agent: N-methyl-N-

alkoxycarbonylaminomethyl chloride or N-aryl-N-methyloxycarbonylaminomethyl

chloride.









Conjugates Synthesized.


Compd X Z R
1 NHCOMe H Me
2 CN H Me
3 CHO H Me
4 H CHO Me
5 COMe H Me
6 COOMe H Me
7 NO2 H Me
8 NO2 H 4'-C6H4-OMe
9 NO2 H 4'-C6H4-COOEt
10 NO2 H C6H5


Table 2-2 N-methyl-N-methyloxycarbonylaminomethyl and NArNAOCAM Conjugates
of Naproxen.


.0 NY CH3

0


Compd R

11 Me

12 C6H5

13 4'-C6H4-OMe

14 4'-C6H4-COOEt


Table


X-









N




0

0 0 N O0
R
R = R' = CH3


0



NOZ N

N N
H


R = R'= CH3


Figure 2-2 Structures of N-methyl-N-methyloxycarbonylaminomethyl prodrugs of
dimethylaminobenzoic acid and 6MP.


CH2=0 +

R


rN +
RN N
R 'R



0

ROJ N Cl
I
R


RNH2


NaOH N

rN N
~~ i


0

O/R'
CH2C12



CH2 Cl2
+ Y-H A*-
TEA


0

\R N Cl
R


0

N Y
R
Y = 6MP,phenol or
carboxylic acid
R, R' = CH3
Compounds 1-7, 11, 15, 16


Figure 2-3 Synthesis of NANAOCAM prodrugs









U 0

2 HN O C1 N O 3
Py TMSC1

CH3OCOC1
(CH2O)n
XX X
0 0

Cl1-1\ N O"CH 3 Naproxen Y1 N '0CH3
or

p-nitrophenol
TEA
X X


X= OCH3, COOC2H,, H.
YH = Naproxen, p-nitrophenol.
Compounds 8-10, 12-14.
Figure 2-4 Synthesis of NArNAOCAM prodrugs

N-Methyl-N-methyloxycarbonylaminomethyl chloride synthesis (Figure 2-3):

The N-methyl-N-methyloxycarbonylaminomethyl chloride was synthesized as

reported by Siver et al., (1990) from 1, 3, 5-trimethylhexahydrotriazine.

(a) 1, 3, 5-Trimethylhexahydrotriazine was synthesized from equimolar equivalents of

aqueous formaldehyde, methyl amine and NaOH according to the protocol originally

developed by Graymore et al., (1932) and modified by Siver et al., (1990). Methyl amine

(0.4 mol, 40% aqueous) was placed in an ice bath and an equivalent of 37% aqueous

formaldehyde was added dropwise over a period of 10 min. The solution was allowed to

equilibrate to room temperature and stirred for one hour, then an equivalent of NaOH was

added and the contents were stirred for 1.5 h more. The solution was extracted with 4 x

50 mL CH2C12, dried over Na2SO4, filtered and concentrated to a clear colorless CH2Cl2

solution containing the hexahydrotriazine derivative. Complete concentration of CH2C12









wasn't carried out as it resulted in some loss of the desired product. For quantification

purposes the CH2C2 'methylene peak'at 65.3 and N-CH2-N peak of hexahydrotriazine

derivative at 63.2 were used: yield = 82% in CH2C2, 1H NMR (400 MHz; CDCl3;

Me4Si):63.2(s, 6H), 62.3(s, 9H).

(b) N-Methyl-N-methyloxycarbonylaminomethyl chlorides were synthesized from 1, 3,

5-trimethylhexahydrotriazine by reacting it with three equivalents of methyl

chloroformate in CH2C2 .To well stirred solution of methylchloroformate in CH2C2

cooled with an icebath was added an equivalent of 1, 3, 5-trimethylhexahydro-triazine

(freshly prepared) in CH2C2 over a period of 10 minutes. The white suspension that was

observed was allowed to equlilibrate to room temperature and stirred overnight. The

suspension was filtered and the filtrate concentrated to oil. The oil contained the desired

product and some of the corresponding bis (N-methyl-N-methyloxycarbonyl-

amino)methane byproduct. The oils were purified by trituration with hexane overnight

followed by ether overnight. The clear solution was decanted leaving the white residue

(bis derivative) behind. The clear solution was then concentrated. The peak at 64.8 due to

the 'CH2 of the bis derivative and peak at 65.3 due to the 'CH2' of N-methyl-N-methyl-

oxycarbonylaminomethyl chloride were used to quantitate the amount of product formed:

yield = 90%, 1H NMR (400 MHz; CDCl3; Me4Si):65.31-5.33 (2s, 2H), 63.73-3.79 (2s,

3H), 62.9-3.0 (2s, 3H).

N-Aryl-N-methyloxycarbonylaminomethyl chloride synthesis (Figure 2-4):

N-Aryl-N-methoxycarbonylaminomethyl chlorides were made from aromatic amines in

two steps. Methylchloroformate was reacted with aromatic amines in the presence of

pyridine and dichloromethane to give N-aryl carbamic acid methyl esters. The

chloromethyl derivative of the carbamic acid derivative was then made by a synthetic









procedure reported by Moreira et al., (1994) for N-methylamides. The N-Methyl

carbamic acid alkyl ester was refluxed with thirteen equivalents of trimethylsilyl chloride

and 1.7 equivants of paraformaldehyde to give the appropriate N-aryl-N-alkoxycarbonyl-

aminomethyl chloride.

(a) N-Aryl carbamic acid methyl ester: To a solution of methylchloroformate (3.5 mmol)

in 15 mL CH2C2 was added dropwise an equivalent of pyridine and aromatic amine. The

reaction mixture was allowed to warm to room temperature and subsequently stirred

overnight. The clear solution was washed with 10 mL brine 3 times, and the organic layer

was dried over Na2SO4 and concentrated to a solid. The solids obtained were

recrystallized from CH2C2: hexane.

N- (Phenyl)carbamic acid methyl ester: yield = 98 %, mp = 44-450C, 1H-NMR (400

MHz; CDCl3; Me4Si): 67.2-7.37(m, 4H), 67.06 (m, 1H), 66.63(s, 1H), 63.76 (s, 3H).

N-(4'-Ethoxycarbonylphenyl)carbamic acid methyl ester: yield = 97 %, mp = 152-1540C,

1H-NMR (400 MHz; CDCl3; Me4Si): 68.0 (d, 2H), 67.45 (d, 2H), 66.9(s, 1H), 64.37 (q,

2H), 63.8 (s, 3H), 61.39 (t, 3H).

N-(4'-Methoxyphenyl)carbamic acid methyl ester: yield = 94 %, mp = 85-860C, 1H-

NMR (400 MHz; CDCl3; Me4Si): 67.27 (d, 2H), 66.85 (d, 2H), 6 3.78(s, 3H), 6 3.75(s,

3H).

(b) N-Aryl-N-methyloxycarbonylaminomethyl chloride: A suspension of N-aryl

carbamic acid alkyl ester (2.5 mmol), 1.7 equivalents of paraformaldehyde and 13

equivalent of trimethylsilyl chloride was refluxed using a CaCl2 drying tube and a water

condenser for 18 h over an oil bath. The suspension was diluted with -10 mL CH2C2 and

filtered to get rid of the unreacted paraformaldehyde. The clear filterate was concentrated









using a rotavapor at 400C under reduced pressure. The yellow oil obtained was triturated

with hexane overnight. The white suspension obtained was filtered and the filtrate was

concentrated to give the desired alkylating agent.

N- (Phenyl)-N-methyloxycarbonylaminomethyl chloride: yield = 75%, 1H NMR (400

MHz; CDCl3; Me4Si): 67.32-7.44 (m, 5H), 65.56(s, 1H), 63.78 (s, 3H).

N- (4'- Ethyloxycarbonylphenyl)-N-methyloxycarbonylaminomethyl chloride: Complete

conversion to product wasn't seen in this case, the crude mixture was thus used in the

next step. The peak at 66.9 due to the 'NH of the N-(4'-ethyloxycarbonylphenyl)

carbamic acid methyl ester and the peak at 65.56 due to the 'CH2' of N-(4'-

ethyloxycarbonylphenyl)-N-methyloxycarbonylaminomethyl chloride were used to

quantify the amount of product formed: yield = 27%, 1H NMR (400 MHz; CDCl3;

Me4Si): 68.1 (d, 2H), 67.5 (d, 2H), 65.56(s, 1H), 64.38 (q, 2H), 63.81 (s, 3H), 61.39 (t,

3H).

N-(4'-methoxyphenyl)-N-methyloxycarbonylaminomethyl chloride: yield = 82%, 1H

NMR (400 MHz; CDCl3; Me4Si): 67.23 (d, 2H), 66.93 (d, 2H), 65.52(s, 1H), 63.82 (s,

3H), 63.76 (s, 3H).

Alkylation of Phenols, 6-Mercaptopurine, Dimethylaminobenzoic acid and
Naproxen with N-methyl-N-methyloxycarbonylaminomethyl chloride or N-aryl-
N-methyloxycarbonylaminomethyl chloride (Figures 2-3 and 2-4).

Typical procedure for phenols: In a round bottom flask was dissolved 1 equivalent of

phenol (-0.01 mol) and 1.1 equivalents triethylamine in 20 mL CH2Cl2. The contents

were stirred for an hour under reflux conditions with a water condenser and oil bath. N-

Methyl-N-alkoxycarbonyl aminomethyl chloride or N-aryl-N-methyloxycarbonyl-

aminomethyl chloride (1.1 equivalents) in 5 mL CH2Cl2 was then added dropwise. An

exothermic reaction occurred and white fumes could be seen. The clear solution or









suspension (depending on the phenol used) was stirred overnight. The reaction was

worked up by filtering the suspension [NEt3 C:, H- NMR(CDCl3): 63.1 (q,2H), 61.5

(t,3H)], diluting the filtrate to 40 mL with CH2C2 and washing it with 3 x 30 mL water.

The CH2C12 solution was dried over Na2SO4 for an hour and filtered. The solution was

concentrated using a rotavapor under vacuum at 400C until solvent free. The resulting

material was purified by recrystallization and, if necessary, column chromatography until

a sharp melting point was obtained, a single spot was seen on TLC and a clean 1H- NMR

was obtained. The particular results for each synthesis are listed below.

1 was prepared from N-methyl-N-methyloxycarbonylaminomethyl chloride,

triethylamine and acetaminophen in CH2Cl2. Recrystallization from CH2Cl2: hexane (1:3)

twice gave white crystals: yield = 70%, mp = 86-880C, Rf (0.36, ether). Elemental

analysis (Found: C, 56.81; H, 6.34; N, 11.1. Calc. for C12H16N204 : C, 57.13; H, 6.39; N,

11.1%). UV: Xmax (pH 8.8 buffer)/nm 243.4 (E/LmolEcm10.95 x 10 4). UV: Xmax (pH

7.1 buffer)/nm 240 (E/LmolEcm"1 1.01 x 10 4, 0.09 x 10 4).1H NMR(400 MHz; CDCl3;

Me4Si): 67.6(s,1H), 6 7.39(d,2H), 66.96-6.87(2d,2H), 65.28-5.21(2s,2H), 63.72-

3.7(2s,3H), 6 3.0-2.97(2s,3H), 6 2.0(s,3H).

2 was prepared from N-methyl-N-methyloxycarbonylaminomethyl chloride,

triethylamineand p-cyanophenol in CH2C2. Recrystallization from ethyl acetate:hexane

(1:2) twice gave white crystals: yield = 55%, mp = 56-580C, Rf (0.17, ethyl

acetate:hexane, 1:4). Elemental analysis (Found: C, 59.54; H, 7.27; N, 9.94. Calc. for

C11H12N203: C, 60.02; H, 7.14; N, 10%). UV: Xmax (pH 8.8 buffer)/nm 245 (E/Lmol"

1cm1 1.26 x 10 4).1H NMR (400 MHz; CDCl3; Me4Si):6 7.6(d, 2H), 6 7.1-6.96(2d, 2H), 6

5.38-5.31(2s, 2H), 6 3.8(s, 3H), 6 3.06-3.0(2s, 3H).









3 was prepared from N-methyl-N-methyloxycarbonylaminomethyl chloride,

triethylamine and p-hydroxybenzaldehyde in CH2C12. Purification on a silica gel column

with hexane: acetone (4:1) as the eluent gave a colorless oil: yield = 67%, Rf (0.37, ethyl

acetate:hexane, 1:2). Elemental analysis (Found: C, 59.08; H, 5.86; N, 6.3 Calc. for

ClnH13N04: C, 59.19; H, 5.87; N, 6.27%).UV: Xmax (pH 8.8 buffer)/nm 276.6

(E/L mol 1cm1 1.37 x 10 4).1H NMR(400 MHz; CDCl3; Me4Si): 69.9(s,1H),6 7.83(d,2H),

6 7.1-7.03(2d,2H),65.42-5.35(2s,2H), 6 3.8(s,3H), 6 3.07-3.01(2s,3H).

4 was prepared from N-methyl-N-methyloxycarbonylaminomethyl chloride,

triethylamine and salicyldehyde in CH2C2. Recrystallization from acetone gave white

crystals: yield = 87%, mp = 65-670C, Rf(0.76, ether).Elemental analysis (Found: C,

59.06; H, 6.01; N, 6.24 Calc. for CjlH13N04: C, 59.19; H, 5.87; N, 6.27%). UV: Xmax

(pH 8.8 buffer)/nm 315 (E/LmolPcm11.27 x 10 4). 1H NMR(400 MHz; CDCl3; Me4Si): 6

10.49(s,1H), 6 7.85(d,1H), 6 7.55(t,1H) 6 7.2(m,2H), 6 5.47-5.38(2s,2H), 63.76(s,3H),

63.08-3.03(2s,3H).

5 was prepared from N-methyl-N-methyloxycarbonylaminomethyl chloride,

triethylamine and p-hydroxyacetophenone in CH2C2. Purification on a silica gel column

with hexane: acetone (4:1) as the eluent gave a colorless oil: yield = 82%, Rf (0.75,

ether). Elemental analysis (Found: C, 60.45; H, 5.91; N, 6.22 Calc. for CjlH13N04: C,

60.75; H, 6.37; N, 5.9%). UV: (pH 8.8 buffer)/nm 269.7(E/LmolPlcm-1 1.12 x 10 4).1H

NMR(400 MHz; CDCl3; Me4Si): 67.93(d,2H), 6 7.05-6.94(2d,2H), 6 5.39-5.32(2s,2H), 6

3.8(s,3H), 6 3.05-3.0(2s,3H), 6 2.6(s,3H).

6 was prepared from N-methyl-N-methyloxycarbonylaminomethyl chloride,

triethylamine and methyl p-hydroxybenzoate in CH2C2. Purification on a silica gel









column with ethyl acetate: hexane as eluent gave a colorless oil. This oil, after

recrystalization from CH2C2 :hexane (1:2), gave white crystals: yield = 21%, mp = 52-

540C, Rf(0.49, ethyl acetate:hexane,1:4).Elemental analysis (Found: C, 56.98; H, 5.93;

N, 5.49.Calc. for C12H15N05: C, 56.91; H, 5.97; N, 5.53%). UV: Xmax (pH 8.8

buffer)/nm 253 (E/Lmol-1 cm-1 1.49 x 10 4). 1H NMR(400 MHz; CDCl3; Me4Si): 6

8.0(d,2H), 6 7.02-6.92(2d,2H), 6 5.37-5.3(2s,2H), 6 3.9(s,3H), 6 3.8(s,3H), 6 3.05-

2.99(2s,3H).

7 was prepared from N-methyl-N-methyloxycarbonylaminomethyl chloride,

triethylamine and p-nitrophenol in CH2C2. Recrystallization from CH2C2 :hexane (1:4)

gave pale yellow crystals: yield = 76%, mp = 77-790C, Rf (0.37,ether: hexane,

1:1).Elemental analysis(Found: C, 49.99; H, 4.94; N, 11.57.Calc. for CloH12N205: C,

50.0; H, 5.04; N, 11.66%). UV: Xmax (pH 8.8 buffer)/nm 310.6 (E/Lmolfcm10.76 x 104).

UV: Xmax (pH 7.1 buffer)/nm 310 (E/Lmolfcm-1 0.75 x 10 4). 1H NMR (400 MHz;

CDCl3; Me4Si): 6 8.21(d, 2H), 6 7.1-6.99(2d, 2H), 6 5.42-5.35(2s, 2H), 6 3.77(s, 3H),6

3.07-3.01(2s, 3H).

8 was prepared from N-(4'-methoxyphenyl)-N-methyloxycarbonylaminomethyl

chloride, triethylamine and p-nitrophenol in CH2C2. Purification on a silica gel column

with ethyl acetate: hexane as eluent gave a colorless oil: yield = 64%, Rf (0.17, ethyl

acetate:hexane, 1:5), UV: Xmax (pH 8.8 buffer)/nm 310 nm (E/Lmol cm-1 1.06 x 10 4).

1H NMR(400 MHz; CDCl3; Me4Si):6 8.2(d,2H), 6 7.16(d,2H), 6 7.05(d,2H), 6

6.89(d,2H), 6 5.64(s,2H), 6 3.81(s,3H), 6 3.73(s,3H).









9 was prepared from N-(4'-ethyloxycarbonylphenyl)-N-methyloxycarbonylaminomethyl

chloride, triethylamine and p-nitrophenol in CH2C2. A white solid residue was seen. This

residue was dissolved in CH2C2 and hexane was added till the solution became turbid

and white crystals were seen. The white crystals were filtered and discarded while the

filterate was reconcentrated to give more white crystals which were filtered. The filterate

was purified on a silica gel column with ethyl acetate: hexane as eluent to give a colorless

oil: yield = 17%, Rf (0.29, ethyl acetate:hexane, 1:3). UV: Xmax (pH 8.8 buffer)/nm 309

(E/Lmolfcm-1 1.03 x 10 4). 1H NMR(400 MHz; CDCl3; Me4Si): 6 8.21(d,2H),

68.07(d,2H), 6 7.37(d,2H), 6 7.05(d,2H), 65.69(s,2H), 6 4.38(q,2H), 63.77(s,3H), 6

1.39(t,3H).

10 was prepared from N-(phenyl)-N-methyloxycarbonylaminomethyl chloride,

triethylamine and p-nitrophenol in CH2C2 Purification on a silica gel column with ethyl

acetate: hexane as eluent gave a colorless oil which, on recrystallization from ether:

petroleum ether (1:2), gave white crystals: yield = 69%, mp = 105-1060C, Rf (0.26, ethyl

acetate: hexane, 1:9). Elemental analysis (Found: C, 59.49; H, 4.63; N, 9.21. Calc. for

C15H14N205: C, 59.6; H, 4.67; N, 9.27%). UV: Xmax (pH 8.8 buffer)/nm 305 nm

(E/Lmolfcm-1 1.19 x 10 4).1H NMR (400 MHz; CDCl3; Me4Si): 6 8.2(d, 2H), 6 7.27-

7.42(m, 5H), 6 7.05(d, 2H), 6 5.67(s, 2H), 6 3.75(s, 3H).

11 was prepared from naproxen (0.01 mol) stirred with triethylamine (O.Olmol) in 50 mL

CH2C2 for an hour. A clear colorless solution was seen. N-Methyl-N-methyloxy-

carbonylaminomethyl chloride (O.Olmol) was added to the reaction mixture and the

contents were stirred overnight. The clear solution was washed 3-times with 10 mL

water. The CH2C2 solution was dried over Na2SO4 for an hour and then evaporated









under reduced pressure to give white crystals: yield = 97%, mp = 100-1010C, Rf (0.49,

ethyl acetate:hexane,1:4). Elemental analysis (Found: C, 65.05; H, 6.45; N, 4.2 Calc. for

Ci8H21N05: C, 65.24; H, 6.39; N, 4.23%). UV: Xmax (pH 7.1 buffer)/nm 271.5 and 245

(E/Lmolfcm-1 0.52 x 10 4, 0.77 x10 4 L/mole.1H NMR(400 MHz;CDCl3;Me4Si):

67.69(t,3H), 67.4(d,lH), 67.11-67.15(m,2H), 6 5.41-5.29(dd,2H), 63.91(s,3H),

63.85(q,1H), 63.7-3.65(2s,3H), 62.89-2.85(2s,3H), 61.57(d,3H).

12 was prepared from naproxen (0.001 mol) stirred with triethylamine (0.001mol) in 50

mL CH2C2 for an hour. A clear colorless solution was seen. N-(Phenyl)-N-methyloxy-

carbonylaminomethyl chloride (0.001mol) was added to the reaction mixture and the

contents were stirred overnight. The clear solution was washed 3-times with 10 mL

water. The CH2C2 solution was dried over Na2SO4 for an hour and then evaporated

under reduced pressure to give a yellow oil. Purification of the oil on a silica gel column

with ethyl acetate: hexane as eluent gave a colorless oil: yield = 60%, Rf (0.17, ethyl

acetate:hexane, 1:5), UV: Xmax (pH 7.1 buffer)/nm 272.8 and 245, (E/Lmolfcm10.35 x

10 4, 0.83 x10 4). 1H NMR(400 MHz;CDCl3;Me4Si): 67.69(t,3H), 67.36-7.39(dd,1H),

67.13-67.17(m,5H), 66.9(d,2H), 6 5.5-5.68(dd,2H), 63.93(s,3H), 63.88(q,lH),

63.65(s,3H), 61.57(d,3H).

13 was prepared from naproxen (0.001 mol) stirred with triethylamine (0.001mol) in 50

mL CH2Cl2 for an hour. A clear colorless solution was seen. N-(4'-Methoxyphenyl)-N-

methyloxycarbonylaminomethyl chloride (0.001mol) was added to the reaction mixture

and the contents stirred overnight. The clear solution was washed 3-times with 10 ml

water. The CH2C2 solution was dried over Na2SO4 for an hour and then evaporated

under reduced pressure to give to give a yellow oil. Purification of the oil on a silica gel









column with ethyl acetate: hexane as eluent gave a colorless oil: yield = 70%, Rf (0.1,

ethyl acetate:hexane, 1:5), UV: kmax (pH 7.1 buffer)/nm 271.5 and 245, (E/Lmolfcm

10.26 x 10 4, 0.91 x10 4 ). 1H NMR(400 MHz;CDCl3;Me4Si):67.69(t,3H), 67.36-

7.39(dd,lH), 67.13-67.17(m,2H), 67.69(t,3H), 67.37(dd,lH), 66.8(d,2H), 66.5(d,2H),6

5.5-5.64(dd,2H), 63.92(s,3H), 63.88(q,lH), 63.68(s,3H), 63.65(s,3H), 61.57(d,3H).

14 was prepared from naproxen (0.001 mol) stirred with triethylamine (0.001mol) in 50

mL CH2C12 for an hour. A clear colorless solution was seen. N-(4'-Ethyloxycarbonyl-

phenyl)-N-methyloxycarbonylaminomethyl chloride (0.001mol) was added to the

reaction mixture and the contents were stirred overnight. The clear solution was washed

3-times with 10 mL water. The CH2Cl2 solution was dried over Na2SO4 for an hour and

then evaporated under reduced pressure to give white crystals. This residue was dissolved

in CH2Cl2 and hexane was added till the solution became turbid and white crystals were

seen. The white crystals were filtered and discarded while the filterate was reconcentrated

to give more white crystals. After purification on a silica gel column with ethyl acetate:

hexane as eluent the filtrate gave a colorless oil: yield = 13%, Rf (0.25, ethyl acetate:

hexane, 1:5), UV: Xmax (pH 7.1 buffer)/nm 270 and 245 (E/Lmolfcm10.25 x 10 4, 0.95

x10 4. UV: Xmax (pH 4 buffer)/nm 245(E/Lmolflcm-10.88 x10 4). UV: Xmax (pH 6

buffer)/nm 245 (E/Lmolfcm-1 0.9 x10 4). UV: Xmax (pH 8.25 buffer)/nm 245 (E/L

mol 1cm1 0.92 x10 4). UV: Xmax (pH 9.2 buffer)/nm 245 (E/LmolPcm10.94 x10 4).

1H NMR(400 MHz;CDCl3;Me4Si): 67.76(d,2H), 67.63-7.71(m,3H),67.34-7.37(dd, 1H),

67.13-7.17(m,2H), 66.98(d,2H),6 5.5-5.7(dd,2H), 64.3(q,2H), 63.94(s,3H), 63.88(q,lH),

63.66(s,3H), 61.57(d,3H), 61.38(t,3H).

15 was prepared from dimethylaminobenzoic acid (0.01 mol) stirred with triethylamine









(O.Olmol) in 50 mL CH2C2 for an hour. To the white suspension that formed was added

N-methyl-N-methyloxycarbonylaminomethyl chloride (0.01 mol) and the contents were

stirred overnight. The clear solution was washed 3-times with 10 mL water. The CH2C2

solution was dried over Na2SO4 for an hour and then evaporated under reduced pressure

to give white crystals: yield = 97%, mp = 100-1010C, Rf (0.49, ethyl acetate: hexane,

1:4), Elemental analysis: (Found: C, 57.73; H, 7.02; N, 10.6. Calc. for C13H18N204: C,

57.64; H, 6.81; N, 10.52 %). UV: Xmax (pH 8.8 buffer)/nm 315 and 287 (E/Lmol-1

cm1 1.41 x 10 4, 1.15 x10 4). 1H NMR(400 MHz;CDCl3;Me4Si): 6 7.9(d,2H), 6

6.6(d,2H), 6 5.58-5.54(2s,2H), 63.76-3.72(2s,3H), 63.05(s,9H).

16 was synthesized using the protocol developed by Siver et al. To 0.01 mol 6MP in 15

mL DMSO was added 1.1 equivalents of N-methyl-N-methyloxycarbonylaminomethyl

chloride. The yellow solution was stirred at room temperature for 1.5 h. Triethyl amine,

2.5 equivalents and 45 mL CHCl3 was added and stirring continued for one more hour.

The reaction mixture was then diluted with 100 mL CH2C2 and washed with 10 ml IN

HC1, 20 ml saturated NaHCO3 solution and 3 x50 mL brine solution. The organic layer

was dried over Na2SO4 and concentrated to give a yellow oil. The oil was triturated with

hexane to give yellow solids which were further recrystallised with CH2C2: hexane (1:5)

to give yellow crystals: yield = 62%, mp = 162-1640C (lit mp = 1630C), Rf= 0.46(10:3,

ether: methanol), Elemental analysis (Found: C, 42.9; H, 4.08; N, 27.78. Calc. for

C9H11N502S: C, 42.68; H, 4.38; N, 27.65 %). UV: Xmax (pH 7.1 buffer)/nm 288.4

(E/Lmolfcm-1 0.71x 10 4). UV: Xmax (pH 8.8 buffer)/nm 310.2 (E/Lmolfcm10.55 x10 4)

.UV: (MeOH)/nm 286 (E/ Lmolflcm 10.41 x 10 4). 1H NMR (400 MHz;CDCl3;Me4Si): 6

8.71(s,1H), 6 8.1-8.37(2s,1H), 6 5.68(s,2H), 63.75-3.85 (2s, 3H), 63.13-3.09(2s,3H).









Hydrolysis Studies.

S6-(N-Methyl-N-methyloxycarbonyl) aminomethyl- 6MP was synthesized by Siver

et al. This derivative reverted to 6MP with a half life of 91 minutes in pH 7.1 buffer at

320 C. An SN1 mechanism of hydrolysis was proposed but not firmly established. We

decided to probe the reactivity of NANAOCAM-phenol conjugates with acetaminophen

(APAP) acting as our model phenolic drug.

Since NANAOCAM-APAP type derivatives were designed as a hydrolytically

labile prodrugs, their rates of hydrolysis in aqueous buffers were investigated. Hydrolysis

of NANAOCAM derivatives of APAP at 320 pH and 7.1 buffer were much slower than

those of 6MP. The rate of hydrolysis of_N-Methyl-N-methoxycarbonylaminomethyl-

APAP (1) at pH 8.8 and 460 C was found to be 60 h. so all rates of hydrolysis were

determined at 46 'C and pH 8.8 buffer S6-(N-Methyl-N-methyloxycarbonyl)

aminomethyl- 6MP (16), resynthesized according to Siver et al. (1990), had a half life of

19 minutes under the same conditions. The anion of APAP (pKa-9.5) is not a good

leaving group and as a result the hydrolysis of its prodrug is slow compared to 6MP

(pKa-7.5). On the other hand, the NANAOCAM derivative of p-nitrophenol (7, Tables 2-

1 and 2-3) had a half life of 22 min at 460C and pH 8.8 buffer. The hydrolysis rate was

higher and t 1/2 lower since the pKa of the leaving group was much lower:7.4. Thus N-

Methyl-N-alkyloxycarbonylaminomethyl derivatives of phenols with an electron

withdrawing group like -NO2 on the ring hydrolyzed faster than those with an electron

releasing group on the ring like -NHCOCH3.

In order to establish the mechanism of hydrolysis and probe the effect of electron

withdrawing groups on hydrolysis rates ofNANAOCAM-phenol conjugates, five more

phenols (2-6, Table 2-1) with an electron withdrawing group on the aromatic ring were









synthesized using the synthetic protocol developed for APAP and rates of hydrolysis

were determined at pH 8.8 and 460 C (Table 2-3). To further establish the effect of pKa

on rates of hydrolysis, the N-Methyl-N-methyloxycarbonylaminomethyl ester of

dimethylaminobenzoic acid (15, Figures 2-1 and 2-3) was also synthesized and its rate of

hydrolysis determined.Thus, rates of hydrolysis were determined where YH = 6MP,

phenol or carboxylic acid and the corresponding anions acted as the leaving group

(Figure 2-5).



-+ OH,
Y OR' Y CH2 OR HO OR'

R R
+
SYH

I-CH2=0


0
H
\N OR'

R
Figure 2-5 Hydrolysis of NANAOCAM-Y in aqueous buffers.

Rates of hydrolysis of NANAOCAM-Y were dependent on the acidity (leaving

group ability) of the parent compound. The more acidic YH was, the faster the rate was.

When YH was cyanophenol with a pKa of 7.95, the half life was 149 minutes. Increasing

the pKa by 0.1 unit as in the case of the methyl paraben derivative, 5 raised the half life to

189 minutes. This sensitivity to small changes in pKa of YH derivative suggests that Y-

was acting as the nucleofuge by either a SN2 or SNI mechanism. The effect of election

withdrawing groups in the para position of the phenol on rates hydrolysis was also









quantified using the Hammett plots between cy substituents versus log K. (Table 2-3).

Plot of log K and sigma (o ) were linear and indicated that rates of hydrolysis were

dependent on the electron withdrawing effect imparted by the substituent at the para

position (Figure 2-6).

Table 2-3 Correlation of Rates of Hydrolysis of NANAOCAM-Y with pKa and Sigma
Values of the Leaving Group (Y).
Compd Y log k (sec-1) t 1/2(min.) pKa o


15 p-Me2N-C6H4COO- -1.30 0.22 5.03
4 o-OHC-C6H40- -3.03 12 6.79
16 6MP -3.22 19 7.5
7 p-02N-C6H40- -3.28 22 7.14 1.25
3 p-OHC-C6H40- -3.81 75 7.66 0.94
2 p-NC-C6H40- -4.11 149 7.95 0.99
5 p-MeOC-C6H40- -4.21 189 8.05 0.82
6 p-MeOOC-C6H40- -4.39 290 8.47 0.74
1 p-MeOCHN-C6H40- -5.49 3600 9.5 0.19

When log k (rate constant) was plotted against the pKa of the leaving group (Y'), a

negative correlation was found with a r2 of 0.98 (Figure 2-7). Rate of hydrolysis were

also independent of the pH of the buffer. When rates were studied in various buffers (pH

4.6, 7.1, 8.8) at 460 C for 7 and log K was plotted against buffer pH a straight line with

zero slope was obtained (Table 2-4, Figure 2-8). The pH rate profile of 14 will be

discussed later.

Table 2-4 Effect of pH of Buffer on Rates of Hydrolysis of 7
pH t1/2 (min) log k
4.6 22.83 -3.29588
7.1 21.83 -3.27655
8.8 22.83 -3.29596







62



y = 2.0288x 5.889
R2 = 0.968


* -N02
-CHO
*-CN
* -COCH3
-COOMe
* -NHCOCH3


Sigma


Figure 2-6 Pseudo-first order rate constants (sec-1) versus a- of parent phenol.
15


y = -0.9325x + 3.4017
R2 = 0.9828


0-4
0

-4


pKa


Figure 2-7 Pseudo-first order rate constants (sec-1) versus acidic pKa of parent.


-2

0 -3

-4

-5

-6


. 16











-3
-3.2 -
-3.4
-3.6
-3.8

0 -4 14
-1 -4.2 -

-4.4
-4.6
-4.8
-5

3.5 5.5 7.5 9.5

pH

Figure 2-8 pH rate profiles of compds 7 and 14

Several mechanisms for base catalysed hydrolysis of NANAOCAM-Y can be

proposed (see above).Figure 2-9A shows an SN2 type of pathway where hydroxide ion

acts as a nucleophile and displaces Y- to give a hydroxymethyl-N-alkylcarbamic acid

alkyl ester. This hemiaminal derivative spontaneously hydrolyzes in water to give

formaldehyde and N-alkylcarbamic acid alkyl ester in the same way that hydroxymethyl

amides do (Johansen et al., 1979 and Bundgard et al., 1980). Similarly the hydroxide

anion can act as a nucleophile and displaces the N-alkylcarbamic acid alkyl ester anion as

in Figure 2-9B. The hydroxymethyl conjugate of the drug, YH, formed as an intermediate

reverts to the parent drug by loss of formaldehyde. This particular mechanism however

can be ruled out because of the lower pKa of YH (-5.2-9.5) compared to the N-

alkylcarbamic acid alkyl ester (-14) making Y- a better leaving group than the N-

alkylcarbamic acid alkyl ester anion.






64



0 -CH2=0O 0

O l N OH 0 N H
R R


OH,
o YH + OH


0 OH

0 N Y


- CH2=O


Y-OH


0 0
OH2 H
O N O N +C
R R
Figure 2-9 SN2 type of hydrolysis of NANAOCAM-Y with Y- or carbamate as the
leaving group.


OH,
Y YH + OH-
+


OH,
20


-CH2=0


0



R


Figure 2-10 SNI type of hydrolysis of NANAOCAM-Y.


-OH


H









A SNI type of pathway is another possibility (Figure 2-10). The lone pair of

electrons on nitrogen can donate its electrons to stabilize an incipient carbocation with Y-

leaving. This carbocation can subsequently react with water to give hydroxymethyl-N-

alkylcarbamic acid alkyl ester which then falls apart to give N-alkylcarbamic acid alkyl

ester. However none of the data plotted in Figures 2-6 and 2-7 provides an insight into

the mechanism of hydrolysis since both SNI and SN2 mediated hydrolysis are dependent

on the leaving group ability of the ionized parent molecule, Y-. However the fact that the

hydrolyses were pH independent strongly favors a SNI mechanism. There is also some

precedent in literature suggesting the mechanism of hydrolysis to be SNl. N-

Alkylamidomethyl esters of carboxylic acid hydrolyze by a SNI mechanism (Moreira et

al., 1994) (Figure 2-11).

0 0 0 0

Ph N 0- R' Ph N -- Ph N OH
R R H2 R
+

R'COO
Figure 2-11 Hydrolysis of N-alkylamidomethylcarboxylic acid esters in aqueous buffers.

Hydrolysis proceeds by donation of electrons from the amide nitrogen resulting in a

carbocation intermediate with the carboxylate group leaving. In the subsequent step, the

carbocation reacts with water to give the hydroxymethylamide. When R' was naproxen

and R was methyl the t 1/2 for hydrolysis of the prodrug was 14 min at pH 7.4 and 320 C,

while when R was -CH2COOEt, t 1/2 for hydrolysis was 2520 min. Thus, introducing an

electron withdrawing group on the amide nitrogen diminishes its ability to donate

electrons and slows down reaction rate, thereby suggesting the mechanism of hydrolysis

to be SNI.










The hydrolysis of 0-Imidomethyl derivatives of phenols has been reported by Getz

et al., (1992). Figure 2-12 above compares the halflives of imidomethyl derivatives of p-

nitrophenol (PNP) with (N-Methyl-N-methoxycarbonyl) aminomethyl-PNP. The

proposed hydrolytic mechanism for the imidomethyl derivatives where saccharin,

phthalimide and succinimide served as the imide, was SN2. Thus, presence of two electron

withdrawing groups on the nitrogen disfavors SNI, (as observed in case of N-alkylamido

methyl esters of carboxylic acids) to the extent that SN2 becomes the major pathway.

t 1/2(min)

0.75(250C) 4.9(250C) 7.43(250C) 22(460C)

NO2 NO, NO2 NO2





0 0 0
O 00

O NO NO OO N-
i o 0
0 0







NO2





OH
Figure 2-12 Hydrolysis of N-imidomethyl derivatives of phenols in aqueous buffers.
Values are shown for tl/2 at the temperature (in parenthesis).









Implications of NANAOCAM-phenol Hydrolysis on Prodrug Design.

NANAOCAM-phenol conjugates are chemically stable and the suggestion that

their mechanism of hydrolysis is SN1 is based on past literature. We were interested in

designing derivatives which would be labile under physiological conditions and which

would illustrate that the mechanism of hydrolysis was SNI. Hence the syntheses of

NANAOCaminoethylidene (ROCO-NR'CH(CH3)), NANAOCaminobenzylidene

(ROCO-NR'CH(Ph)) and NArNAOCAM (ROCO-N(Ph)CH2) promoieties were

examined. NANAOCaminoethylidene and NANAOCaminobenzylidene derivatives of

phenols will stabilize positive charge on the methine carbon leading to a more labile

derivative if hydrolysis follows a SN1 type of pathway and will hinder an SN2 type path

by creating more steric hinderance. Also if the mechanism is SN2, the rate of hydrolysis

should be accelerated by an electron withdrawing group on an N-aryl group where a

positive charge on CH2 in N-CH2-O is increased making the CH2 a better target for

nucleophilic attack. On the other hand, if the mechanism is SN1, the rate of hydrolysis

should be decelerated by an electron withdrawing group on an N-aryl group where a

positive charge on CH2 in N-CH2-O is destabilized. Unfortunately our attempts to

synthesize NANAOCaminoethylidene and NANAOC-aminobenzylidene derivatives of

phenols were unsuccessful. The synthetic protocol used is shown in Figure 2-13. An

aldehyde was reacted with aqueous methyl amine at room temperature to give N-

methylimines which were reacted with methyl chloroformate under nitrogen at -780C to

00C. The major product obtained was the parent aldehyde, N-methyl carbamic acid

methyl ester and 2-3% of the desired alkylating agent (NANAOCaminoethylidene

chloride or NANAOCaminobenzylidene chloride) based on 1H-NMR.

















0o^. Y 0 0
Y 0 N 0Y O N O

X= H, OCH,, COOEt
NANAOaminoethylidene-Phenol NANAOaminobenzylidene-Phenol








x
X= H, OCH,, COOEt
NArNAOCAM-Phenol
Attempts to isolate the alkylating agent by column chromatography led to complete

decomposition of the alkylating agent. Reaction of phenols with the crude reaction

mixture containing the alkylating agent also lead to hydrolysis of the alkylating agent to

the parent aldehyde and N-methyl carbamic acid methyl ester quantitatively.

CHOCOCI
RCHO +CH3NH2 P RCH=NHCH3 RCHO + CH3NHCOOCH3 + RCH -N

major I COOCH3
R = Methyl or Aryl
NANAOC amino ethylidene chloride
or
NANAOC amino benzylidene chloride

minor
Figure 2-13 Attempts at the synthesis of NANAOCaminoethylidene chloride and
NANAOCaminobenzylidene chloride.

However our attempts to make NArNAOCAM conjugates of phenols were

successful. NArNAOCAM chlorides (R'OCONRCH2-, R = aryl, R'= CH3) were

synthesized and used to alkylate p-nitrophenol (Figure 2-3, Table 2-1, 8-10). The half-

lives of hydrolyses clearly illustrate that when an N-alkyl group (Table 2-3, t 1/2 for 7 =

22 min) was replaced by N-aryl (Table 2-5, t 1/2 for 8 and 10 = 400-520 min while t 1/2 for









9 > 24 h) the rates of hydrolyses were slower. An electron donating substituent on the N-

aryl ring like -OCH3 stabilizes the SNI transition state more than an electron withdrawing

group like -COOC2H5 or -H and the half lives are in the order of 9 >10 >8. Thus,

hydrolysis of NArNAOCAM and N-aryl-N-alkyloxycarbonylaminomethyl derivative of

phenols follow a SNI type of mechanism. However, although NANAOCAM-phenols

hydrolyse by a SNI type of pathway, they are chemically too stable to revert to the parent

drug at a sufficient rate to be effective prodrugs based on chemical hydrolysis.

Hydrolyses of NANAOCAM-phenols does occur enzymatically. When 1 was used

in diffusion cell experiments where hairless mouse skin was the membrane, about 23% of

acetaminophen was regenerated from the prodrug during its passage through the skin. We

will talk more about diffusion cell experiments in Chapter 3.

Thus NANAOCAM-phenols represent a novel class of prodrugs which are

sufficiently chemically stable to allow formulation but may be sufficiently enzymatically

labile to revert to the parent drug at a useful rate. Shorter chain NANAOCAM prodrugs

of phenols have both higher lipid and water solubilities compared to ACOM and

AOCOM phenolic conjugates (solubililities shown in Chapter 3). Thus, replacing the

oxygen atom in O-CH2 of ROCOOCH2 with a substituted nitrogen (N-R, R = alkyl)

makes it possible to increase solubility in a membrane which should increase

permeability across a biological barrier such as skin and oral absorption across the GI

tract.









0

02N /N-0 N 0





x


Table 2-5 N-aryl-N-alkyloxycarbonylaminomethyl Derivatives of p-nitrophenol (PNP) a
Compd X log kobsv(sec-1) t 1 (min.)
8 OMe -4.54 400
9 COOEt >24 h
10 H -4.65 519
a Hydrolysis experiments in pH 8.8 and 460C

Implications of NArNAOCAM-carboxylic acid Hydrolysis on Prodrug Design

Besides NANAOCAM derivatives of phenols, NANAOCAM derivatives of

carboxylic acids have also been examined. Carboxylic acids, like phenols, are polar and

being ionized at physiological pH penetrate membranes with difficulty. By transiently

masking the negative charge it has been shown that it is possible to increase the

permeation of carboxylic acids across biological barriers (Bonina et al., 1995 and Rautio

et al., 1999). N-Alkyl-N-alkyloxycarbonylaminomethyl derivatives of carboxylic acids

are chemically too unstable to serve as useful derivatives (compared to phenols which

were chemically stable) because the leaving group has a pKa of -3-5 (compared to 6.8-

9.5), e.g. 15, t 1/2 ~ 6 min at pH 7.1 and 390C. Thus, the goal here was to design

derivatives which were chemically more stable so that they could be easily formulated

but would hydrolyze independent of enzymes. It has been previously shown that 0-

imidomethyl derivatives of phenols like estradiol were enzymatically stable but had a half

life of 100 min in pH 7.4 buffer (Patel et al., 1994 and 1995). This derivative was about 8

times more potent than estradiol alone when given orally. Thus prodrugs whose









hydrolysis is not dependent on enzymes are useful because their hydrolysis isn't limited

by enzymatic variability and strict substrate specificity of some esterases in the blood

which could lead to incomplete hydrolysis to the parent drug (Beamount et al., 2003).

Here, naproxen was chosen as a model carboxylic acid drug for which to prepare

NANAOCAM prodrugs (Table 2-2). The NANAOCAM derivatives of naproxen were

made by alkylating naproxen with NANAOCAM-Cl in presence of triethyl amine in

dichloromethane (Figure 2-3).

11, Table 2-2 had a half life of 2 min at pH 7.1 and 390 C which would make

pharmaceutical formulation difficult. Since the mechanism of hydrolysis of

NANAOCAM esters of carboxylic acids should also be SNI, the NArNAOCAM

promoiety should lead to a chemically more stable and useful prodrug of a carboxylic

acid since the phenol NArNAOCAM derivatives were much more stable. NArNAOCAM

ester derivatives of naproxen (12-14, Table 2-2) were synthesized by alkylating naproxen

with NArNAOCAM-Cl (Figure 2-4) and their rates of hydrolysis determined (Table 2-6).

As expected the substituted N-aryl derivatives increased the half-lives of the hydrolyses

from 2 min to over 150 min. As in the case of phenols the order of rates of hydrolyses

was, -COOEt < -H < -OMe which is consistent with a SNI type of hydrolysis. All three

NArNAOCAM naproxen derivatives represent potentially useful prodrugs which are

reasonably stable but should be sufficiently labile enough to release the drug at an

appropriate rate to express its pharmacological activity. Moreover, since 14 would

eventually release p-aminobenzoic acid, a component of sunscreens, it may be the most

attractive candidate.









x






0\ O N


CHO 0


Table 2-6 Hydrolysis of N-aryl-N-alkyloxycarbonylaminomethyl Derivatives of
Naproxen.b
Compd X log kobs,,.(sec-1). t / (min.)
12 H -4.015 119.62
13 OMe -3.63 49.83
14 COOEt -4.13 156.46
Hydrolysis experiments in pH 7.1 and 390C

The kinetics of decomposition of N-(4'-ethyloxycarbonylphenyl)-N-methyloxy-

carbonylaminomethyl ester of naproxen, 14, was also studied in aqueous buffers at 390 C

over a wide pH range (Figure 2-3).

Table 2-7 Effect of pH of Buffer on Rates of Hydrolysis of 14.
pH tl/2 (min) log k
4 166.76 -4.16
6 156.29 -4.13
7 156.46 -4.13
8.25 152.72 -4.12
9.2 35.45 -3.49

The rates of hydrolysis were independent of the pH of buffers from pH 4.0-8.25 so

the mechanism of hydrolyses along this pH range is believed to be SNI (Moreira et al.,

1996 and Iley et al., 1997). At pH 9.2, a sharp increase in rates of hydrolysis was

observed. This is probably due to a change in the mechanism of hydrolysis from SNI to

addition-elimination where nucleophilic addition occurs on the carbonyl functional group

of the ester followed by elimination of carboxylate ion to form hydroxymethyl-N-methyl










carbamic acid methyl ester (Bundgard et al., 1991). N-Methyl carbamic acid methyl ester

is formed in the subsequent step by the loss of formaldehyde (Figure 2-14). Attempts to

make NArNAOCAM-naproxen derivatives with enhanced water solubility are currently

under investigation. NArNAOCAM-naproxen conjugates are thus prodrugs of carboxylic

acids which unlike ACOM or AOCOM esters are not dependent on enzymatic hydrolysis

and have sufficient chemical stability to be formulated.



/ / H ,--
R N 0 HO N O N 0
H : -CH2=0O



COOEt COOEt COOEt

+

RCOOH
Figure 2-14 Hydrolysis of NArNAOCAM-carboxylicacid conjugates at pH 9.2.














CHAPTER 3
SYNTHESIS AND TOPICAL DELIVERY OF N-ALKYL-N-
ALKYLOXYCARBONYLAMINOMETHYL PRODRUGS OF A MODEL PHENOLIC
DRUG:ACETAMINOPHEN

Topical approaches to masking a phenolic functional group have been limited to

morphine (Drustrup et al., 1991), naltrexone (Stinchcomb et al., 2002; Pillai et al., 2004;

Valiveti et al., 2005; Hammell et al., 2005; Vaddi et al., 2005, Valivetti et al., 2005a),

buprenorphine (Stinchcomb et al., 1996), nalbuphine (Sung et al., 1998) and

acetaminophen (Wasdo and Sloan, 2004). Only three promoieties, alkylcarbonyl (AC),

alkyloxycarbonyl (AOC) and alkylaminocarbonyl (AAC), have been utilized but no soft

alkylated derivatives of phenol have been used for topical delivery. Both AC and AOC

derivatives of phenols rely on esterase catalysed hydrolysis or addition-elimination type

mechanism to revert to the parent drug.The ACOM and AOCOM soft alkyl approach is

presently being investigated in our lab to improve topical delivery of phenols. Insertion of

'N(R')CH2' into the AOC (ROCO-) promoiety gives the NANAOCAM

(ROCONR'CH2-) promoiety which should act as a soft alkyl promoiety of phenols since

it did for 6MP (Siver and Sloan, 1990). It is of interest to see if this insertion improves

the biphasic properties and flux through skin of phenols since addition of heteroatoms

into the promoiety improves aqueous solubility of drugs. In order to test this hypothesis

acetaminophen (APAP) was chosen as the model phenolic drug.

The performance of the prodrug series in diffusion cell experiments will be

compared with that predicted by the Roberts-Sloan (RS) equation. The results will be









added to the Sloan and coworkers database to generate new coefficients for the

parameters in the equation and give a more robust RS equation.

Experimental Procedure

Materials and Methods

Isopropyl myristate (IPM) was obtained from Givaudan Corp (Clifton, NJ).

Theophylline (ThH) was purchased from Sigma Chemical Co. (St. Louis, MO, USA); all

other reagent chemicals were from Aldrich Chemical Co. (Milwaukee, WI, USA). The

water was obtained from a Millipore Milli-Q water ultra filteration system. Ultraviolet

spectra were recorded on a Shimadzu UV-2501 PC spectrophotometer. A radiometer pH

meter 26 was used to determine pH of solutions.The vertical, Franz type diffusion cells

were from Crown Glass (Somerville, NJ, USA) (surface area 4.9cm2, 20 mL receptor

phase volume, 15 mL donor phase volume). The diffusion cells were maintained at 320C

with a Fischer (Pittsburgh, PA, USA) circulating water bath model 25. The female

hairless mice (SKH-hr-1) were from Charles River (Boston, MA, USA). Statistical

analyses were carried out by using SAS 9.0. All animal sacrifices and preparation of

membranes were carried out by Professor Ken Sloan using IACAC approved protocols.

Synthesis of Prodrug Derivatives

Synthesis begins by the alkylation of APAP with N-alkyl-N-alkyloxycarbonyl

aminomethyl chloride (NANAOCAM-C1) in presence of a base like triethylamine and

CH2C12 as the solvent. In every case it was necessary to synthesize the corresponding

alkylating agent, NANAOCAM-C1.









0

N K Cl 0 R 0
CH2O + CH3NH2 NaOH N OR
N N
DCM
R= CH3 and C2H5
Figure 3-1 Synthesis of NANAOCAM-Cl from 1, 3, 5-trialkylhexahydrotriazine.

Synthesis of NANAOCAM-C1 (Figure 3-1)

a) 1, 3, 5-Trimethylhexahydrotriazine was synthesized from equimolar equivalents of

aqueous formaldehyde, methyl amine and NaOH according to the protocol originally

developed by Graymore et al., (1932) and modified by Siver et al., (1990). Methyl amine

(0.4 mol of 40% aqueous) was placed in an ice bath and an equivalent of 37% aqueous

formaldehyde was added dropwise over a period of 10 min. The solution was allowed to

equilibrate to room temperature and stirred for one hour, then an equivalent of NaOH was

added and the contents were stirred for 1.5 h more. The solution was extracted with 4 x

50 mL CH2Cl2 and the collected CH2Cl2 were dried over Na2SO4, filtered and

concentrated to a clear colorless CH2C2 solution containing the the hexahydrotriazine

derivative. Complete evaporation of CH2Cl2 wasn't carried out as it resulted in loss of the

desired product. For quantification purposes the 'CH2' ofCH2C2 at 65.3 and N-CH2-N

peak of hexahydrotriazine derivative at 63.2 were used, yield = 82% in CH2Cl2, 1H NMR

(CDCl3): 62.3 (s, 9H), 63.2 (s, 6H).

Next the 1, 3, 5-trimethylhexahydrotriazine (freshly prepared) in CH2Cl2 was added

to an ice cold solution of 3 equivalent of alkyl chloroformate in CH2C2 over a period of

10 minutes. The white suspension that was observed was allowed to equlilibrate to room

temperature and stirred overnight. The suspension was filtered and the filtrate

concentrated to an oil. The oil contained the desired product and corresponding bis(N-









Alkyl-N-alkyloxycarbonyl)aminomethanes with a 'CH2' peak at 64.8.The oil was

purified by trituration with hexane overnight followed by ether overnight. The clear

solution was decanted leaving the white residue (bis derivative) behind. The clear

solution was then concentrated. The peak at 64.8 due to the 'CH2 of the bis derivative

and peak at 65.3 due to the 'CH2' of N-alkyl-N-alkyloxycarbonylaminomethyl chloride

were used to quantitate the amount of product formed.

N-Methyl-N-methyloxycarbonylaminomethyl chloride: yield = 90%, 1H NMR (CDCl3):

62.9 (s, 3H), 63.75 (s, 3H), 65.3 (s, 2H).

N-Methyl-N-ethyloxyoxycarbonylaminomethyl chloride: yield = 89%, 1H NMR

(CDCl3):61.3 (t, 3H), 62.9 (s, 3H), 64.22 (q, 2H), 65.33 (s, 2H).

N-Methyl-N-propyloxycarbonylaminomethyl chloride, N-methyl-N-butyloxycarbonyl-

aminomethyl chloride, N-ethyl-N-methyloxycarbonylaminomethyl chloride and N-

methyl-N-hexyloxycarbonylaminomethyl chloride were synthesized in an alternate way

via a N-alkyl carbamic acid alkyl ester (Figure 3-2).


RO Cl






3 ROH


0
TEA H
+ R'NH2 RO NH
SR'




+ A TEA
CIS3CO OCCI0 3


(CH2O)n 0

THF RO N/CN I
TMSCI R'


0

RO Cl


R = C3H7, C4H9, CH13
R' = CH3, C2H5
Figure 3-2 Preperation of alkyl chloroformate in situ from alcohol and synthesis of
NANAOCAM-Cl from alkyl amine.









b) N-Alkyl carbamic acid alkyl ester.

To a solution of chloroformate (33 mmol) in 75 mL CH2Cl2 mounted in an ice bath, was

added an equivalent of triethylamine and aqueous alkyl amine. The reaction mixture was

allowed to warm to room temperature and stirred overnight. The clear solution was

washed with 3 x 10 mL brine and the organic layer was dried over Na2SO4 and

concentrated to an oil. This oil was purified by trituration with hexane overnight. The

suspension that resulted was filtered and the filtrate evaporated to a yellow oil.

N-Methyl carbamic acid propyl ester: yield = 60 %, 1H-NMR (CDCl3): 60.97 (t, 3H),

61.7 (m, 2H), 62.9 (d, 3H), 64.06 (t, 2H) ,64.58 (s, 1H).

N-Methyl carbamic acid butyl ester: yield = 49 %, 1H-NMR (CDCl3): 60.92 (t, 3H),

61.38 (m, 2H), 61.6 (t, 2H), 62.8 (d, 3H), 64.06 (t, 2H) ,64.58 (s, 1H).

In the case of N-methyl carbamic acid hexyl ester it was necessary to synthesize the

chloroformate by reacting three molar equivalents of hexanol with an equivalent of

triphosgene and three equivalents of a poorly nucleophilic base, triethylamine (Figure 3-

2). A solution of triphosgene (11 mmol) in 100 mL CH2Cl2 was first prepared and placed

in an ice bath. To this solution was added a mixture of triethylamine (33 mmol) and the

alcohol (33 mmol) in 25 mL CH2Cl2 dropwise over 20 min. The exothermic reaction was

allowed to proceed, the contents were allowed to warm to room temperature and stirred

overnight. To this preformed chloroformate was added methyl amine and triethylamine as

stated above in the synthesis of N-alkyl carbamic acid alkyl esters.

N-Methyl carbamic acid hexyl ester: yield = 59%, 1H NMR (CDCl3): 6 0.89 (t, 3h), 6

1.31 (m, 6H), 6 1.58 (m, 2H), 62.79 (d, 3H), 64.04 (t, 2H) ,64.69(s, 1H).

(c) N-Alkyl-N-alkyloxycarbonylaminomethyl chloride:









A suspension of N-alkyl carbamic acid alkyl ester, 1.7 equivalents of paraformaldehyde

and 13 equivalents of trimethylsilyl chloride were refluxed using a CaCl2 drying tube and

water condenser for 2.5 h over an oil bath. The suspension was diluted with CH2Cl2 and

filtered to get rid of the unreacted paraformaldehyde. The clear filterate was concentrated

with a rotavapor at 400C under reduced pressure. The yellow oil obtained was triturated

with hexane overnight, the white suspension obtained was filtered, and the filtrate was

concentrated to give the desired alkylating agent.

N-Methyl-N-propyloxycarbonylaminomethyl chloride: yield = 85%, 1H NMR (CDCl3):

60.97 (t, 3H), 61.7 (m, 2H), 63.0 (s, 3H), 64.12 (t, 2H), 65.35 (s, 2H).

N-Methyl-N-butyloxycarbonylaminomethyl chloride yield = 82%, 1H NMR (CDCl3):

60.95 (t, 3H), 61.41 (m, 2H), 61.65 (m, 2H), 63.0 (s, 3H), 64.17 (t, 2H), 65.33 (s, 2H).

N-Methyl-N-hexyloxycarbonylaminomethyl chloride: yield = 83%, 1H NMR (CDCl3): 6

0.97 (t, 3H), 6 1.3 (m, 6H), 6 1.6 (m, 2H), 63.02 (s, 3H), 64.15 (t, 2H), 65.33 (s, 2H).

Synthesis ofNANAOCAM-APAP prodrugs (Table 3-1): Briefly a suspension of

APAP (0.Olmol) and triethylamine (0.011 mol) in 25 mL dichloromethane was stirred

under reflux conditions for an hour followed by the addition of the alkylating agent,

NANAOCAM-Cl (0.01 Imol). The contents were stirred overnight and subsequently

diluted with 50 mL dichloromethane and washed with water (3 x 10 mL). The

dichloromethane layer was, dried over Na2SO4 and concentrated at 400C over a rotary

evaporator. The yellow oil obtained was then purified by column chromatography using

ethyl acetate: hexane as eluent.









Table 3-1. NANAOCAM Prodrugs of APAP.
0

0 N O, R





NHCOCH3

Compound N-R' O-R

1 CH3 CH3
17 CH3 C2H5
18 CH3 C3H7
19 CH3 C4H9
20 CH3 C6H13


1 was prepared in 70 % yield from N-methyl-N-methyloxycarbonylaminomethyl

chloride, triethylaminie and acetaminophen in CH2C2 after recrystallization of the oil

obtained using CH2C2: hexane 1:3: mp = 86-880C, Rf (0.36,ether), 1H NMR(CDCl3):

67.6(s,1H), 6 7.39(d,2H), 66.96-6.87(2d,2H), 65.28-5.21(2s,2H), 63.72-3.7(2s,3H), 6 3.0-

2.97(2s,3H), 6 2.0(s,3H). Anal calcd for C12H16N204 : C, 57.13; H, 6.39; N, 11.1. Found:

C, 56.81; H, 6.34; N, 11.

17 was prepared in 65% yield from N-Methyl-N-ethyloxycarbonyl aminomethyl chloride,

triethylamine and acetaminophen in CH2C2 after trituration with hexane overnight to

give yellow crystals which were recrystallized from ethyl acetate:hexane (1:3) to give

white crystals: mp = 75-770C, Rf (0.6, Et20), 1H NMR(CDCl3): 6 7.38(d,2H),

67.12(s,1H), 6 6.96(dd,2H), 65.3(s,2H), 64.15(q,2H), 6 3.0(s,3H), 6 2.15(s,3H),

61.25(3H,t). Anal calcd for C13H18N204 : C, 58.64; H, 6.81; N, 10.52. Found: C, 58.53;

H, 6.93; N, 10.47.









18 was prepared in 76% yield from N-Methyl-N-propyloxycarbonyl aminomethyl

chloride, triethylamine and acetaminophen in CH2C2 after silica gel column

chromatography in ethyl acetate:hexane (3:2) followed by trituration in hexane overnight

to give pale white crystals. These crystals were further recrystallized from ethyl

acetate:hexane (3:2) to give white crystals: mp = 58-590C, Rf (0.2, ethyl acetate:hexane

3:2), H NMR(CDCl3): 6 7.38(d,2H), 67.14(s,1H), 6.96(dd,2H), 65.3(s,2H), 64.05(t,2H),

6 3.0(s,3H), 6 2.15(s,3H), 61.65(2H,m) 60.91(3H,t). Anal calcd for C14H20N204: C,

60.01; H, 5.45; N, 12.73. Found: C, 60.1; H, 5.35; N, 12.58.

19 was prepared in 66 % yield from N-Methyl-N-butyloxycarbonyl aminomethyl

chloride, triethylamine and acetaminophen in CH2C2 to give a colorless oil which was

purified after silica gel column chromatography in ethylacetate:hexane (3:2) followed by

trituration in hexane overnight to give a colorless oil: Rf= 0.13(3:7, Ethyl acetate:

hexane), H NMR(CDCl3): 6 7.38(d,2H), 67.14(s,lH), 6.96(dd,2H), 65.3(s,2H),

64.05(t,2H), 6 3.0(s,3H), 6 2.15(s,3H), 61.65(2H,m), 61.4(2H,m), 60.91(3H,t). Anal calcd

for C15H23N204 : C, 61.21; H, 7.53; N, 9.52. Found: C, 61.16; H, 7.65; N, 9.24.

20 was prepared in 63 % yield from N-methyl-N-hexyloxycarbonylaminomethyl

chloride, triethylamine and acetaminophen in CH2C2 to give a colorless oil which was

purified after silica gel column chromatography in ethyl acetate:hexane (4:1) followed by

trituration in hexane overnight to give a colorless oil: Rf (0.36, 7:3, ethyl acetate :

hexane), 1H NMR(CDCl3): 6 7.38(d,2H), 67.29(s,1H), 66.95-6.89(2d,2H), 65.29-

5.23(2s,2H), 64.11-4.07(m,2H), 6 3.02-2.98(2s,3H), 6 2.15(s,3H), 61.63(2H,m),

61.3(6H,m), 60.89(3H,m). Anal calcd for C17H27N204 : C, 62.33; H, 8.13; N, 8.69.

Found: C, 61.96; H, 8.14; N, 8.36.









Determination of Solubilities and Partition Coefficients

Molar absorbtivities were determined in triplicate for each member of the series in

acetonitrile (ACN) and pH 7.1 phosphate buffer at the maximum absorption wavelength.

The molar absorbtivities were calculated using Beer's law (E = A/c). The solubilities of

the prodrugs in isopropyl myristate (IPM) were determined in triplicate by stirring

suspensions of the compound in 3 mL IPM with a magnetic stirrer for 24 h at room

temperature (23 1C) (Martin and coworkers, 1985 and Beall et al., 1994). The test

tubes containing the suspensions were sealed and thermally insulated from the stirrer.

After stirring, the suspensions were filtered through a 0.45 [tm nylon membrane filter. An

aliquot (-0.1-0.3 mL) was withdrawn from the clear filtrate of saturated solutions and

diluted to 10 mL in a volumetric flask with ACN. The samples were then analyzed by

UV spectroscopy using absorbances determined at 240 nm for APAP prodrugs. The

solubility in IPM was calculated from the following relationship (equation 3-1):

SIPM = (A/s) (Dilution factor) (3-1)

Dilution factor = (Vfinaialiquot) (3-2)

where A is the absorbance of the sample at 240 nm, s is the molar absortivity of the

sample at 240 nm in ACN and Valiquot is the volume of the saturated filterate aliquot and

Vfinal is the final diluted sample volume. In the case of prodrug derivatives which were

oils, direct solubility measurements were not possible. Thus partition coefficients

between IPM and pH 4.0 buffers were used to estimate SIPM.

Solubilities in water (SAQ) were determined by stirring suspensions in deionized

water for 1 h to limit the extent of hydrolysis of the prodrugs. The samples were filtered

through nylon filters diluted with ACN and analyzed by UV spectroscopy. SAQ was

determined using the following equation:









SAQ = (A/s) (Dilution factor) (3-3)

Determination of SAQ of compounds which were oils was carried out by stirring the

compound in deionized water for 1 h and ensuring that a biphasic solution was present at

all times. The test tubes containing the biphasic solution were centrifuged for 2 minutes;

an aliquot (-0.1-0.3 mL) was withdrawn from the water layer and diluted to 10 mL with

ACN in a volumetric flask. The samples were then analyzed by UV spectroscopy as

above.

For determination of partition coefficients between IPM and pH 4.0 buffer (KIPM:

4.0) for compounds which were solids at room temperature, a measured volume (-0.5 mL-

1 mL) of the filtered saturated IPM solutions from the lipid solubility experiments were

mixed with a measured volume of pH 4.0 acetate buffer (-1-5 mL) in a 10 mL test tube

(Beall et al., 1993). The test tube was capped and vigorously shaken for 10 seconds and

subsequently centrifuged for 2 min to allow the clear separation of two phases. An

aliquot (-0.3 mL) was withdrawn from the IPM layer and diluted to 10 mL with ACN in

a volumetric flask and analyzed by UV spectroscopy. The KIPM: 4.0 was calculated using

the following relationship:

KIPM: 4.0 = (V4.0/VIPM) x AF/ (Al-AF) (3-4)

where V4.0 is the volume of pH 4.0 buffer used, VIPM is the volume of IPM used, AI is the

initial absorbance of the saturated IPM solution before partitioning and AF is the

absorbance of the compound remaining after partitioning. The pH 4.0 buffer solubility

was estimated from KIPM: 4.0 using the following equation:

S4.0 = SIPM / KIPM: 4.0 (3-5)









Determination of partition coefficients between IPM and water (KIPM: AQ) for

compounds which were oils at room temperature was carried out by using a measured

volume (-0.5 mL 1.0 mL) of saturated solution used for determination of SAQ placed in

a 10 mL test tube and a measured volume (-1-10 mL) of IPM. The test tubes were

centrifuged; an aliquot was taken from the water layer and diluted with ACN to 10 mL in

a volumetric flask. Samples were subsequently analyzed by UV spectroscopy. KIPM: AQ

and SIPM were calculated using the following equations:

KIPM: AQ = (V4.0/VIPM) X (AI-AF)/AF (3-6)

SIPM = KIPM: AQ x SAQ (3-7)

Solubility ratios (SR) were calculated from the ratio of SIPM / SAQ. The methylene Rt

values were calculated using equation 3-8:

x7 = (log SR n+m log SR n)/m (3-8)

where n is the number of methylene units in the promoiety of one prodrug (the lowest

member of the homologous series) and m is the number of additional units in the

promoiety in the higher member of the homologous series. Similarly the methylene 7R

values (Leo et al., 1971 and Hansch and Leo, 1979) using K values are also reported.

t = (log K n+m log K n)/m (3-9)

Determination of Flux through Hairless Mice Skins (Beall et al., 1994)

The mice were rendered unconscious using CO2 and sacrificed by cervical

dislocation. Full thickness skins were removed by dissection; the pieces were scraped to

remove excess fat, cut into proper sizes and placed dermal side down on the diffusion

cells. A Franz diffusion cell consists of two compartments; donor compartment and a

receptor compartment. The receptor compartment has a side arm which can be used to

sample the compartment. The skins were held in place using rubber rings, and the two




Full Text

PAGE 1

PRODRUG STRATEGIES AIMED AT IM PROVING TOPICAL DELIVERY OF DRUGS USING THE N-ALKYL-N-ALK YLOXYCARBONYLAMINOMETHYL (NANAOCAM) PROMOIETY By SUSRUTA MAJUMDAR A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Susruta Majumdar

PAGE 3

This document is dedicated to my wife, PUJA and my father, BABA

PAGE 4

iv ACKNOWLEDGMENTS I would like to thank my advisor, Dr. Ken Sloan, wholeheartedly for allowing me to pursue his and my ideas independently in his lab. His constant support, patience, ability to keep me focused, thorough unders tanding of literature, and stress on basic principles proved vital in my finishing this manuscript. The rigorous training imparted to me in and outside the lab was responsible in my doing well in the lab, for the publications that have or will come out of this diss ertation and talking to people in various conferences and interviews I a ttended during my stay at Flor ida. I greatly appreciate his ability to communicate with me on a daily ba sis the answers to my questions in an articulate manner. I feel confident enough to ta ke myself to the next level in search of something new again. There are two people who inspire me the mo st: one of them is my father whom I call baba and second is my wife, Puja. I would like to dedicate 4 years of my work to my father and my wife. My father and my advisor have two th ings in comman. They both believe in keeping things simple and have great faith in basic principles and honestly I thought I had those under control till I joined the Sloan lab. My father introd uced me to science at an early age. He along with my mother made it possible for me to have the best education. I went to the best private schools back home ju st because they valued education a lot in spite of facing financial hardsh ips sometimes. I did not pick my college major, chemistry;

PAGE 5

v he did. He said, “Just trust me on this on e” and I am so glad I did. My dad being a pharmacist influenced me to do what he did. I got my chance at Flor ida and I loved it. I dated my present wife for 6 years before we got married. She keeps me going everyday day after day with t hose constant words of encouragement. If my father was responsible for me being in science, my wi fe ensured that his dream of my being a scientist gets fulfilled. I admire the strength in her character to stay from me to finish her education. I would like to thank my mother and sister for their love and affection. My mother was responsible for me being independent, tole rant and disciplined. I admire my sister’s ability to fight ‘time’ against all odds. I would like to thank my graduate collea gues Dr. Scott Wasdo, Dr. Joshua Thomas and Maren Muellar Spaeth. Scott actually was my second advisor; sometimes his explanations for a physical chemistry concept were just too easy to follow. I hope he decides to teach one day. I do not have words to describe Josh. He is too nice, too good and focused. I would like to thank him for the critical discussions we had in the lab on science, religion, politics and everything on earth. I admire hi m as a wonderful scientist; I do not know of too many people who will run an experiment for 36 hrs non stop week after week. Such dedication to work amazes me. Maren was an undergraduate exchange student who worked with me. Together we finished couple project s in the limited time she had in her disposal. I appreciate her a dding insight and thoughtfulness into our work. I would like to thank my committee members Dr. M.O.James, Dr. Raymond G. Booth and Dr. William Dolbier. I would like to thank Dr. James for her constant encouragement and her writing good recommen dation letters for me. Dr. Dolbier was

PAGE 6

vi largely responsible for a project I did with Ma ren; I thank him for adding insight into the project and for being a part of my committee. Finally I would like to thank Dr. Gee ti Bansal (MIT), Dr. Sushma Chauhan (Ranbaxy) and Dr. Gagan Kukreja (Ranbaxy) w ho taught me basically how to operate in an organic chemistry lab like running TLCÂ’s, column chromatography and doing crystallizations during my stay as an underg rad student at the Un iversity of Delhi.

PAGE 7

vii TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES...............................................................................................................x LIST OF FIGURES..........................................................................................................xii ABSTRACT....................................................................................................................... xv CHAPTER 1 INTRODUCTION........................................................................................................1 The Skin....................................................................................................................... .1 Topical Delivery...........................................................................................................8 Approaches to Increase Permeation across the Skin..................................................11 Theory of Percutaneous Diffusion..............................................................................13 Mathematical Modeling of Flux through Human Skin. Derivation of RobertsSloan equation from FickÂ’s law of Diffusion..........................................................17 Prodrugs......................................................................................................................2 1 Prodrugs for Dermal and Transdermal delivery.........................................................33 Research Objectives....................................................................................................41 2 DESIGN, SYNTHESIS, HYD ROLYSIS OF N-ALKYL-NALKYLOXYCARBONYL-AMINOMETHYL DRUG DERIVATIVES AND ITS IMPLICATIONS ON PRODRUG DESIGN.......................................................43 Introduction.................................................................................................................43 Synthesis of N-methyl-N-methylox ycarbonylaminomethyl and N-aryl-Nmethyloxy-carbonylaminomethyl Prodrugs.....................................................45 N-Methyl-N-methyloxycarbonylaminomethyl chloride synthesis...............48 N-Aryl-N-methyloxycarbonylaminomet hyl chloride synthesis...................49 Alkylation of Phenols, 6-Mercaptopu rine, Dimethylaminobenzoic acid and Naproxen with N-methyl-N-methyloxy carbonylaminomethyl chloride or N-aryl-N-methyloxycarbonylaminomethyl chloride.......................................51 Hydrolysis Studies......................................................................................................59 Implications of NANAOCAM-phenol Hydrolysis on Prodrug Design.......67 Implications of NArNAOCAM-carboxy lic acid Hydrolysis on Prodrug Design.......................................................................................................70

PAGE 8

viii 3 SYNTHESIS AND TOPICAL DELIVERY OF N-ALKYL-NALKYLOXYCARBONYLAMINOMETHYL PRODRUGS OF A MODEL PHENOLIC DRUG:ACETAMINOPHEN.................................................................74 Experimental Procedure..............................................................................................75 Materials and Methods.................................................................................75 Synthesis of Prodrug Derivatives.................................................................75 Synthesis of NANAOCAM-Cl.....................................................................76 Determination of Solubilities and Partition Coefficients.............................82 Determination of Flux through Hairless Mice Skins...................................84 Determination of Prodrug Hydrolysis by UV Spectroscopy........................87 Calculation of Maximum Flux.....................................................................87 Physicochemical Properties of NANAOCAM Prodrugs of APAP............................88 Solubilities....................................................................................................88 Diffusion Cell Experiments..........................................................................91 Prodrug Bioconversion to Parent Drug........................................................92 Permeability Coefficients and Solubility Parameter Values........................92 Residual Amounts in Skin............................................................................93 Second Application Fluxes...........................................................................94 Modelling the Flux of NANAOCAM prodr ugs of APAP through Hairless Mouse Skin from IPM using the RS equation....................................................................95 Conclusions.................................................................................................................97 4 SYNTHESIS AND TOPICAL DELIVERY OF N-ALKYL-NALKYLOXYCARBONYLAMINOMETHYL PRODRUGS OF AN IMIDE CONTAINING DRUG: THEOPHYLLINE...............................................................99 Experimental Procedure............................................................................................101 Materials and Methods...............................................................................101 Synthesis of Prodrug Derivatives...............................................................102 Synthesis of NANAOCAM-Cl...................................................................102 Determination of Solubilities and Partition Coefficients...........................108 Determination of Flux through Hairless Mice Skins.................................110 Determination of Prodrug Hydrolysis by UV Spectroscopy......................113 Calculation of Maximum Flux...................................................................113 Physicochemical Properties of NA NAOCAM Prodrugs of Theophylline...............114 Melting Point Behaviour of NANAOCAM Prodrugs of Theophylline.....114 Solubilities..................................................................................................114 Diffusion Cell Experiments........................................................................116 Prodrug Bioconversion to Parent Drug......................................................118 Permeability Coefficients and Solubility Parameter Values......................120 Residual Amounts in Skin..........................................................................121 Second Application Fluxes.........................................................................121 Modelling the Flux of NANAOCAM Prodrugs of Acetaminophen and Theophylline through Hairless Mouse Skin from IPM using the RS equation....122 Conclusions...............................................................................................................125

PAGE 9

ix 5 SUMMARY OF RESULTS OBTA INED AND FUTURE WORK........................126 LIST OF REFERENCES.................................................................................................129 BIOGRAPHICAL SKETCH...........................................................................................143

PAGE 10

x LIST OF TABLES Table page 1-1 x, y, z, r2 and Average Residual Errors for Various Databases Fit to RS................21 1-2 Physicochemical Characterization of 1-Alkyloxycarbonyl esters of 5-FU..............35 1-3 Physicochemical Characteriza tion of Alkyloxycarbonyl esters of Acetaminophen.........................................................................................................38 1-4 Physicochemical Characterization of PEG esters of Indomethacin.........................40 2-1 N-methyl-N-methyloxycarbonylaminomet hyl-phenol Conjugates Synthesized.....46 2-2 N-methyl-N-methyloxycarbonylaminomet hyl and NArNAOCAM Conjugates of Naproxen..................................................................................................................46 2-3 Correlation of Rates of Hydrolysis of NANAOCAM-Y with pKa and Sigma Values of the Leaving Group (Y-)............................................................................61 2-4 Effect of pH of Buffer on Rates of Hydrolysis of 7 .................................................61 2-5 N-aryl-N-alkyloxycarbonylaminomethyl Derivatives of p-nitrophenol (PNP).......70 2-6 Hydrolysis of N-aryl-N-alkyloxycarbonyl aminomethyl Derivatives of Naproxen.72 2-7 Effect of pH of Buffer on Rates of Hydrolysis of 14. ..............................................72 3-1 NANAOCAM Prodr ugs of APAP...........................................................................80 3-2 Molecular Weights, Melting Points, L og Solubilities in Isopropyl Myristate Log Solubilities in Water and Log Solubilities in pH 4.0 Buffer............................90 3-3 Molar Absortivities in Acetonitrile a nd Buffer, Log Solubility Ratios between IPM and Water, the Differences Between Log SRIPM:AQ, the Log of Partition coefficients Between IPM and pH 4.0 Bu ffer, and the Differences Between Log KIPM:4.0......................................................................................................................90 3-4 Solubilities in IPM, Solub ilities in Water and Flux through in vitro Hairless Mouse Skins from IPM............................................................................................92

PAGE 11

xi 3-5 Log Permeability Values for the APAP from IPM through Hairless Mouse Skins and Solubility Parameter values...............................................................................93 3-6 Residual Skin Concentrations of To tal APAP and Ratios of Dermal versus Transdermal Fluxes..................................................................................................94 3-7 Second Application Theophylline Flux data for Flux of Theophylline from Propylene Glycol......................................................................................................94 3-8 Experimental Flux, Calculated Flux a nd Error in Predicting Flux for Compounds 1, 17-26 through Hairless Mouse Skins from IPM..................................................96 4-1 NANAOCAM Prodrugs of Theophylline..............................................................107 4-2 Melting Point Comparisons of N ANAOCAM-Th with AOC-Th and ACOM-Th.115 4-3 Molecular Weights, Melting Points, Log Solubilities in Isopr opyl Myristate, Log Solubilities in Water and Estimated Log Solubilities in pH 4.0 Buffer.................117 4-4 Molar Absortivities in Acetonitrile a nd Buffer, Log Solubility Ratios between IPM and Water, the Differences Between Log SRIPM:AQ, the Log of Partition Coefficients Between IPM and pH 4.0 Bu ffer, and the Differences Between Log KIPM:4.0....................................................................................................................118 4-5 Solubilities in IPM, Solubilities in Water and Fl ux through in vitro Hairless Mouse Skins from IPM..........................................................................................119 4-6 Log Permeability values for Theophylline Prodrugs from IPM through Hairless Mouse Skins and Solubility Parameter Values......................................................120 4-7 Residual Skin Concentrations of Total Theophylline Species ..............................121 4-8 Second Application Theophylline Flux data for Flux of Theophylline from Propylene Glycol....................................................................................................122 4-9 Experimental Flux, Calculated Flux a nd Error in Predicting Flux for Compounds 1, 17-26 through Hairless Mouse Skins from IPM................................................124

PAGE 12

xii LIST OF FIGURES Figure page 1-1 Cross sectional view of human skin ..........................................................................2 1-2 Brick and mortar model of dr ug absorption through the skin....................................6 1-3 Structures of ceramides seen in th e lipid bilayers of the lamellar bodies..................7 1-4 Illustration of a typical flux profile..........................................................................14 1-5 Two compartment diffusion model..........................................................................15 1-6 Esterase mediated hydrolysis of prodrugs................................................................23 1-7 Soft alkyl prodrug hydrolysis...................................................................................24 1-8 Hydrolysis pathways for Saccharin base d O-imidomethyl derivative of estradiol..26 1-9 Hydrolysis of phosphoryloxymethyl prodrug mediated by phosphatases...............27 1-10 Hydrolysis of PEG conjugated to drug using a hydroxyl-benzylalcohol linker in vivo ...........................................................................................................................30 1-11 PEG-Ala-Campothecan............................................................................................30 1-12 Ionic complexes formed by oligoarginin e of Tat with the pho sphate groups in cell membranes.........................................................................................................32 1-13 Chemical hydrolysis of oligoarginin e conjugates of cyclosporine and taxol...........33 1-14 Structure of testosterone and indomethacin esters with polarisable side chains......39 2-1 Design of NANA OCAM promoiety........................................................................44 2-2 Structures of N-methyl-N-met hyloxycarbonylaminomethyl prodrugs of dimethylaminobenzoic acid and 6MP......................................................................47 2-3 Synthesis of NANAOCAM prodrugs......................................................................47 2-4 Synthesis of NArNAOCAM prodrugs.....................................................................48

PAGE 13

xiii 2-5 Hydrolysis of NANAOCAM-Y in aqueous buffers.................................................60 2-6 Pseudo-first order rate constants (sec-1) versus of parent phenol.......................62 2-7 Pseudo-first order rate constants (s ec-1) versus acidic pKa of parent.....................62 2-8 pH rate profiles of compds 7 and 14 ........................................................................63 2-9 SN2 type of hydrolysis of NANAOCAM-Y with Yor carbamate as the leaving group.......................................................................................................................... .64 2-10 SN1 type of hydrolysis of NANAOCAM-Y.............................................................64 2-11 Hydrolysis of N-alkylamidomethylcarboxyl ic acid esters in aqueous buffers........65 2-12 Hydrolysis of N-imidomethyl deri vatives of phenols in aqueous buffers................66 2-13 Attempts at the synthesis of NA NAOCaminoethylidene chloride and NANAOCaminobenzylidene chloride.....................................................................68 2-14 Hydrolysis of NArNAOCAM-carbo xylicacid conjugates at pH 9.2.......................73 3-1 Synthesis of NANAOCAM-Cl from 1, 3, 5-trialkylhexahydrotriazine...................76 3-2 Preperation of alkyl chloroformate in situ from alcohol and synthesis of NANAOCAM-Cl from alkyl amine.........................................................................77 3-3 A Franz diffusion cell...............................................................................................86 3-4 Plot of solubility parameter versus log PMIPM for 1 17 20 ......................................93 3-5 Experimental versus calculated log maximum flux values through hairless mouse skin from IPM using equation 3.15..............................................................96 3-6 Experimental versus calculated log maximum flux values through hairless mouse skin from IPM using equation 3.16..............................................................97 4-1 Carbocations formed as intermedia tes by hydrolysis of Mannich bases of theophylline and NANAOCAM-Th.......................................................................100 4-2 Synthesis of NANAOCAM-Cl from 1, 3, 5-trialkylhexahydrotriazine.................102 4-3 Synthesis of NANAOCAM-Cl from al kyl amine and preperation of alkyl chloroformate in situ from alcohol.........................................................................104 4-4 Alkylation of theophylline with NANAOCAM-Cl................................................106 4-5 Plot of solubility parameter versus log PMIPM for 21-26........................................120

PAGE 14

xiv 4-6 Experimental versus calculated log maximum flux values through hairless mouse skin from IPM using equation 4.13 for n = 69............................................124 4-7 Experimental versus calculated log maximum flux values through hairless mouse skin from IPM using equation 4.14 for n = 74............................................125

PAGE 15

xv Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PRODRUG STRATEGIES AIMED AT IM PROVING TOPICAL DELIVERY OF DRUGS USING THE N-ALKYL-N-ALK YLOXYCARBONYLAMINOMETHYL (NANAOCAM) PROMOIETY By Susruta Majumdar August 2006 Chair: Kenneth. B. Sloan Major Department: Medicinal Chemistry Topical delivery is an attr active route to deliver drugs into systemic circulation; however the poor biphasic solubility of drug molecules limits delivery across the skin. NAlkyl-N-alkyloxycarbonylaminomethyl (NANA OCAM)-drug conjugates were designed as prodrugs with good biphasic so lubilities to increase the topical de livery of phenol and imide containing drugs. Prodrugs are biological inactive deri vatives of a drug which hydrolyse in vivo to the parent drug molecule. These derivatives mask polar functional groups present in a drug molecule which, in this case, leads to incr eased solubility of the drug in the skin. Prodrugs must hydrolyse to exhibit their pha rmacological activity. To elucidate the mechanism of chemical hydrolysis, a seri es of NANAOCAM conjugates of phenols, thiols and carboxylic acids were synthesized and their rates of hydrolysis determined in aqueous buffers. The hydrolysis followed ps eudo-unimolecular firs t order kinetics and was dependent on the nucleofugacity of the leaving group. The hydrolysis was also

PAGE 16

xvi independent of the pH of buffers. To furthe r elucidate the mechanism of hydrolysis, Naryl-N-alkyloxycarbonylaminomethyl (NAr NAOCAM) conjugates of phenols and carboxylic acids were also synthesized and thei r rates of hydrolysis determined. Since the hydrolysis of NArNAOCAM conjugates were slower than NANOCAM conjugates, the NANAOCAM conjugates are proposed to hydrolyse by a SN1 type of pathway with the lone pair on the nitrogen stabilizing the carbocation formed as an intermediate. To investigate if the NANAOCAM promoiet y increases the dermal delivery of phenolic drugs and imide containing dr ugs, a homologous series of NANAOCAMacetaminophen and NANAOCAM-the ophylline were synthesi zed. These derivatives were characterized by determin ation of their solubilities in IPM and water, partition coefficients between IPM and pH 4.0 buffer and flux through hairless mouse skins from IPM. Only two prodrugs of acetaminophen, C1 alkyloxy and C2 alkyloxy, increased flux through the skin. These derivatives were the mo st water soluble prodrug s in the series of more lipophilic prodrugs. In the theophylline series, only one derivative, C2 alkyloxy, increased delivery through the skin. This de rivative was the most lipid and most water soluble member of the series. The fl ux of NANAOCAM prodrugs from IPM was accurately predicted by the Roberts-Sloan equation.

PAGE 17

1 CHAPTER 1 INTRODUCTION The Skin Skin is the largest organ in the human body, accounts for 10% of the total body weight and has a surface area of 2m2 (Schaefer and Redelmeir, 1996). It acts as a physical and chemical barrier, protecting the body from the surrounding environment. Besides this, it enables the body to control fluid lo ss, regulate body temper ature and intercept external stimuli. The skin can be divided in to three distinct laye rs, the hypodermis, the dermis and the epidermis (Block, 2000). A micr oscopic view of skin is shown in Figure 1-1. The hypodermis is made up of a network of connective fibers and adipocytes which play an important role in energy storage a nd metabolism as well as providing insulation and protection against injury. Th e intracellular fat droplets may also act as a reservoir of hydrophobic compounds which may have penetrated the stratum corneum (the outermost layer).The hypodermis is 1-2 mm thick. It houses blood vessels whose prime function is to deliver nutrients to the skin and rem ove waste products, metabolites and xenobiotics from the skin. The dermis is the thickest layer of the skin and is 2 mm thick. It is responsible for maintaining the structural integrity of the skin. The dermis consists of a network of collagen and elastin fibers which provide flex ibility and tensile st rength to the dermis. Proteoglycans fill up most of the intracellu lar space and these macromolecules are responsible for the water-retaining propertie s of the dermis. These are polysaccharides

PAGE 18

2 covalently bound to a polypeptide backbone. The polysaccharide side chains are frequently sulfated. Examples of proteoglycans frequently seen in the dermis are heparin, Figure 1-1 Cross sectional view of human sk in (Reproduced with permission from K.G. Siver, Ph.D. Dissertation, Un iversity of Florida, 1987). chondroitin sulfate and dermatin sulfate. Th e anionic charge associated with these proteoglycans also allows a lot of water to be associated with the carboxylate or the sulfate groups and impedes th e ability of lipophillic compounds to cross the dermis. The dermis has an extensive vascular network wh ich participates in various processes like nutrition, exchange of gases, repair of ti ssues, immune responses and thermoregulation. Nerve endings sensitive to vibration, pressure pain and temperature are seen in dermis along with mast cells, macrophages and t-cell s which are present for protection against foreign antigens. A variety of appendages which permeate the stratum corneum like sweat glands, hair follicles and sebaceous gl ands are derived from this region. The hair

PAGE 19

3 follicles and exocrine sweat glands have pores that extend to the skin surface and provide openings to the epidermis which contains the stratum corneum: the main permeability barrier of the skin. Thus xenobi otics can use this route, also called the shunt route, to penetrate the dermis; this is especially true for highly polar molecules whose permeability is limited by the stratum corneum. However these pores constitute only 0.1 % of the total skin surface area and are responsible for early diffusion processes (Scheuplein and Blank, 1971; Stuttgen, 1982). Sebaceous glands are loca ted within the hair follicles and secrete sebum which makes its way on to the skin surf ace. Sebum is a complex mixture of fatty acids, triglycerides, squalene and waxes. Eccrine sweat glands are also present in the dermis and their primary purpose is thermore gulation. Since the epidermis is devoid of any vasculature, it depends on the dermis fo r supply of nutrients, oxygen and removal of waste metabolites. The presence of papillae fr om each layer interlocked with one another at the border of the dermis a nd epidermis facilitates this pr ocess of exchange between the two layers by increasing the surface area. Th e papillary layer is 50 nm thick and has capillaries which feed the epidermis with the required nutrients (Bisset, 1987). The epidermis, contains the stratum corneu m, provides the main barrier function to the skin and prevents the entry of foreign particles and loss of water. Keratinocytes (keratin producing cells) make up most of the viable epid ermis which is 50-100 m thick. This protein is very polar with every th ird amino acid in the backbone of keratin containing an ionizable side group (-COOH, -NH2 and –SH) (Flynn, 1990). Besides the keratinocytes the viable epidermis also has melanocytes which are responsible for skin pigmentation, langerhans cells for immune response and merkel cells responsible for sensory reception. The epidermis also acts as a store for antioxidants like ascorbic acid,

PAGE 20

4 tocopherols and tocotrienols, glutathi ones, urates, ubiquinone-10 and enzymatic antioxidants like catalases, superoxide dismut ases, glutathione peroxidases. The primary function of these antioxidants is to offer prot ection to the skin against the ultra violet radiation of the sun (Thiele et al., 2002). Division and differe ntiation of the keratinocytes lead to the formation of five distincts layers within the epidermis: stratum basale, stratum spinoum, stratum granulosum, stratum lucidum and finally stratum corneum. The cells migrate from the basal cell layer to the skin surface while gradually changing their structure and function resulting in the formation of physically and chemically resistant cell remnants called corneocytes which ma ke up the stratum corneum. Once on the surface of the skin the cells are shed off as a part of desquamation. This cycle of cell division followed by differentiat ion, proliferation and migrati on from basal to the surface and finally desquamation repeats itself over and over again with the new cells rapidly replacing the old ones. The transition from a terminally differentiated keratinocyte to corneocyte takes 24 h in the epidermis. The process of cornification and desquamation is intimately linked; synthesis of stratum corneum occurs at the same rate as loss. The outermost layer of the epidermis is called the stratum corneum, and it is the main rate limiting barrier to drug delivery (Stuttgen, 1982; Berti and Lipsky, 1995). It is 10-20 m thick consisting of polyhedral flat and nonnucleated cells called the corneocytes (Berti and Lipsky, 1995).The pH of the stratum corneum lipids is 4-6.5. The stratum corneum has 5-20 layers of corneo cytes. These cells are remnants of the terminally differentiated cells called kera tinocytes which are found in the viable epidermis. Their cell organelles and cytopl asm have disappeared during the process of cornification. Corneocytes are composed of insoluble kera tins surrounded by a cell

PAGE 21

5 envelope stabilized by cross linked protei ns and covalently bound lipids. These corneocytes are further interconnected by lipids. The presence of these intracellular lipids is required for a competent skin barrier and follows a tortuous path. These lipids are synthesized and assembled into lamellar st ructures which surround the corneocytes. These lamellar bodies which envelope the st ratum corneum store a complex mixture of lipids and lipid-like materials like free fa tty acids, choleste rol and hydroxyamide derivatives called ceramides. Interconnecti ng corneocytes of the stratum corneum are polar structures such as corneodesmosom es which contribute to stratum corneum adhesion. The stratum corneum is not uniform ly homogeneous and the layers represent various stages of corneocyte and lipid maturation. In short, the stratum corneum can be visualized as a brick and a mortar model with the flattened cornified cells being the offset layers of bricks lying one over another and the lipid rich intracellular matrix as the mortar separating the bricks or corneocytes (Figur e 1-2). By physically stripping the outermost layer of the skin an increase in permeability of various ch emicals and water can be seen proving stratum corneum to be the main rate li miting barrier. Thus t opical approaches at increasing flux through the skin have relied on increasing the solubi lity of compounds in the stratum corneum (Sloan et al.1989, 1992 and 2003). It is primarily believed that drugs cannot penetrate the protein rich corneocytes cells but need to ta ke a tortuous path alongside the corneocyte s through the interconnecting lipid s (Michaels et al., 1975). Thus the drug enters systemic circulation through th is narrow channel of lip ids and need to be hydrophobic so that they can permeate thes e semi-fluid matrix lipid barriers.

PAGE 22

6 Figure 1-2 Brick and mortar model of drug absorption through the skin. Unlike cell membranes in most parts of the body which are composed of phospholipids, the skin is devoid of any phospholipids. However, the presence of functional groups which are capable of hydr ogen bonding in ceramides (Figure 1-3), free fatty acids, and cholesterol su lfate in the lamellar bodies which surround the corneocytes makes the stratum corneum hydrophilic as well as lipophilic. The stratum corneum is about 30% water by weight and most of it is due to the association of water with these polar groups seen in the lamellar lipids. Thus, drugs with good biphasi c solubility can permeate the stratum corneum better than drugs which are merely hydrophobic. Lipid matrix Epidermis Corneocyte Dermis

PAGE 23

7 O O O N H OH OH O N H OH OH O N H OH OH OH O O O N H OH OH OH O N H OH OH OH O N H OH OH OH OH O N H OH OH OH OH O N H OH OH OH Ceramide 1 Ceramide 2 Ceramide 3 Ceramide 4 Ceramide 5 Ceramide 6 Ceramide 7 Ceramide 8 Figure 1-3 Structures of ceramides seen in the lipid bilayers of the lamellar bodies. The heterogenous composition of the lipids and disorder in packaging of these lipids leadS to the formation of lipid micr odomains. Ruthenium te troxide staining of normal skin shows the presence of altern ate hydrophilic region s connecting lipophilic bands in the bilayer structur e (Hou et al., 1991). The struct ure and arrangement of the barrier is such that drugs need to pa ss through alternate layers of hydrophobic and

PAGE 24

8 hydrophilic regions, formed as a result of the pa cking disorder of the lipids present, to reach the systemic circulation and thereby must possess adequate lipid as well as water solubilities (van Hal et al., 1996). The experimental evidence for this concept of biphasic solubility being important in terms of perm eability through the skin was first provided by Scheuplein and Blank (1973) for the diffusion of a homologous series of alcohols from water through human skin in vitro.The al cohol which gave maximum flux was not the most lipid soluble but was the most water so luble. Sloan et al. (1997) confirmed these results in in vitro hairless mice skin. Topical Delivery Most drugs developed today ar e delivered through the oral route. This is primarily because it is economical for the company de veloping the drug. The delivery of the drug does not require a visit to a physician or a trai ned medical representati ve as would be the case for an intravenous or parental formulati on. Drugs given orally also have the ability to be given in large doses because of the high permeability of the enterocyte cells present in the GI tract. However, the pH of the stomach is acidic, so the stability of molecules is compromised. Oral bioavailability of drugs is sometimes limited by efflux of active drugs by Mrp or P-gp transporters which prevent drugs from reaching intestinal vasculature (Wacher et al., 1996; Watkins, 1997; Schinkel, 1997 and De Mario et al., 1998) and CYP 450 catalysed metabolism of drugs (Watkins, 19 92). The portal vein carries the drug from the intestine to the liver by a process known as “enterohepatic cycling” before the drug finally enters systemic circulation. The liver is the largest drug me tabolizing organ in the body and has enzymes like CYP 450, GSH/GST and UDP/UGT which do not have strict substrate structure specificity. These enzymes transform the drug molecule to an inactive

PAGE 25

9 metabolite or lead to the formation of toxi c by-products and hence limit the therapeutic effectiveness of the drug. To circumvent these problems associated with oral drug deliver y, an alternate and useful site of delivery can be the skin. De rmal delivery evades first pass metabolism of drugs because once the drug permeates the skin it reaches the systemic circulation and hence the bioavailability is not limited by me tabolism, and side effects arising out of toxic metabolites are minimized. Dermal drugs are more likely to reach systemic circulation intact because of the absence of high concentrations of drug-metabolizing enzymes like CYP 450 enzymes, GSH/GST (con centration in nmol amounts, Shindo et al., 1994) or P-gp efflux transporters. CYP isoz ymes have been shown to be induced in skin (Ahmad et al., 1996, 2004 and Swanson, 2004) and P-gp transporters are seen in epidermal cells but their primary function seem s to be limited to clearing out endogenous compounds (Laupeze et al., 2001). The concentra tion of the enzymes is low and there is no systematic arrangement of the enzymes and transporters on the surface of the skin so they cannot clear the drugs efficiently or aff ect the therapeutic efficacy and availability of drugs. Topical delivery can be used for delivery of potent molecules whose dose requirements are low or require slow and sustained serum concen tration levels of drug for a prolonged period of time. It is po ssible to deliver about 30 mg/4.9 cm2 of an appropriately soluble drug through the skin over a period of 24 h (Beall et al., 1996). Dermal delivery means delivery into the skin, i.e., the epidermal and dermal skin cells. Such delivery is particularly useful for skin cancers. Transdermal de livery means delivery through the skin into systemic circula tion. Drugs with high transdermal delivery invariably show high dermal delivery too, t hough it may be possible to increase dermal

PAGE 26

10 absorption by making a hydrolytically labile prodrug which hydr olyses to give a lipophilic promoiety whose permeation th rough the dermis is limited and hence selectively accumulates in the dermis. Skin is a useful target organ for skin disorders and local action. Attempts are currently being made to deliver soft estroge ns locally because estrogen receptors are expressed in the skin (Labaree et al., 2001 and 2003). The idea is to increase uptake locally and prevent systemic circ ulation and hence the side effects that may result from it. These locally active estrogens can be used to treat vaginal dyspare unia without the risk involved with systemic administration. These so ft drugs are inactivated by esterases so that their action is geographically limited to the site of action. Delivery of 5-FU through the skin enables treatment of actinic keratosis (Dillaha et al., 1965) and may be useful for psoriarsis (T suji and Sugai, 1972) and yet spare the body from most of its systemic effects. Nicotine, fentanyl, estradiol and nitroglycerine are used in some transdermal patches available in the market today. Some attractive targets for transdermal delivery include alcohol cessation agents like naltrexone, testosterone, and antioxidants like vitamin C and vitamin E. Currently, the annual US market for transdermal patches is greater than $3 billion dollars; many companies are thus recognizing the value of transdermal delivery (Prausnitz et al., 2004). The problems associated with topical delive ry include local irritation or allergic reactions on the skin because of formulati on contents and poor penetration of drugs through the stratum corneal ba rrier due to poor physicochemical properties of the molecule.

PAGE 27

11 Approaches to Increase Permeation across the Skin The three approaches commonly used in incr easing transdermal delivery of drugs are. 1. Mechanical or electromechanical me thods like ultrasound, iontophoresis and microneedles. 2. Penetration enhancers and hydration. 3. Prodrugs. Ultrasound can be used to increase the tr ansport of both low molecular weight compounds and macromolecules like insulin a nd interferon. It reli es on usage of low frequency waves which causes disorganiza tion of lipid bilayers and increased permeability (Mitragotri and Kost, 2004). Iontophoresis is the use of an electric curre nt applied across the skin to drive drugs through the epithelium. The flow of electric current increas es the permeability of the skin. The current does not pass through the skin uniform ly but transappendageally through pores or sometimes through lipid channe ls. The electrical potential provides an electromotive force that is capable of dr iving charged molecules through the stratum corneum (Riviere and Heit, 1997). The formati on of HCl and NaOH by th e electrolysis of NaCl leads to a lowering of skin pH and are responsible for the irritation associated with iontophoresis (Mitragotri et al., 1996). Microneedles have also been used for tr ansdermal delivery of macromolecules. Micron sized needles in the pa tch pierce the skin into 10-15 m of the upper stratum corneal layer so that they do not stimulate th e nerves found in the lower layers and hence cause no pain while increasing the permeability of the drug used (Henry et al., 1998).

PAGE 28

12 The second approach to increase permeati on is by using penetration enhancers and hydration. The two distinct mechanisms by wh ich penetration enhan cers work are, the ‘push’ and the ‘pull’ mechanisms. The ‘pus h’ mechanism relies on the use of volatile components in a formulation to drive a drug into the skin. Evaporation of the volatile components leads to a supersaturated soluti on of the drug with thermodyanamic activity greater than one and results in increased flux through the membrane (Kadir, 1987). The ‘pull’mechanism on the other hand increases topical absorption by th e use of vehicles which interact with the skin and decreas e the diffusional resistance by disrupting the barrier, i.e., increase the de gree of hydration, leach the li pid components in the stratum corneal barrier or increases the solubilizing cap acity of the skin, so that flux of drug can increase (Barry, 1987). It is possible to incr ease the flux of a compound by about 10 fold by merely altering the vehicle properties before damage to the skin occurs (Sloan, 1992). Commonly used penetration enhancers in clude azone, DMSO, propylene glycol and decylmethyl sulphoxide. Keratolytic agents like salicyclic acid also increase penetration through the skin but cause damage. Hydration of skin leads to an increase in elasticity and permeability of the stratum corneal barrier. Occlusion is the most comma n way to increase the hydration of the skin as it prevents loss of moisture from the surface of skin by evaporation. It also leads to an increase in temperature and humidity on the surface and thereby to increased permeability of the drug. Topical formulations with fats and oils like glycols, glycerols, paraffins, waxes, silicones and clay also increase hydration (Block, 2000). Prodrugs represent yet another way in increasing topical ab sorption. They are biologically inactive and labile derivatives of a drug which revert to the active drug

PAGE 29

13 molecule by chemical or enzymatic hydrolysis. In this the case the dr ug in the form of a prodrug has better solubility in the skin and he nce gets pulled into the skin. An advantage of prodrugs over penetration enhancers is that prodrugs are a 1:1 molecular combination of drug and the promoiety compared to pe netration enhancers wh ere excess amounts are used which may cause allergic reactions or local irritation. Prodr ugs thus increase solubility of the drug in the skin while pene tration enhancers increa se the solubalization capacity of the skin for a drug molecule. Co mbination of prodrug with a penetration enhancer in the same formulation may prove to be more useful (Waranis and Sloan, 1987) but hasnÂ’t been fully utilized yet. Theory of Percutaneous Diffusion The mechanism of permeation across human skin relies on passive diffusion which is driven by a difference in the concentrati on gradient across the membrane. Any solute diffusing through the skin shows an initial, nonlinear increase in drug concentration which represents a build up of the drug in ba rrier region. The emergence of drug from the barrier region is different from drug to drug because of the time required to saturate the membranes. This is also true for most controlled release systems like transdermal patches. This is called a lag-time effect. Once steady state is reached a linear increase of drug concentration is seen which represents a condition where the exodus of solute from the dermal side is equal to the mass of ma terial entering the epidermal side (Poulsen, 1971). By simply extrapolating the steady stat e line to the absci ssa it is possible to determine the time to reach steady state. It also gives an estimate of how much drug permeates through the transfollicular shunt rout e. This is very well illustrated by Figure 1-4.

PAGE 30

14 Figure 1-4 Illustration of a typical flux profile. The rate of mass transfer across a membra ne or flux (J) is proportional to the concentration gradient expressed across the membrane. Imagine two solution filled compartments containing different concentrations of some compound (CD and CR) and separated by a permeable membrane of thic kness h (Figure 1-5).The rate at which the drug diffuses across the membrane from one co mpartment to the other is described by FickÂ’s first law. Equation 1-1 shows the rela tionship in a mathematical form and forms the basis of FickÂ’s first law (Fick, 1855). J = (dM/dt) unit area = -D (dC/dx) (1-1) where dM is the amount in mass or moles of solute passing through the membrane in time dt, dC/dx is the concentr ation gradient within the membrane over infinitely small distances and D is called the diffusion coefficient (diffusivity): its units are area per unit time (cm2/s). The negative sign indicates that th e concentration decreases with distance.

PAGE 31

15 CD Donor compartment (CD>>C1) CR Receptor compartment (C2>>CR) h C1 C2 Figure 1-5 Two compartment diffusion model. To utilize Fick’s law experimentally, th e concentration in both compartments (CD and CR) are maintained constant so as to ha ve a constant concentration gradient (C1-C2) across the membrane once equilibrium is esta blished and sink conditions are maintained i.e., CR 0. The partial differential can be replaced by C1, C2 and h. The concentration gradient can further be simplified as dC/dt = (C1-C2)/h (1-2) where C1 is concentration of the solute in the donor side of the membrane, C2 is the concentration of the solute on th e receptor side of the membra ne and h is the thickness of the membrane as shown in Figure1-5. Co mbining (1-1) and (1-2) we get J = (dM/dt) = -D (C1 – C2)/h (1-3) Both C1 and C2 are difficult to measure be ing inside the membrane. However under sink conditions, C1-C2 C1, C1 = SSkin (solubility in the skin), J = JMV (maximum flux from a saturated solution of drug in a vehicle). To maintain sink conditions it is necessary that the concentra tion of drug in the donor compartment doesn’t change with time and hence a suspension in equilibrium with a saturated solution is generally used for in vitro diffusion experiments and excess drug is used in a transdermal patch formulation. This type of trans port process follows zero order kinetics.

PAGE 32

16 JMV = D SSkin (1-4) h SSkin = SV. K Skin: V (1-5) Inserting (1-5) in (1-4) JMV = (D K Skin: V SV) (1-6) h where SV is the solubility of solute in the vehicle, JMV/ (SV) = P: P = (K Skin: V D) (1.7) h P is known as the permeability coefficient and can be determ ined experimentally. Since DA/h is constant, permeability coefficient is proportional to vehicl e/membrane partition coefficient of the drug. The more soluble th e drug is in the vehicle, the lower the permeability coefficient for the delivery of the drug from that vehicle (Sloan, 1992). The partition coefficient between drug in vehicle and in the membrane can be calculated from theory using equation (1-8). ln K = [( i – v) 2 v 2– ( i – s) 2 s 2]Vi (1-8) RT where i is the solubility parameter of the drug, v of the vehicle (8.5 (cal/cm3)1/2) for IPM (isopropyl myristate) and s is the solubility parameter of skin (10 (cal/cm3)1/2), Vi is the molar volume of the drug (cm3/mol), R is the gas constant (1.98 cal/K) and T is the temperature (305 K) (Sloan et al., 1986; Sh erertz et al., 1987; Sl oan et al.,1986a). The solubility parameter of the drug can be dete rmined experimentally or from individual

PAGE 33

17 group contribution methods (Sloan et al., 1986; Sherertz et al., 1987; Sloan et al., 1986a; Fedors et al., 1974; Martin et al., 1985; Martin et al., 1985a). Mathematical Modeling of Flux through Human Skin. Derivation of Roberts-Sloan equation from FickÂ’s law of Di ffusion (Roberts and Sloan, 1999). When the vehicle is water, equation (1-7) which relates permeability P with partition coefficient between skin and the vehicle can replaced by equation (1-9). P = (K Skin: AQ D)/h (1-9) The diffusivity or diffusion coefficient can be estimated from molecular volume through the following relationship (Cohen and Turnbull, 1959). D = D0e-z MV (1-10) The partition coefficient between skin and the vehicle is difficult to measure and thus can be estimated from the partition coefficien t between OCT (octanol ) or IPM (isopropyl myristate) and water (AQ) where a lipid like vehicle (OCT or IPM) replaces skin, which is considered to be lipid-like. K Skin: AQ = (KOCT: AQ) y (1-11) Here y represents the difference between the pa rtitioning domain of th e skin with respect to the solvent that replaces it, in this case OC T. The closer y is to 1, the closer the solvent is to being is a good surrogate of stratum co rneum lipids, e.g., ether has a y value of 0.53 while OCT has a y value of 0.7. Thus, OCT mi mics the skin more closely than ether does. By substituting equations (1-10) and (1-11) in equation (1-9) we get equation (1-12) P = [(KOCT: AQ) y D0 e-z MV]/h (1-12)

PAGE 34

18 By collecting the constants, taking the log of both sides and repl acing MV (molecular volume) by MW (molecular weight), the Po tts-Guy equation (Potts and Guy, 1992) can be obtained. log P = y log KOCT: AQ – z MW + x (1-13) where x = Do/h The Roberts-Sloan equation previously known as the transformed Potts-Guy model can be derived from the Potts-Guy model. The Potts-Guy is useful in predicting permeability through the skin but its prime lim itation is that it correlates permeability with partition coefficient between a lipid vehi cle and polar vehicle a nd hence is inversely proportional to solubility of the drug in a polar vehicle like water. This has serious implications in drug design and misleads medi cinal chemists to synthesize more lipid soluble derivatives in order to optimize topi cal delivery of a drug. We will talk about the importance of biphasic solubility on topical deli very in the chapters that follow. A model for flux (J) instead of permeability (P) is more clinically relevant because it actually tells us how much solute in moles (amount) is go ing into the skin compared to permeability. Since: JMAQ = (P) (SAQ) (1-14) Inserting equation (1-14) in equation (1-13) we get log (J/SAQ) = y log KOCT: AQ – z MW + x (1-15) log JMAQ – log SAQ = y log KOCT: AQ – z MW + x log JMAQ – log SAQ = y log SOCT – y log S AQ – z MW + x log JMAQ = y log SOCT – y log SAQ –z MW + log SAQ + x log JMAQ = x + y log SOCT + (1-y) log SAQ – z MW (1-16)

PAGE 35

19 Equation (1-16) is commonly referred as the Roberts-Sloan equation ( RS ) and in it flux is proportional to both the lipi d and the water solubilitie s of a drug and inversely proportional to molecular weight. If IPM instead of water was used as th e vehicle in diffusion cell studies, because some prodrugs were unstable, a diffe rent derivation is required. Starting at equation (1-6), wh ich was previously derived, bu t using IPM as the vehicle and D = Do exp (-z MW) as in equation (1-10) we get JMIPM = Do exp (-z MW) K Skin: IPM.SIPM h Taking the log of both sides and combining constants (x = Do/h) we get log JMIPM = x + log K Skin: IPM + log SIPM – z MW (1-17) Since, K Skin: IPM = K Skin: AQ/ K IPM: AQ (1-18) And since equation 1-11 relate s the partition coefficient betw een skin and the vehicle to the partition coefficient between OCT and AQ; where the vehicle is IPM, we can replace OCT with IPM to give equation (1-19). K Skin: AQ = (K IPM: AQ) y (1-19) Taking the log of both sides of equation (1-18) and substituting (1-19) for KSkin:AQ we get equation (1-20) log KSkin: IPM = y log K IPM: AQ – log K IPM:AQ log KSkin:IPM = y log SIPM – y log SAQ – log SIPM + log SAQ (1-20) Substituting equation (1-20) into equation (1-17) log JMIPM = x + y log SIPM – y log SAQ – log SIPM + log SAQ + log SIPM – z MW

PAGE 36

20 Collecting the terms we arrive at the RS equation. log JMIPM = x + y log SIPM + (1-y) log SAQ – z MW (1-21) or log JMV = x + y log SLIPID + (1-y) log SPOLAR – z MW (1-22) Equation 1-22 is the general form of the RS equation. From equation 1-4 JMV = D SSkin (1-4) h Inserting equation 1-10 (D = D0 exp (MV)) into equation 1-6 we arrive at: JMV = D0 exp (MV) SSkin (1-23) h Taking the log of both sides and collecting the constants we get: log JMV = x + log SSkin – z MV (1-24) where x = D/h and z = Correlating equation 1-24 with equation 1-22 ( RS ) we get log SSkin = y log SLIPID + (1-y) log SPOLAR (1-25) Equation 1-25 illustrates that solubility in skin can be modeled by solubility in two phases: solubility in a lipid phase and solubility in a polar phase. Biphasic solubility thus becomes an important determinant in optimizing flux through the skin. The RS equation has been used to predict th e topical delivery of homologous series of prodrugs across hairless mouse skin from their suspensions in isopropyl myristate (Wasdo and Sloan, 2004) and water in vitro (Sloan et al., 2003) and permeation of nonsteroidal anti-inflammatory drugs through human skin from mineral oil in vivo

PAGE 37

21 (Wenkers and Lippold, 1999 and Roberts and Sloan, 2001), permeation of solutes, chemicals and drugs from water in vitro (Flynn, 1990 and Majumdar et al., 2004) and penetration of sunscreens and antimicrobials in in vivo human skin from PG/water (Leweke and Lippold, 1995 and Majumdar and Sloan, 2005). In all cas es a significant dependence of flux on water solubility was obs erved regardless of the vehicle used. Table 1-1 shows the coefficients of x, y, z and r2 of these databases. Th e RS equation can used to predict flux of any drug molecule acr oss skin from any vehicle if three physicochemical parameters SLIPID (solubility in a lipid lik e vehicle like OCT or IPM), SPOLAR (solubility in a polar vehicle like wa ter or propylene glycol) and MW are known. Table 1-1 x, y, z, r2 and Average Residual Errors fo r Various Databases Fit to RS. Fit to RS (see text) Model Database Vehicle n x y z r2 logJM in vivo Human Skin Wenkers and Lippold (1999) Mineral oil 10 -1.459 0.722 0.00013 0.93 0.133 in vitro Hairless Mouse Skin Sloan and coworkers (2003) Water 18 -1.497 0.66 0.00469 0.77 0.193 in vitro Hairless Mouse Skin Wasdo and Sloan (2004) Isopropyl Myristate 61 -0.491 0.52 0.00271 0.91 0.15 in vivo Human Skin Flynn (1990) Water 103 -2.571 0.56 0.00444 0.9 0.44 in vivo Human Skin Leweke and Lippold (1995) 30% PG/ water 10 -2.116 0.45 0.00048 0.97 0.11 Prodrugs The word prodrug was coined by Adrien Albert to descri be compounds that undergo biotransformation prior to eliciting th eir pharmacological effects (Albert, 1958). It is impossible to review prodrugs of all cl asses; we will therefore cover only the basic

PAGE 38

22 principles involved in prodrug de sign and their applications in clinical practice (Ettmayer et al., 2004). Prodrugs have been primarily designed to increase oral bioava ilability, solubility, enhance chemical stability, prevent prematur e metabolism, decrease toxicity and improve taste (Stella, 1985). The primary function of a prodrug is to mask a polar functional group (-XH) where X can be –OH (phenolic or alc oholic), -COOH, -SH, -NH (amide, imide or amine) in a transient manner so that once the pr odrug is in the target site (which may be a tissue, cell or membrane) it hydrolyses to release the active drug molecule (Bundgaard, 1991). Since the polar functional group is mask ed, the tendency of the drug molecule to form intramolecular hydrogen bonding is also blocked. This leads to improved solubility properties of the prodrug in both lipid a nd aqueous phases with respect to the drug (Bundgaard, 1991). By merely increasing biphasic solubility of drugs it is possible to improve absorption and permeability across biological membrane barriers like the enterocyte cells of the GI tract or stratu m corneum of skin (Sloan and Wasdo, 2003). An increase in absorption and permeability resu lts in increased bioa vailability. Prodrugs containing charged promoieties like phosphate s, amino acids, hemisuccinates and amines have often been used in the promoiety to in crease the aqueous solubility of drugs which leads to increased dissolution rates and e nhanced oral (Fleisher et al., 1996) and transdermal absorption of the drug (Sloan and Wasdo, 2003). Masking of a functional group transiently also may allow drugs to evad e P-gp mediated efflux across the GI tract. (Jain et al., 2004). To make a hydrolytically la bile prodrug it is necessary to atta ch the drug to a promoiety which is stable enough to reach the ta rget site but labile enough to release the

PAGE 39

23 drug efficiently. Most prodrugs utilize an es ter promoiety because of the presence of esterases which are largely nonspecific and found in most tissues, biological fluids, organs and blood etc, and which hydrolyse the prodrug to give the corresponding drug (Beamount et al., 2003). A sc hematic representation is shown below (Figure 1-6). X O R' XH R'COOH O OR COOH ROH DRUG ESTERASES DRUG + DRUG DRUG + ESTERASES X = N, O, S; R = ALKYL; R' = OALKYL, N-ALKYL Figure 1-6 Esterase mediated hydrolysis of prodrugs. Carbonates of drugs have been made where the esters were too labile. Insertion of an oxygen atom in the promoiety led to grea ter stability of the corresponding prodrug in vivo because of the +I effect imparted by the oxygen atom which decreases the electrophilicity of the carbonyl group towards esterases. Si milarly a nitrogen atom can be inserted in the promoiety if a more stable derivative is desired. Since a prodrug needs to hydrolyse to be e ffective, the inability of esterases to efficiently transform penicillin esters to peni cillin led to the design of soft alkylated derivatives which are alkyl carbonyloxymethyl (ACOM) pr odrugs. The ACOM promoiety relies on a ‘OCH2’ spacer between the drug functi onal group and the labile ester functionality. Penicillin esters are steri cally hindered towards enzymatic hydrolyses however insertion of a ‘OCH2’ spacer made it easier for the esterases to freely access the ester functionality and hydr olyze it efficiently (Jansen and Russell, 1965).These soft alkylated derivatives rely on esterase hydrolysis to yield a hydroxymethyl derivative of

PAGE 40

24 the parent drug which is chemically lab ile and yields the parent molecule and formaldehyde (Figure 1-7). These types of promoieties rely on both enzymatic and chemical hydrolyses to release the drug molecu le at the target site In cases where the ACOM moiety was not sufficiently stab le, the alkyloxycarbonyloxymethyl (AOCOM) promoiety have been used where the ester is replaced by a carbonate functional group. The use of soft alkylation has been extended to other drugs containi ng various functional groups like imides (Bodor and Sloan, 1977; Buur and Bundgaard, 1985 and Taylor and Sloan, 1998), amines, phenols (Sloan et al., 1983a and 1983b; Bundgaard, et al., 1986 and Seki et al., 1988); alcohols (Beamount et al.), thiols (Sloan, 1983) etc. O O O O R O O OH COOH O O R X X OH XH DRUG ESTERASES DRUG DRUG -HCHO DRUG ESTERASES DRUG DRUG -HCHO X = N, O, S; R = ALKYL or OALKYL Figure 1-7 Soft alkyl prodrug hydrolysis. One of the critical issues involving prodrugs is that they should be labile but if they are too labile they wonÂ’t reach the target s ite and their purpose will be lost. Thus it is important to design derivatives accordingly and use esters or carbonates (ACOMs or AOCOMs) depending on the problem in hand a nd fine tune stabil ity to meet design directives. Sometimes it is necessary to rel ease drugs slowly and in such cases prodrug derivatives need to be more st able so that the residence time in the body is enhanced. We will look at some examples of such prodrugs later on. One of the prime reasons for esters being so widely used as prodrug promoieties is that they are easy to make and are economi cal. They release non-toxic side products on

PAGE 41

25 hydrolysis which is an importa nt factor in drug developmen t. Soft alkylated prodrugs release formaldehyde which is relatively innocuous and biocom patible since it is released as a byproduct during O or N-dealkylation of drugs by CYP 450 catalyzed oxidative reactions in liver and intes tines (Beamount et al., 2003). Another type of soft alkylating promoiety which has been useful in increasing oral bioavailability of phenolic dr ugs like estradiol is O-imidomethyl saccharin. The oral bioavailability of estradiol is drastically re duced by sulfation and glucuronidation of the phenolic functional group which leads to ra pid elimination of the drug from the body before or after it is absorbed. Such premat ure metabolism can be prevented by transiently masking the phenolic group with an imidom ethyl group such as O-imidomethyl-saccharin (Figure 1-8). The potency of the drug deliver ed as a prodrug was increased by about 8 times compared to an equimolar dose of estr adiol alone (Patel et al., 1991). The prodrug has a half life of 100 min under physiological conditions and the mechanism of hydrolysis is believed to be SN2 (Getz and Sloan, 1993), al though others suggest an addition to the carbonyl group followed by a vi nylogous elimination (Iley et al., 1998). Both mechanisms may be operating. This t ype of prodrug approach is independent of enzymatic hydrolysis and can be useful in cas es where enzymatic va riability becomes an issue; the ester promoiety is enzymatically too labile or efficient hydr olysis to the active principle is limited by strict structure specificity of the enzymes as in case of blood (Beamount et al., 2003). Although pharmaceutical scientists have often utilized esterases to biotransform a prodrug to its active parent drug, human alkaline phosphatases are another class of

PAGE 42

26 O N S O O O OH O O H NH S O O O O H -HCHO + SN2 S N O O O O OH O OH S O O N OH O H -OH + Figure 1-8 Hydrolysis pathways for Sacch arin based O-imidomethyl derivative of estradiol. enzymes which are found throughout various ti ssues (Strigbrand a nd Fishman, 1984) and which have been used by chemists to dephosphorylate phosphor yloxymethyl prodrugs. Fosphenytoin (Stella, 1996) is one such dr ug which is activated by phosphatases to give hydroxymethyl phenytoin which spontaneously loses formaldehyde to generate phenytoin (Figure 1-9). Delivery of phenytoin was a probl em because of its poor aqueous solubility

PAGE 43

27 (20-25 g/mL) and was formulated at pH 12 when given parentally; crystallization of drug at the injection site was seen. The a queous solubility was enhanced by using the phosphoryloxymethyl promoiety by about 4410 ti mes. The water solubility of the prodrug increased to 142 mg/mL (88. 2 mg/mL equivalent of phenytoin). O O N N H O P O H O O H O O N N H O H O O N H N H FOSPHENYTOIN PHOSPHATASES -HCHO PHENYTOIN Figure 1-9 Hydrolysis of phosphoryloxymet hyl prodrug mediated by phosphatases. This promoiety has also been used to incr ease the solubility of amines and alcohols (Safadi et al., 1993 and Krise et al., 1999). Thus phosphoryloxymethyl is another example of soft alkylating promoiety. The ge neric chemical structure of soft alkylating promoities is shown below. Phosphoryloxymethyl has two X’’’R’s, i.e., X’-P=X”(X”’R’)2.

PAGE 44

28 X R X' Y X'' X''' R' DRUG X = N, O, S; R = H or ALKYL; X' = N or O; Y= C or P; X" = O; X"' = ALKYL, N or O; R' =ALKYL and H The following discussion about the use of pr odrugs to increase delivery covers some miscellanenious examples which include the use of a novel enzyme system (amidoreductases), use of a water solubl e polymer (polyethylene glycol) and a polypeptide (Tat protein). Amidoximes are bioconverted to amidines by reductases expressed in the kidneys, liver, brain, lungs and GI tract (Clement, 2002). The double prodrug Ximelagatran relies on amidoreductases and esterases to yield the active drug melagatran. N H N O N H N N H2 OH O O N N O N H N N H2 H O O H H H MELAGATRAN + + XIMELAGATRAN Melagatran is a charged mo lecule at physiological pH with the carboxylate anion (acidic pKa ~ 2), amidine (basic pKa ~ 11. 5) and secondary amine (basic pKa ~ 7) functional groups making it very hydrophilic. As a result absorption acr oss the GI tract is decreased and its oral availability compromi sed. In the prodrug, the carboxylate ion is blocked as an ester. Since the ester is electr on withdrawing the pKa of the secondary amine is reduced to 4.5 and it is no longer charged at physiological pH. The amidine is protected as an amidoxime whose basic pKa wa s reduced to 5.2 so it is no longer charged

PAGE 45

29 either. Thus, the resultant molecule is ne utral at physiological pH, 170 times more lipophilic based on K values and about 3-6 fold more bioavailable as a result of increased absorption (Gustafsson et al., 2001 and Clement and Lopian, 2003). In the prodrugs we have talked about till now, the drug was attached to the labile promoiety directly or through a linker. In th e following example, a drug is conjugated to the linker (which should be a labile and bifunctional group), through a bridging functional groups, the linker in tu rn is attached to a promoi ety (which may be something that increases the solubility of the drug or increases its uptake). This type conjugate is a new class of prodrugs called tripartate prodrugs. Amino acids, aminoethoxyalcohols, hydroxybenzylalcohols, aminobenzylalcohols and o-hydroxyphenylprop ionic acids have all been used as bifunctional linkers through which PEGs can be transiently attached to drugs (Greenwald et al., 1999 and 2000). Th e active drug molecule is released by hydrolysis in two steps. An example utiliz ing hydroxyl-benzylalcohol as the linker and CO2 as the bridging functional group is shown below for an amine drug like Daunorubicin (Figure 1-10).These polyethylenegl ycol ether (PEGs) prodrugs have been used to increase the systemic circulati on time. Greenwald and coworkers (1996, 2001, 2003) have evaluated a series of PEGyla ted conjugates of anti cancer drugs like Campothecan and Taxol with PEGs of various molecular weights. PEG40000-AlaCampothecan (Prothecan) is presently in phase II clinical trials for treatment of lung, pancreatic and gastric cancers. Here the mo lecule has the aminoacid alanine acting as a linker between the Campothecan and PEG (Figure 1-11).

PAGE 46

30 O O PEG O O NHDRUG O N H DRUG O O O CO2DRUG-NH2 OH2 OH O H ESTERASES + + Figure 1-10 Hydrolysis of PE G conjugated to drug using a hydroxyl-benzylalcohol linker in vivo. O N O N O O O N H O O PEG OH N O N O O Figure 1-11 PEG-Ala-Campothecan These conjugates have higher water sol ubility than Campothecan which allows increased dissolution of the drug. Also, since the molecular weight of the conjugates is close to 45KDa (which is the renal threshol d limit), these drugs are retained in the body longer because of a phenomenon known as ‘e nhanced permeation retention’ (EPR) (Maeda et al., 1992). These prodrugs need to circulate in the body fo r longer periods and release the drug slowly in a c ontrolled manner; they have ha lf lives >10 h compared to other traditional prodrugs with half-lives of 20 min. These molecules evade glomerular filteration of kidneys which excretes molecu les whose size is 30KDa (Yamaoka et al., 1994). Thus the drug selectively accumulates in the cancer cells taking advantage of the

PAGE 47

31 leaky tissues and reduced drainage seen in cancerous tissues. This selectively towards cancerous cells leads to reduced toxicity a nd better targeting of anti cancer agents. Drugs have also been conjugated to oligoa rginine residues to increase their uptake across cell membranes and skin. Certai n proteins contain subunits like Tat 49-57 (RKKRRQRRR) of HIV-1, that en able the translocation of proteins across the plasma membrane into the cells (Green and Lo evenstein, 1988; Ande rson et al., 1993 and Lindgren et al., 2000). The intact Tat protein has been conjugat ed to proteins to enhance their delivery (Kim, 1997 and Nagahara, 1998). The chirality of the arginine residue (D or L) didnÂ’t affect cellular uptake. The Tat pr otein efficiently crosses membranes of cells in an energy dependent fashion after endocyt osis (Mann and Franke l, 1991; Vives, 2003 and Fuchs and Raines, 2004). There is some evidence for guanidium rich molecules undergoing receptor mediated endocytosis due their big size i.e. MW >3000. An alternative mechanism of uptak e is believed to be electrostatic interaction between the positively charged guanidine (pKa ~ 12-13) and the carboxylate ion of the lipid bilayer or phosphate group of the phospholipids bilayer to fo rm a transient ion pair complex (Figure 1-12) which is less polar than the drug-Arg conjugate (Luedt ke et al., 2003 and Rothbarg et al., 2002). The driving force for the passage through the membrane is voltage potential across most cell membranes. The positively charged complex formed due to incomplete ion pairing with anionic cell components moves along the direction of transmembrane potential (intracellular K+levels >>>extracellular K+ levels). Uptake is decreased when membrane potential decreases (e xtracellular and intracellular K+ levels equal) while it increases when a peptide antibiotic like valinomycin which selectively shuttles K+ across the membrane is used in the assay. The co mplexes dissociate on th e inner leaf of the

PAGE 48

32 membrane. Truncation studies and synthe sis of a series of analogues of Tat 49-57 revealed that short chain oligomers (7-9 residues) of arginines are more efficient in translocation than the entire protein (We nder et al., 2000). Wender and coworkers analyzed a series of Cyclosporin-Arg7 and Paclitaxel-Arg8 conjugates for uptake across human, mice skins and various in vitro cell line surrogates of plasma membrane. They found that it is necessary to have the guanidin e moiety; oligolysine, oligohis tidine or citru line residues were not as successful as oligomers of arginine (Mitchell et al., 2000). When the guanidine residues were al kylated the complex formation was hindered and uptake diminished. A bidentate ligand like guanidine is able to form stronger ionic complexes with a phosphate than the regular amine of lysine : as a result uptake is higher for arginine rich residues than other charged ami no acid residues (Rothbard et al., 2005). N H O N H NH2 N H P O O O O Figure 1-12 Ionic complexes formed by oligoargin ine of Tat with the phosphate groups in cell membranes. Too few arginine residues di minishes cell surface adhere nce while too many positively charged guanidines lead to reduced escape fr om the inner leaf of the membrane. The active drug is released from the conjugate hydrol ytically with half lives depending on the pH sensitive linker being used (Figure 1-13). CyclosporineArg7 conjugates release the active drug at pH 7.4 and 37C with a half lif e of 90 min. The mechanism of hydrolysis is intramolecular nucleophilic a ttack by the secondary am ine on the ester carbonyl functional group (Rothbard et al., 2000) Similarly in case of TaxolArg8 conjugates, the hydrolysis rates vary depending on the R group a ttached to the amine from 1 min for R =

PAGE 49

33 H to 107 min for R = Boc. Thus it is possible to control the rates of hydrolysis just by varying the R group from an electron dona ting to an electron withdrawing group (Kirschberg et al., 2003). N H O N O O CSA O Ph NHArg7-COOH N N O O O NHArg7-COOH Ph O O S RHN NH-Arg8-CONH2 O TAXOL RNS O O H2NCO-8grA-HN + CSA pH 7.4 pH 7.4 TAXOL + R = H, Ac, Piv, Boc; CSA Cyclosporin. Figure 1-13 Chemical hydrolysis of oligoarginine conjugates of cyclosporine and taxol. Prodrugs for Dermal and Transdermal delivery Although the previous paradigm (Guy and Hadgraft, 1988; Flynn, 1990) for increasing dermal delivery focused on increasing KOCT:AQ, most effective prodrug approaches aimed at improvi ng the delivery of drugs acro ss the skin have relied on increasing the solubility in a lipid vehicle (SLIPID) like octanol or IPM and the solubility in a polar vehicle like water (SAQ) of the prodrug over the parent drug. It is believed that an increase in biphasic solubility of the drug le ads to an increase in permeability of drugs across the stratum corneum (Sloan et al., 1984; Sloan, 1989 and 1992; Sloan and Wasdo, 2003). In a homologous series of lipophilic pr odrugs, the more water soluble of the more lipid soluble member of the series gave the highest flux (Sloan et al., 1984, 1984, 1992, 2003 and 2004). Thus the balance between so lubility in lipid and aqueous phases

PAGE 50

34 becomes paramount and flux is better modele d by a combination of these two parameters rather than K (partition coefficient). Previ ously it was shown that by merely masking polar functional groups transien tly it is possible to increa se biphasic solubility and thereby flux. While it is easy to see ho w addition of carbon atoms to the promoiety conjugated to the drug increases SLIPID, it is difficult to visualize why a simulataneous increase in SAQ occurs. For example, one may compare caffeine, theophylline and theobromine to illustrate this increase in SAQ (Windholz et al., 1983). Theobromine has two intramolecular hydrogen bonding sites; theophylline has one hydrogen bonding site while caffeine has none. A look at the melting points and SAQ of these three compounds reiterates the fundamental idea behind prodrug synthesis. Th e masking of polar functional groups leads to decreased intramolecular hydrogen bonding and in turn leads to a decrease in lattice energy, melting points and increase in biphasic solubilities. N N N N H O O N N N N O O N H N H N N O O CAFFEINE THEOPHYLLINE THEOBROMINE mp: 357 270-274 238 Saq: 3 mmol/l 46.55 mmol/l 112.64 mmol/l Codeine, the ether derivative of morphine is 40 times more water soluble than morphine and 2400 times more lipid soluble. Thus masking of a phenolic functional group increases solubility in both phases. There are two types of prodrug promoieties th at have been used in dermal delivery; the acyl based approach (N-acyl, O-acyl a nd S-acyl) and soft alkyl based approach (ACOM or AOCOM). In the discussion that follows we will look at the importance of

PAGE 51

35 solubilities in optimizing flux across the skin with examples. Consider the table below for series of 1-alkyloxycarbonyl -5-FU prodrugs and 5-FU with their physicochemical properties (Table 1-2). Melting points (mp), molecular weight (MW), log solubility in isopropyl myristate (log SIPM), log solubility in water (log SAQ), log partition coefficients between IPM and pH 4 buffer (log KIPM: AQ) and log maximum flux of total species delivered by the prodrug (and 5-FU) through hairless mice skins in vitro from an IPM donor phase (log JMIPM) (Beall et al., 1993) are given. Table 1-2 Physicochemical Characterization of 1-Alkyloxycarbonyl esters of 5-FU. Compound mp MW log SIPM log SAQ log KIPM:AQ log J MIPM C1 160 188 0.328 2.05 -1.72 0.42 C2 128 202 1.117 2.24 -1.12 0.77 C3 126 216 1.182 1.63 -0.45 0.36 C4 98 230 1.529 1.37 0.16 0.35 C6 67 258 2.186 0.7 1.48 0.19 C8 98 286 1.561 -0.89 2.46 -0.53 5-FU 284 130 -1.308 1.93 -3.24 -0.62 The melting points of the prodrugs decr ease with the addition of each CH2 unit from C1 to C6 while the melting point of C8 prodrug is higher than the C6 probably because of van der Waals interaction between the alkyl side chains (Yalkowsky, 1977). This increase in melting point al so leads to a decrease in SIPM. All members of the series are more lipid soluble than 5-FU. The C6 pr odrug derivative is the mo st lipid soluble of the series, has the highest K and SIPM. So, if SIPM were the most important determinant, then its log JMIPM should be the highest instead it is the ne xt to lowest in the series. On the other hand, the C2 prodrug has a higher lipid so lubility than 5-FU (not the highest) and highest SAQ in the series (1.8 times more water so luble than 5-FU) and gives the highest JMIPM (25 fold higher than 5-FU) across the sk in. Sloan and coworkers (1983, 1984 and 2003) have reported such dependence on bi phasic solubility fo r a series of 6-

PAGE 52

36 mercaptopurine (6MP), ThH and 5-FU prodr ugs. It was first observed for 7-ACOM prodrugs of ThH (Sloan et al., 1982) but not recognized as a new paradigm until later (Sloan et al., 1984). There are two schools of thought for optimizi ng flux across the sk in. The first relies on increasing the lipid solubility without incr easing aqueous solubility of the prodrug and uses increased KOCT: AQ as an indicator of increased li pophicity. Since skin is primarily a lipophilic barrier it relies on ma king a prodrug with higher K than the parent drug so that penetration across the barrier can be enhanced. In this school of thought, any dependence of flux on aqueous solubility is attributed to the use of a polar ve hicle, the nature of mouse skins (which are more hydrophilic than human skin), in vitro experimental conditions and metabolic ra tes of conversion. In in vitro cell experiments, the skin is abnormally hydrated particularly when water is used as a vehicle. Thus some argue that the dependence of flux on SAQ could be due to the necessity fo r increased solubility in the vehicle. This school of thought also argues th at lipid soluble prodru gs undergoing fast rates of hydrolysis to the more water solubl e parent drug bypasses the hydrophilic region of the skin dermis easily and permeates the skin better. However in case of lipid soluble prodrugs undergoing slower hydrolysis, the dermis becomes a rate limiting barrier and the prodrug gets trapped in the dermis, co mpromising delivery across the skin. Thus water solubility becomes important only if the designed prodrugs are stable to hydrolysis during their passage through the skin (Stin chcomb and coworkers, 2002 and 2005). The second school of thought relies on the importance of biphasic solubility in order to increase flux. It reli es on increasing both lipid and water solubility at the same time to optimize delivery. It is important to note that flux is inde pendent of metabolic

PAGE 53

37 conversion rates of prodrug to the drug because flux is dependent only on the solubility in the membrane (Equation 1.4). The observations that flux depends on SAQ and SLIPID holds true for prodrugs with a half -life of 1-5 min (7-AC-Th, Sloan et al., 2000) to prodrugs which permeate intact in diffusion cells experiments (1-ACOM-FU, >50% intact prodrug obtained, Taylor and Sloan, 1998) run on hair less mice skins. The argument against the dependence of water solubility due to the pres ence of stagnant water layers in presence of a polar vehicle or in in vitro conditions when the skin is heavily hydrated comes from experiments carried out with isopropyl myrist ate as a vehicle (a non polar vehicle in which the longer chain prodrugs were more soluble). The same dependence on SAQ was again observed where IPM was the vehicle, reemphasizing the fact that this dependence is a result of the inhere nt properties of the skin. In Vivo human skin (human skin is less hydrophilic than mouse) experiments using mi neral oil also show dependence on water solubility. Irrespective of the ve hicle used, the best performing prodrug of a series from a vehicle like IPM is also the be st performing prodrug from a di fferent vehicle like water. The best prodrug was the one with the best biphasic solubility. The increase in flux though varies from vehicle to vehicle dependi ng on the interaction of the skin with the vehicle. IPM causes more damage than water does; it leaches out the lipids from the stratum corneum hence compromises the barrier and leads to increased permeability of prodrugs compared to water (Sloan et al., 2003a). Water so lubility is important in improving flux across the skin whether it is hum an skin or mouse skin, polar or non polar vehicle, in vitro or in vivo, and prodrugs which are labile or stable to hydrolysis. In the 5-FU prodrug series we considered above, the prodrug exhibiting the best flux had a higher SAQ and SLIPID than 5-FU. It is not alwa ys possible to increase SAQ of the

PAGE 54

38 prodrug compared to the parent drug. In such cases, the best perfor ming prodrug is the one which shows minimum decrease in SAQ with respect to the parent drug, e.g. methyloxycarbonyl-APAP prodrug shows the highest flux through skin in the homologous series of alkyloxycarbonyl-APA P prodrugs investigated by Wasdo and Sloan (Table 1-3). This particular derivativ e was 6.3 times more lipid soluble and about 3.5 times less water soluble than APAP, yet was the most water soluble prodrug of the series being investigated. Table 1-3 Physicochemical Characteri zation of Alkyloxycarbonyl esters of Acetaminophen. Compound mp MW log SIPM log SAQ log KIPM:AQ log J MIPM C1 112 209 1.08 1.31 -0.16 0.0 C2 120 223 0.97 0.58 0.32 -0.76 C3 104 237 1.37 0.43 0.9 -0.45 C4 118 251 1.14 -0.43 1.5 -1.01 C6 108 279 1.22 -0.37 2.71 -1.49 APAP 171 151 0.28 2.0 -1.72 -0.29 Similarly in the 7-alkyloxycar bonyl-Th series (Sloan, 2000), C3-Th was most water soluble prodrug member of the lipid soluble series. This prodrug derivative was 1.3 times less water soluble than ThH and s till gave higher flux than ThH. Ever since SAQ of a drug was recognized as being important to optimize dermal delivery, attempts have been made by Sloan and coworkers (Sloan et al., 1984; Sloan et al., 1988; Saab et al., 1989 and 1990) to in corporate basic amino groups into the promoiety. The rationale behind introducing a polar functional group like amine into a promoiety is to increase the wa ter solubility of poorly solubl e drugs. Insertion of simple alkyl groups into the promoiet y does increase lipid solubil ity but increases in water solubility are modest and are restricted to the shorter alkyl chain members. Mannich bases (DrugX-CH2-NR2, X = S, N, O; R = alkyl) of 6MP, 5-FU and ThH were thus

PAGE 55

39 designed to improve SAQ of the prodrugs without decreasing SLIPID compared to the parent drug and hence to increase flux. Hussain and coworkers (Milosovich et al., 1993 and Jona et al., 1995) added polarisable amino groups to the alcohol portion of testosterone and indomethacin ester prodrugs. When the 4-dime thylaminobutyrate hydrochloride ester of testosterone was examined in vitro using human skin, a 35 fold flux enhancement was observed (Figure 1-14). Similarly when 2-di ethylaminoethyl group was built into the ester promoiety of indomethacin the SAQ was enhanced compared to the parent drug and flux through skin increased by a bout six fold (Figure 1-14). N O Cl O MeO O N(CH2CH3)2 O O N(CH3)2.HCl O Figure 1-14 Structure of testos terone and indomethacin este rs with polarisable side chains. Also, attempts have been made by Bonina and coworkers (1995) to increase SAQ and SLIPID by incorporating short ch ain polyethylene glycolic et hers into the promoiety. Thus NSAIDS esters with small PEGs have been evaluated for their flux across human skin. These polyethers did increase both SAQ and SLIPID, e.g. for the indomethacin esters addition of a -OCH2CH2 unit increased SAQ by 1.3 fold and SLIPID by 1.02 times and as a result flux also increased. Although all the derivatives were more lipid soluble than indomethacin, the prodrug containing n = 5 -OCH2CH2 units, which was 1.77 times more water soluble and 9 fold more lipid soluble than indomethacin, gave the highest flux, (Table 1-4).

PAGE 56

40 N O Cl O O(CH2CH2O)nCH2CH2OH MeO PEG-esters of Indomethacin Prodrug promoieties containing s hort PEGs can thus increase the biphasic solubility of lipid soluble drugs with very poor aqueous solubi lity. It is interesting to note even in this series of ‘ethylene oxy’ homol ogous series it is the more water soluble derivative that gave the highest flux just as in case of simple alkyl derivatives where a ‘CH2’unit is inserted for ‘O’along the series. Thus, the importance of the balance between SLIPID and SAQ of the prodrugs is important in optimizing delivery of drug molecules topically. Table 1-4 Physicochemical Characteriza tion of PEG esters of Indomethacin. COMPOUNDSSOCT (mmol mL-1102) SAQ (mol mL-1 102) JM (mol cm-2 102) Indomethacin 27.7 22 0.84 n = 1 221.9 6.7 0.4 n = 2 225.8 9.2 0.65 n = 3 240.1 17 0.73 n = 4 244.3 25 2.83 n = 5 251.8 39 3.12

PAGE 57

41 Research Objectives N-alkyl-N-alkyloxycarbonylamin omethyl (NANAOCAM) promoi ety has been used to make prodrugs of 6-mercaptopurine for topical delivery (Siver, 1990). The mechanism of hydrolysis however wasnÂ’t clearly established. The purpose of this research project was to explore the use of NANAOCA M promoiety as prodrugs of ot her functional groups like phenols, carboxylic acids, imides and evaluate its potential use for enhancing the skin penetration of model phenolic and imide containing drugs. 1. Design, synthesize, and investigate ra tes of hydrolysis and mechanism of hydrolysis of N-alkyl-N-alkyloxycar bonylaminomethyl (NANAOCAM) and Naryl-N-alkyloxycarbonylaminomethyl (NAr NAOCAM) prodrug derivatives of phenols and carboxylic acid. 2. Synthesis of homologous series of NANAOCAM prodrugs of acetaminophen (APAP, model phenolic drug) and their physicochemical characterization which includes: measuring solubilities in isopropyl myristate (IPM) and water, partition coefficients of prodrugs between IPM and pH 4.0 buffer, investigation of flux of compounds through hairless mice skins from IPM. 3. Synthesis of homologous series of NANAOC AM prodrugs of theophylline (ThH, imide containing drug) and their physicoch emical characterization which includes: measuring solubilities in isopropyl myristat e (IPM) and water, partition coefficients of prodrugs between IPM and pH 4.0 buffer, investigation of flux of compounds through hairless mice skins from IPM. 4. Mathematical modeling of flux These goals are discussed in three separa te chapters. Chapter 2, talks about the design of NANAOCAM promoiety, synthesi s of NANAOCAM-drug derivatives by alkylation of drug with N-alkyl-N-al kyloxycarbonylaminomethyl chlorides, determination of rates of hydrolysis in aqueous buffers, its implication on prodrug design leading to the synthesis and hydrolysis of N-aryl-N-alkyloxycarbonyl-aminomethyl (NArNAOCAM) prodrugs of phe nols and carboxylic acids.

PAGE 58

42 Chapter3 deals with alkylation of acetaminophen with N-alkyl-Nalkyloxycarbonylaminomethyl chlorides to synthesize a homologous series of NANAOCAM prodrugs, evaluation of physicochemical properties and transdermal and dermal penetration of these prodrugs and pred iction of flux of APAP prodrugs using the RS equation. Chapter4 deals with alkylation of the ophylline with N-alkyl-N-alkyloxycarbonylaminomethyl chlorides to synthesize a homologous series of NANAOCAM prodrugs, evaluation of physicochemical properties and tr ansdermal and dermal penetration of these prodrugs and prediction of fl ux of ThH prodrugs using the RS equation, and all NANAOCAM prodrugs when fitted to the existing n = 63 database.

PAGE 59

43 CHAPTER 2 DESIGN, SYNTHESIS, HYDROLYSIS OF N-ALKYL-N-ALKYLOXYCARBONYLAMINOMETHYL DRUG DERIVATIVES AN D ITS IMPLICATIONS ON PRODRUG DESIGN. Introduction Drugs containing polar functional gr oups pose problems of membrane permeability, solubility and premature metabolism which limit their oral and dermal delivery. The prodrug approach, which involves masking these polar functional groups as labile derivatives which then hydrolyze to the native drug either enzymatically or chemically, has proved useful in numerous cases (Bundgard, 1991 and Sloan and Wasdo, 2003). In most cases drug functional groups ar e masked as simple esters. Acyloxymethyl (ACOM, RÂ’COOCH2-) and alkyloxycarbonyloxymethyl (AOCOM, RÂ’OCOOCH2-) promoieties have been used to derivatize car boxylic acids in cases where the simple ester approach wasnÂ’t useful because, although si mple esters were reasonably stable chemically so that they could be convenie ntly formulated, they were not sufficiently labile enzymatically (Jansen and Russell, 1965 ; Bundgard et al., 1986; Seki et al., 1988 and Beaumont et al., 2003). Similarly, simple alkylation of phenolic groups with alkyl halides gives derivatives that are not efficiently reversible in vivo, while ACOM or AOCOM gives derivatives that are too unstable during the or al absorption process to protect the phenolic drug from premature metabolism. Replacing the oxygen atom in OCH2 with nitrogen (N-R, R = alkyl) in RÂ’OCOOCH2to give a N-alkyl-Nalkyloxycarbonylaminomethyl (NANAOCAM, RÂ’OCONRCH2-) promoiety could give medicinal chemists an additional handle and flexibility to improve solubility (better

PAGE 60

44 balance between solubilities in lipid and wate r) and stability (enzymatic versus chemical) of prodrugs. The evolution of NANAOCAM promoiety from simple esters to ACOM/AOCOM to NANAOCAM is shown in Figure 2-1. O O O R N H O O R N O O R R' R O O O R O R O ESTER OR CARBONATE-DRUG CONJUGATE Insert 'CH2O' Spacer Insert 'O' 'CH2O' Spacer DRUG-X AOCOM DRUG CONJUGATE X = O, S, N DRUG-X alkylate N DRUG-X R= ALKYL N-ALKYL-N-ALKOXYCARBONYLAMINOMETHYL DRUG CONJUGATE (NANAOCAM) X = O, S, N DRUG-X DRUG-X DRUG-X ACOM DRUG CONJUGATE X = O, S, N R= ALKYL R, R' = ALKYL Replace 'O' with 'N' Figure 2-1 Design of N ANAOCAM promoiety. Only one example of the use of a NA NAOCAM promoiety has been reported: 6mercaptopurine (Siver and Sloan, 1989). Th e use of a close analogue of NANAOCAM (RÂ’CONRCH2-) in which the promoiety is an amid e instead of a carbamate has been reported for carboxylic acids (Bundg ard et al., 1991; Moreira et al., 1996 and Iley et al., 1997). Here we extend the use of the NANAOCA M promoiety to other drug functional groups and investigate the mech anism of chemical hydrolysis and its implications in

PAGE 61

45 prodrug design of these soft alkylat ed derivatives with the goal of being able to tailor the promoiety to give the appropri ate rates of conversion to th e parent drug based on the route of administration. In the discussion that follows we will talk about the synthesis of N-methyl-Nmethyloxycarbonylaminomethyl derivatives of phenols, carboxylic acids and 6mercaptopurine (6MP) and N-aryl-N-a lkyloxycarbonylaminomethyl (NArNAOCAM) derivatives of phenols and carboxylic acids. We will then investigat e the mechanism of hydrolysis of N-methyl-N-methylox ycarbonylaminomethyl and NArNAOCAM derivatives of phenols, carboxylic acids and 6MP. Synthesis of N-methyl-N-methyloxycarbo nylaminomethyl and N-aryl-N-methyloxycarbonylaminomethyl Prodrugs. Ten N-methyl-N-methyloxycarbonylaminomethyl and six N-aryl-Nmethyloxycarbonyl aminomethyl derivatives we re synthesized (Tables 2-1 and 2-2). Alkylation of the parent compound with N-methyl-N-methyloxycarbonylaminomethyl chloride or N-aryl-N-methyloxycarbonylaminom ethyl chloride was accomplished in the presence of a base like triethylamine with CH2Cl2 or DMSO as the solvent. In every case it was necessary to synthesize the co rresponding alkylating agent: N-methyl-Nalkoxycarbonylaminomethyl chloride or Naryl-N-methyloxycarbonylaminomethyl chloride.

PAGE 62

46 Table 2-1 N-methyl-N-methyloxycarbonylaminom ethyl-phenol Conjugates Synthesized. O X N R O O CH3 Z Compd X Z R 1 NHCOMe H Me 2 CN H Me 3 CHO H Me 4 H CHO Me 5 COMe H Me 6 COOMe H Me 7 NO2 H Me 8 NO2 H 4Â’-C6H4-OMe 9 NO2 H 4Â’-C6H4-COOEt 10 NO2 H C6H5 Table 2-2 N-methyl-N-methyloxycarbonylam inomethyl and NArNAOCAM Conjugates of Naproxen. CH3O O O N R O O CH3 Compd R 11 Me 12 C6H5 13 4Â’-C6H4-OMe 14 4Â’-C6H4-COOEt

PAGE 63

47 N O O N R O O R' N N N H N S N R O O R' R = R' = CH3R = R' = CH3 15 16 Figure 2-2 Structures of N-methyl-N-m ethyloxycarbonylaminomethyl prodrugs of dimethylaminobenzoic acid and 6MP. CH2=O RNH2 NN N R R R NN N R R R O O Cl R' O O N R Cl R' NaOH O O N R Cl R' Y-H TEA O O N R Y R' CH2Cl2CH2Cl2R, R' = CH3 + + + Y = 6MP,phenol or carboxylic acid Compounds 1-7, 11, 15, 16 Figure 2-3 Synthesis of NANAOCAM prodrugs

PAGE 64

48 NH2 X N H X O O CH3 X N Cl O O CH3 X N Cl O O CH3 (CH2O)n X N Y O O CH3 Naproxen p-nitrophenol CH3OCOCl X = OCH3, COOC2H5, H. YH = Naproxen, p-nitrophenol. Py TMSCl TEA or Compounds 8-10, 12-14 Figure 2-4 Synthesis of NArNAOCAM prodrugs N-Methyl-N-methyloxycarbonylaminomethyl chloride synthesis (Figure 2-3): The N-methyl-N-methyloxycarbonylaminomet hyl chloride was synthesized as reported by Siver et al., ( 1990) from 1, 3, 5-trimethylhexahydrotriazine. (a) 1, 3, 5-Trimethylhexahydrotriazine was s ynthesized from equimolar equivalents of aqueous formaldehyde, methyl amine and NaOH according to the protocol originally developed by Graymore et al., (1932) and modi fied by Siver et al., (1990). Methyl amine (0.4 mol, 40% aqueous) was placed in an i ce bath and an equivalent of 37% aqueous formaldehyde was added dropwise over a period of 10 min. The solution was allowed to equilibrate to room temperature and stirred for one hour, then an equivalent of NaOH was added and the contents were stirred for 1.5 h more. The solution was extracted with 4 50 mL CH2Cl2, dried over Na2SO4, filtered and concentrated to a clear colorless CH2Cl2 solution containing the hexahydrotriazine de rivative. Complete concentration of CH2Cl2

PAGE 65

49 wasn’t carried out as it resulted in some lo ss of the desired product. For quantification purposes the CH2Cl2 ‘methylene peak’at 5.3 and N-CH2-N peak of hexahydrotriazine derivative at 3.2 were used: yield = 82% in CH2Cl2, 1H NMR (400 MHz; CDCl3; Me4Si): 3.2(s, 6H), 2.3(s, 9H). (b) N-Methyl-N-methyloxycarbonylaminomethyl ch lorides were synthesized from 1, 3, 5-trimethylhexahydrotriazine by reacting it with three equivalents of methyl chloroformate in CH2Cl2 .To well stirred solution of methylchloroformate in CH2Cl2 cooled with an icebath was added an equivale nt of 1, 3, 5-trimethylhexahydro-triazine (freshly prepared) in CH2Cl2 over a period of 10 minutes. The white suspension that was observed was allowed to equlilibrate to r oom temperature and stirred overnight. The suspension was filtered and the filtrate concentr ated to oil. The oil contained the desired product and some of the correspondi ng bis (N-methyl-N-m ethyloxycarbonylamino)methane byproduct. The oils were purif ied by trituration with hexane overnight followed by ether overnight. The clear soluti on was decanted leaving the white residue (bis derivative) behind. Th e clear solution was then concentrated. The peak at 4.8 due to the ‘CH2 ’ of the bis derivative and peak at 5.3 due to the ‘CH2’ of N-methyl-N-methyloxycarbonylaminomethyl chloride were used to quantitate the amount of product formed: yield = 90%, 1H NMR (400 MHz; CDCl3; Me4Si): 5.31-5.33 (2s, 2H), 3.73-3.79 (2s, 3H), 2.9-3.0 (2s, 3H). N-Aryl-N-methyloxycarbonylaminomethyl chloride synthesis (Figure 2-4): N-Aryl-N-methoxycarbonylaminomet hyl chlorides were made from aromatic amines in two steps. Methylchloroformate was reacted wi th aromatic amines in the presence of pyridine and dichloromethane to give N-aryl carbamic acid methyl esters. The chloromethyl derivative of the carbamic acid derivative was then made by a synthetic

PAGE 66

50 procedure reported by Moreira et al., (1994) for N-methylamides. The N-Methyl carbamic acid alkyl ester was refluxed with thirteen equivalents of trimethylsilyl chloride and 1.7 equivants of paraformaldehyde to gi ve the appropriate Naryl-N-alkoxycarbonylaminomethyl chloride. (a) N-Aryl carbamic acid methyl ester: To a solution of methylchlo roformate (3.5 mmol) in 15 mL CH2Cl2 was added dropwise an equivalent of pyridine and aromatic amine. The reaction mixture was allowed to warm to r oom temperature and subsequently stirred overnight. The clear solution was washed with 10 mL brine 3 times, and the organic layer was dried over Na2SO4 and concentrated to a solid. The solids obtained were recrystallized from CH2Cl2: hexane. N(Phenyl)carbamic acid methyl este r: yield = 98 %, mp = 44-45C, 1H-NMR (400 MHz; CDCl3; Me4Si): 7.2-7.37(m, 4H), 7.06 (m, 1H), 6.63(s, 1H), 3.76 (s, 3H). N-(4Â’-Ethoxycarbonylphenyl )carbamic acid methyl ester: yield = 97 %, mp = 152-154C, 1H-NMR (400 MHz; CDCl3; Me4Si): 8.0 (d, 2H), 7.45 (d, 2H), 6.9(s, 1H), 4.37 (q, 2H), 3.8 (s, 3H), 1.39 (t, 3H). N-(4Â’-Methoxyphenyl)carbamic aci d methyl ester: yield = 94 %, mp = 85-86C, 1HNMR (400 MHz; CDCl3; Me4Si): 7.27 (d, 2H), 6.85 (d, 2H), 3.78(s, 3H), 3.75(s, 3H). (b) N-Aryl-N-methyloxycarbonylaminomethyl chloride: A suspension of N-aryl carbamic acid alkyl ester (2.5 mmol), 1.7 e quivalents of paraformaldehyde and 13 equivalent of trimethylsilyl ch loride was refluxed using a CaCl2 drying tube and a water condenser for 18 h over an oil bath. The suspension was diluted with ~10 mL CH2Cl2 and filtered to get rid of the unreacted paraformaldehyde. The clear filterate was concentrated

PAGE 67

51 using a rotavapor at 40C under reduced pressu re. The yellow oil obtained was triturated with hexane overnight. The white suspension obtained was filtered and the filtrate was concentrated to give the desired alkylating agent. N(Phenyl)-N-methyloxycarbonylaminom ethyl chloride: yield = 75%, 1H NMR (400 MHz; CDCl3; Me4Si): 7.32-7.44 (m, 5H), 5.56(s, 1H), 3.78 (s, 3H). N(4’Ethyloxycarbonylphenyl)-N-methyloxy carbonylaminomethyl chloride: Complete conversion to product wasn’t seen in this ca se, the crude mixture was thus used in the next step. The peak at 6.9 due to the ‘NH’ of the N-(4’-ethy loxycarbonylphenyl) carbamic acid methyl ester and the peak at 5.56 due to the ‘CH2’ of N-(4’ethyloxycarbonylphenyl)-N-methyloxycarbonylaminom ethyl chloride were used to quantify the amount of product formed: yield = 27%, 1H NMR (400 MHz; CDCl3; Me4Si): 8.1 (d, 2H), 7.5 (d, 2H), 5.56(s, 1H), 4.38 (q, 2H), 3.81 (s, 3H), 1.39 (t, 3H). N-(4’-methoxyphenyl)-N-methyloxycarbonylaminomethyl chloride: yield = 82%, 1H NMR (400 MHz; CDCl3; Me4Si): 7.23 (d, 2H), 6.93 (d, 2H), 5.52(s, 1H), 3.82 (s, 3H), 3.76 (s, 3H). Alkylation of Phenols, 6-Mercaptopurin e, Dimethylaminobenzoic acid and Naproxen with N-methyl-N-methyloxycarbony laminomethyl chloride or N-arylN-methyloxycarbonylaminomethyl ch loride (Figures 2-3 and 2-4). Typical procedure for phenols: In a round bo ttom flask was dissolved 1 equivalent of phenol (~0.01 mol) and 1.1 equiva lents triethylamine in 20 mL CH2Cl2. The contents were stirred for an hour under reflux conditi ons with a water condenser and oil bath. NMethyl-N-alkoxycarbonyl aminomethyl chlo ride or N-aryl-N-methyloxycarbonylaminomethyl chloride (1.1 equivalents) in 5 mL CH2Cl2 was then added dropwise. An exothermic reaction occurred and white fumes could be seen. The clear solution or

PAGE 68

52 suspension (depending on the phenol used) was stirred overnight. The reaction was worked up by filtering the suspension [NEt3 +Cl-:, 1HNMR(CDCl3): 3.1 (q,2H), 1.5 (t,3H)], diluting the filtrate to 40 mL with CH2Cl2 and washing it with 3 30 mL water. The CH2Cl2 solution was dried over Na2SO4 for an hour and filtered. The solution was concentrated using a rotavapor under vacuum at 40C until solvent free. The resulting material was purified by recrys tallization and, if necessary, column chromatography until a sharp melting point was obtained, a si ngle spot was seen on TLC and a clean 1HNMR was obtained. The particular results for each synthesis are listed below. 1 was prepared from N-methyl-N-methyloxycarbonylaminomethyl chloride, triethylamine and acetaminophen in CH2Cl2. Recrystallization from CH2Cl2: hexane (1:3) twice gave white crystals: yi eld = 70%, mp = 86-88C, Rf (0.36, ether). Elemental analysis (Found: C, 56.81; H, 6.34; N, 11.1. Calc. for C12H16N2O4 : C, 57.13; H, 6.39; N, 11.1%). UV: max (pH 8.8 buffer)/nm 243.4 ( /Lmol-1cm-10.95 10 4). UV: max (pH 7.1 buffer)/nm 240 ( /Lmol-1cm-1 1.01 10 4 0.09 10 4).1H NMR(400 MHz; CDCl3; Me4Si): 7.6(s,1H), 7.39(d,2H), 6.96-6.87(2d,2H), 5.28-5.21(2s,2H), 3.723.7(2s,3H), 3.0-2.97(2s,3H), 2.0(s,3H). 2 was prepared from N-methyl-N-m ethyloxycarbonylaminomethyl chloride, triethylamineand p-cyanophenol in CH2Cl2 Recrystallization from ethyl acetate:hexane (1:2) twice gave white crystals : yield = 55%, mp = 56-58C, Rf (0.17, ethyl acetate:hexane, 1:4). Elemental analysis (Found: C, 59.54; H, 7.27; N, 9.94. Calc. for C11H12N2O3 : C, 60.02; H, 7.14; N, 10%). UV: max (pH 8.8 buffer)/nm 245 ( /Lmol1cm-1 1.26 10 4).1H NMR (400 MHz; CDCl3; Me4Si): 7.6(d, 2H), 7.1-6.96(2d, 2H), 5.38-5.31(2s, 2H), 3.8(s, 3H), 3.06-3.0(2s, 3H).

PAGE 69

53 3 was prepared from N-methyl-N-m ethyloxycarbonylaminomethyl chloride, triethylamine and p-hydroxybenzaldehyde in CH2Cl2. Purification on a silica gel column with hexane: acetone (4:1) as the eluent gave a colorless oil: yield = 67%, Rf (0.37, ethyl acetate:hexane, 1:2). Elemental analysis (Found: C, 59.08; H, 5.86; N, 6.3 Calc. for C11H13NO4 : C, 59.19; H, 5.87; N, 6.27%).UV: max (pH 8.8 buffer)/nm 276.6 ( /L mol-1cm-1 1.37 10 4).1H NMR(400 MHz; CDCl3; Me4Si): 9.9(s,1H), 7.83(d,2H), 7.1-7.03(2d,2H),5.42-5.35(2s,2H), 3.8(s,3H), 3.07-3.01(2s,3H). 4 was prepared from N-methyl-N-m ethyloxycarbonylaminomethyl chloride, triethylamine and salicyldehyde in CH2Cl2. Recrystallization fr om acetone gave white crystals: yield = 87%, mp = 65-67C, Rf (0.76, ether).Elemental analysis (Found: C, 59.06; H, 6.01; N, 6.24 Calc. for C11H13NO4 : C, 59.19; H, 5.87; N, 6.27%). UV: max (pH 8.8 buffer)/nm 315 ( /Lmol-1cm-11.27 10 4). 1H NMR(400 MHz; CDCl3; Me4Si): 10.49(s,1H), 7.85(d,1H), 7.55(t,1H) 7.2(m,2H), 5.47-5.38(2s,2H), 3.76(s,3H), 3.08-3.03(2s,3H). 5 was prepared from N-methyl-N-m ethyloxycarbonylaminomethyl chloride, triethylamine and p-hydroxyacetophenone in CH2Cl2. Purification on a silica gel column with hexane: acetone (4:1) as the eluent gave a colorless oil: yield = 82%, Rf (0.75, ether). Elemental analysis (Found: C, 60.45; H, 5.91; N, 6.22 Calc. for C11H13NO4 : C, 60.75; H, 6.37; N, 5.9%). UV: (pH 8.8 buffer)/nm 269.7( /Lmol-1cm-1 1.12 10 4).1H NMR(400 MHz; CDCl3; Me4Si): 7.93(d,2H), 7.05-6.94(2d,2H), 5.39-5.32(2s,2H), 3.8(s,3H), 3.05-3.0(2s,3H), 2.6(s,3H). 6 was prepared from N-methyl-N-methyloxycarbonylaminomethyl chloride, triethylamine and methyl p-hydroxybenzoate in CH2Cl2. Purification on a silica gel

PAGE 70

54 column with ethyl acetate: hexane as elue nt gave a colorless oil. This oil, after recrystalization from CH2Cl2 :hexane (1:2), gave white crystals: yield = 21%, mp = 5254C, Rf (0.49, ethyl acetate:hexane,1:4).Elementa l analysis (Found: C, 56.98; H, 5.93; N, 5.49.Calc. for C12H15NO5 : C, 56.91; H, 5.97; N, 5.53%). UV: max (pH 8.8 buffer)/nm 253 ( /Lmol-1 cm-1 1.49 10 4). 1H NMR(400 MHz; CDCl3; Me4Si): 8.0(d,2H), 7.02-6.92(2d,2H), 5.37-5.3(2s,2H), 3.9(s,3H), 3.8(s,3H), 3.052.99(2s,3H). 7 was prepared from N-methyl-N-methyloxycarbonylaminomethyl chloride, triethylamine and p-nitrophenol in CH2Cl2. Recrystallization from CH2Cl2 :hexane (1:4) gave pale yellow crystals: yield = 76%, mp = 77-79C, Rf (0.37,ether: hexane, 1:1).Elemental analysis(Found: C, 49. 99; H, 4.94; N, 11.57.Calc. for C10H12N2O5 : C, 50.0; H, 5.04; N, 11.66%). UV: max (pH 8.8 buffer)/nm 310.6 ( /Lmol-1cm-10.76 104). UV: max (pH 7.1 buffer)/nm 310 ( /Lmol-1cm-1 0.75 10 4). 1H NMR (400 MHz; CDCl3; Me4Si): 8.21(d, 2H), 7.1-6.99(2d, 2H), 5.42-5.35(2s, 2H), 3.77(s, 3H), 3.07-3.01(2s, 3H). 8 was prepared from N-(4Â’-methoxyphenyl )-N-methyloxycarbonylaminomethyl chloride, triethylamine and p-nitrophenol in CH2Cl2. Purification on a silica gel column with ethyl acetate: hexane as eluent gave a colorless oil: yield = 64%, Rf (0.17, ethyl acetate:hexane, 1:5), UV: max (pH 8.8 buffer)/nm 310 nm ( /Lmol-1cm-1 1.06 10 4 ). 1H NMR(400 MHz; CDCl3; Me4Si): 8.2(d,2H), 7.16(d,2H), 7.05(d,2H), 6.89(d,2H), 5.64(s,2H), 3.81(s,3H), 3.73(s,3H).

PAGE 71

55 9 was prepared from N-(4Â’-ethyloxycarbonylphe nyl)-N-methyloxycarbonylaminomethyl chloride, triethylamine and p-nitrophenol in CH2Cl2. A white solid residue was seen. This residue was dissolved in CH2Cl2 and hexane was added till th e solution became turbid and white crystals were seen. The white crystals were filtered and discarded while the filterate was reconcentrated to give more white crystals which were filtered. The filterate was purified on a silica gel column with ethyl acet ate: hexane as eluent to give a colorless oil: yield = 17%, Rf (0.29, ethyl acetat e:hexane, 1:3). UV: max (pH 8.8 buffer)/nm 309 ( /Lmol-1cm-1 1.03 10 4). 1H NMR(400 MHz; CDCl3; Me4Si): 8.21(d,2H), 8.07(d,2H), 7.37(d,2H), 7.05(d,2H), 5.69(s,2H), 4.38(q,2H), 3.77(s,3H), 1.39(t,3H). 10 was prepared from N-(phenyl)-N-methyloxycarbonylaminomethyl chloride, triethylamine and p-nitrophenol in CH2Cl2 Purification on a silica gel column with ethyl acetate: hexane as eluent gave a colorless oil which, on recrystallization from ether: petroleum ether (1:2), gave white crys tals: yield = 69%, mp = 105-106C, Rf (0.26, ethyl acetate: hexane, 1:9). Elemental analysis (Found: C, 59.49; H, 4.63; N, 9.21. Calc. for C15H14N2O5 : C, 59.6; H, 4.67; N, 9.27%). UV: max (pH 8.8 buffer)/nm 305 nm ( /Lmol-1cm-1 1.19 10 4).1H NMR (400 MHz; CDCl3; Me4Si): 8.2(d, 2H), 7.277.42(m, 5H), 7.05(d, 2H), 5.67(s, 2H), 3.75(s, 3H). 11 was prepared from naproxen (0.01 mol) stirre d with triethylamine (0.01mol) in 50 mL CH2Cl2 for an hour. A clear colorless solu tion was seen. N-Methyl-N-methyloxycarbonylaminomethyl chloride (0.01mol) was added to the reaction mixture and the contents were stirred over night. The clear solution was washed 3-times with 10 mL water. The CH2Cl2 solution was dried over Na2SO4 for an hour and then evaporated

PAGE 72

56 under reduced pressure to give white cr ystals: yield = 97%, mp = 100-101C, Rf (0.49, ethyl acetate:hexane,1:4). Elemental analysis (Found: C, 65.05; H, 6.45; N, 4.2 Calc. for C18H21NO5 : C, 65.24; H, 6.39; N, 4.23%). UV: max (pH 7.1 buffer)/nm 271.5 and 245 ( /Lmol-1cm-1 0.52 10 4, 0.77 10 4 L/mole.1H NMR(400 MHz;CDCl3;Me4Si): 7.69(t,3H), 7.4(d,1H), 7.11-7.15(m,2H), 5.41-5.29(dd,2H), 3.91(s,3H), 3.85(q,1H), 3.7-3.65(2s,3H), 2.89-2.85(2s,3H), 1.57(d,3H). 12 was prepared from naproxen (0.001 mol) st irred with triethylamine (0.001mol) in 50 mL CH2Cl2 for an hour. A clear colorless soluti on was seen. N-(Phenyl)-N-methyloxycarbonylaminomethyl chloride (0.001mol) wa s added to the reaction mixture and the contents were stirred over night. The clear solution was washed 3-times with 10 mL water. The CH2Cl2 solution was dried over Na2SO4 for an hour and then evaporated under reduced pressure to give a yellow oil. Purification of the oil on a si lica gel column with ethyl acetate: hexane as eluent gave a colorless oil: yield = 60%, Rf (0.17, ethyl acetate:hexane, 1:5), UV: max (pH 7.1 buffer)/nm 272.8 and 245, ( /Lmol-1cm-10.35 10 4, 0.83 10 4). 1H NMR(400 MHz;CDCl3;Me4Si): 7.69(t,3H), 7.36-7.39(dd,1H), 7.13-7.17(m,5H), 6.9(d,2H), 5.5-5.68(dd,2H), 3.93(s,3H), 3.88(q,1H), 3.65(s,3H), 1.57(d,3H). 13 was prepared from naproxen (0.001 mol) st irred with triethylamine (0.001mol) in 50 mL CH2Cl2 for an hour. A clear colorless soluti on was seen. N-(4Â’-Methoxyphenyl)-Nmethyloxycarbonylaminomethyl chloride (0.001 mol) was added to the reaction mixture and the contents stirred overnight. The clea r solution was washed 3-times with 10 ml water. The CH2Cl2 solution was dried over Na2SO4 for an hour and then evaporated under reduced pressure to give to give a yell ow oil. Purification of the oil on a silica gel

PAGE 73

57 column with ethyl acetate: hexane as el uent gave a colorless oil: yield = 70%, Rf (0.1, ethyl acetate:hexane,1:5), UV: max (pH 7.1 buffer)/nm 271.5 and 245, ( /Lmol-1cm10.26 10 4, 0.91 10 4 ). 1H NMR(400 MHz;CDCl3;Me4Si):7.69(t,3H), 7.367.39(dd,1H), 7.13-7.17(m,2H), 7.69(t,3H), 7.37(dd,1H), 6.8(d,2H), 6.5(d,2H), 5.5-5.64(dd,2H), 3.92(s,3H), 3.88(q,1H), 3.68(s,3H), 3.65(s,3H), 1.57(d,3H). 14 was prepared from naproxen (0.001 mol) st irred with triethylamine (0.001mol) in 50 mL CH2Cl2 for an hour. A clear colorless soluti on was seen. N-(4Â’-Ethyloxycarbonylphenyl)-N-methyloxycarbonylaminomethyl chloride (0.001mol) was added to the reaction mixture and the conten ts were stirred overnight. Th e clear solution was washed 3-times with 10 mL water. The CH2Cl2 solution was dried over Na2SO4 for an hour and then evaporated under reduced pressure to give white crystals. This residue was dissolved in CH2Cl2 and hexane was added till the solution became turbid and white crystals were seen. The white crystals were filtered and discarded while the filterate was reconcentrated to give more white crystals. After purification on a silica gel column with ethyl acetate: hexane as eluent the filtrate gave a colorless oil: yield = 13%, Rf (0.25, ethyl acetate: hexane, 1:5), UV: max (pH 7.1 buffer)/nm 270 and 245 ( /Lmol-1cm-10.25 10 4, 0.95 10 4. UV: max (pH 4 buffer)/nm 245( /Lmol-1cm-10.88 10 4). UV: max (pH 6 buffer)/nm 245 ( /Lmol-1cm-1 0.9 10 4). UV: max (pH 8.25 buffer)/nm 245 ( /L mol-1cm-1 0.92 10 4). UV: max (pH 9.2 buffer)/nm 245 ( /Lmol-1cm-10.94 10 4). 1H NMR(400 MHz;CDCl3;Me4Si): 7.76(d,2H), 7.63-7.71(m,3H),7.34-7.37(dd,1H), 7.13-7.17(m,2H), 6.98(d,2H), 5.5-5.7(dd,2H), 4.3(q,2H), 3.94(s,3H), 3.88(q,1H), 3.66(s,3H), 1.57(d,3H), 1.38(t,3H). 15 was prepared from dimethyl aminobenzoic acid (0.01 mol) stirred with triethylamine

PAGE 74

58 (0.01mol) in 50 mL CH2Cl2 for an hour. To the white suspension that formed was added N-methyl-N-methyloxycarbonylaminomethyl chlo ride (0.01 mol) and the contents were stirred overnight. The clear solution was washed 3-times with 10 mL water. The CH2Cl2 solution was dried over Na2SO4 for an hour and then evaporated under reduced pressure to give white crystals: yield = 97%, mp = 100-101C, Rf (0.49, ethyl acetate: hexane, 1:4), Elemental analysis: (Found: C, 57.73; H, 7.02; N, 10.6. Calc. for C13H18N2O4 : C, 57.64; H, 6.81; N, 10.52 %). UV: max (pH 8.8 buffer)/nm 315 and 287 ( /Lmol-1 cm-11.41 10 4, 1.15 10 4). 1H NMR(400 MHz;CDCl3;Me4Si): 7.9(d,2H), 6.6(d,2H), 5.58-5.54(2s,2H), 3.76-3.72(2s,3H), 3.05(s,9H). 16 was synthesized using the protoc ol developed by Siver et al.7 To 0.01 mol 6MP in 15 mL DMSO was added 1.1 equivalents of N-methyl-N-methyloxycarbonylaminomethyl chloride. The yellow solution was stirred at r oom temperature for 1.5 h. Triethyl amine, 2.5 equivalents and 45 mL CHCl3 was added and stirring continued for one more hour. The reaction mixture was th en diluted with 100 mL CH2Cl2 and washed with 10 ml 1N HCl, 20 ml saturated NaHCO3 solution and 350 mL brine solution. The organic layer was dried over Na2SO4 and concentrated to give a yellow oil. The oil was triturated with hexane to give yellow solids which we re further recrystallised with CH2Cl2: hexane (1:5) to give yellow crystals: yield = 62%, mp = 162-164C (lit mp = 163C), Rf = 0.46(10:3, ether: methanol), Elemental analysis (Found: C, 42.9; H, 4.08; N, 27.78. Calc. for C9H11N5O2S : C, 42.68; H, 4.38; N, 27.65 %). UV: max (pH 7.1 buffer)/nm 288.4 ( /Lmol-1cm-1 0.71 10 4). UV: max (pH 8.8 buffer)/nm 310.2 ( /Lmol-1cm-10.55 10 4) .UV: (MeOH)/nm 286 ( / Lmol-1cm-10.41 10 4). 1H NMR (400 MHz;CDCl3;Me4Si): 8.71(s,1H), 8.1-8.37(2s,1H), 5.68(s,2H), 3.75-3.85 (2s, 3H), 3.13-3.09(2s,3H).

PAGE 75

59 Hydrolysis Studies. S6-(N-Methyl-N-methyloxycarbonyl) aminomet hyl6MP was synthesized by Siver et al. This derivative reverted to 6MP with a half life of 91 minutes in pH 7.1 buffer at 32 C. An SN1 mechanism of hydrolysis was propos ed but not firmly established. We decided to probe the reactivity of NANAOCAM -phenol conjugates with acetaminophen (APAP) acting as our model phenolic drug. Since NANAOCAM-APAP type derivatives we re designed as a hydrolytically labile prodrugs, their rates of hydrolysis in aqueous buffers were investigated. Hydrolysis of NANAOCAM derivatives of APAP at 32 pH and 7.1 buffer were much slower than those of 6MP. The rate of hydrolysis of N-Methyl-N-methoxycarbonylaminomethylAPAP (1) at pH 8.8 and 46 C was found to be 60 h. so all rates of hydrolysis were determined at 46 C and pH 8.8 buffer S6-(N-Methyl-N-methyloxycarbonyl) aminomethyl6MP (16), resynthesized according to Siver et al. (1990), had a half life of 19 minutes under the same conditions. The anion of APAP (pKa~9.5) is not a good leaving group and as a result the hydrolysis of its prodrug is slow compared to 6MP (pKa~7.5). On the other hand, the NANAOCAM derivative of p-nitrophenol (7, Tables 21 and 2-3) had a half life of 22 min at 46C and pH 8.8 buffer. The hydrolysis rate was higher and t 1/2 lower since the pKa of the leaving group was much lower:7.4. Thus NMethyl-N-alkyloxycarbonylamin omethyl derivatives of phenols with an electron withdrawing group like –NO2 on the ring hydrolyzed faster than those with an electron releasing group on the ring like –NHCOCH3. In order to establish the mechanism of hydrolysis and probe the effect of electron withdrawing groups on hydrolysis rates of NANAOCAM-phenol conjugates, five more phenols (2-6, Table 2-1) with an electron withdraw ing group on the aromatic ring were

PAGE 76

60 synthesized using the syntheti c protocol developed for APAP and rates of hydrolysis were determined at pH 8.8 and 46 C (Table 2-3). To further establ ish the effect of pKa on rates of hydrolysis, the N-Methyl-N -methyloxycarbonylaminomethyl ester of dimethylaminobenzoic acid (15, Figures 2-1 and 2-3) was also synthesized and its rate of hydrolysis determined.Thus, rates of hydrol ysis were determined where YH = 6MP, phenol or carboxylic acid and the correspondi ng anions acted as the leaving group (Figure 2-5). N R Y O OR' N R O H O OR' N R H O OR' N R CH2 O OR' Y OH2 -CH2=O + YH + Figure 2-5 Hydrolysis of NANAOCAM-Y in aqueous buffers. Rates of hydrolysis of NANAOCAM-Y we re dependent on the acidity (leaving group ability) of the parent compound. The mo re acidic YH was, the faster the rate was. When YH was cyanophenol with a pKa of 7.95, the half life was 149 minutes. Increasing the pKa by 0.1 unit as in the case of the methyl paraben derivative, 5 raised the half life to 189 minutes. This sensitivity to small changes in pKa of YH derivative suggests that Ywas acting as the nucleofuge by either a SN2 or SN1 mechanism. The effect of electon withdrawing groups in the para position of the phenol on rates hydrolysis was also

PAGE 77

61 quantified using the Ha mmett plots between substituents versus log K. (Table 2-3). Plot of log K and sigma ( -) were linear and indicated th at rates of hydrolysis were dependent on the electron wit hdrawing effect imparted by th e substituent at the para position (Figure 2-6). Table 2-3 Correlation of Rates of Hydrol ysis of NANAOCAM-Y with pKa and Sigma Values of the Leaving Group (Y-). Compd Y log k (sec-1) t 1/2(min.) pKa 15 p-Me2N-C6H4COO-1.30 0.22 5.03 4 o-OHC-C6H4O-3.03 12 6.79 16 6MP -3.22 19 7.5 7 p-O2N-C6H4O-3.28 22 7.14 1.25 3 p-OHC-C6H4O-3.81 75 7.66 0.94 2 p-NC-C6H4O-4.11 149 7.95 0.99 5 p-MeOC-C6H4O-4.21 189 8.05 0.82 6 p-MeOOC-C6H4O-4.39 290 8.47 0.74 1 p-MeOCHN-C6H4O-5.49 3600 9.5 0.19 When log k (rate constant) was plotted against the pKa of the leaving group (Y-), a negative correlation was found with a r2 of 0.98 (Figure 2-7). Ra te of hydrolysis were also independent of the pH of the buffer. Wh en rates were studied in various buffers (pH 4.6, 7.1, 8.8) at 46 C for 7 and log K was plotted against buffer pH a stra ight line with zero slope was obtained (Table 2-4, Fi gure 2-8). The pH rate profile of 14 will be discussed later. Table 2-4 Effect of pH of Bu ffer on Rates of Hydrolysis of 7 pHt1/2 (min) log k 4.622.83 -3.29588 7.121.83 -3.27655 8.822.83 -3.29596

PAGE 78

62 y = 2.0288x 5.889 R2 = 0.968 -6 -5 -4 -3 -2 -1 0 00.511.5 SigmaLog k -NO2 -CHO -CN -COCH3 -COOMe -NHCOCH3 Figure 2-6 Pseudo-first order ra te constants (sec-1) versus of parent phenol. y = -0.9325x + 3.4017 R2 = 0.9828-6 -5 -4 -3 -2 -1 0 46810pKaLog k 15 4 16 7 3 5 2 6 1 Figure 2-7 Pseudo-first order rate constants (sec-1) vers us acidic pKa of parent.

PAGE 79

63 -5 -4.8 -4.6 -4.4 -4.2 -4 -3.8 -3.6 -3.4 -3.2 -3 3.55.57.59.5pHLog k 7 14 Figure 2-8 pH rate profiles of compds 7 and 14 Several mechanisms for base catalyse d hydrolysis of NANAOCAM-Y can be proposed (see above).Figure 2-9A shows an SN2 type of pathway where hydroxide ion acts as a nucleophile and displaces Yto give a hydroxymethyl-N-alkylcarbamic acid alkyl ester. This hemiaminal derivative s pontaneously hydrolyzes in water to give formaldehyde and N-alkylcarba mic acid alkyl ester in the same way that hydroxymethyl amides do (Johansen et al., 1979 and Bundga rd et al., 1980). Similarly the hydroxide anion can act as a nucleophile and displaces the N-alkylcarbamic acid alkyl ester anion as in Figure 2-9B. The hydroxymethyl conjugate of the drug, YH, formed as an intermediate reverts to the parent drug by loss of form aldehyde. This particular mechanism however can be ruled out because of the lower pKa of YH (~5.2-9.5) compared to the Nalkylcarbamic acid alkyl ester (~14) making Ya better leaving group than the Nalkylcarbamic acid alkyl ester anion.

PAGE 80

64 O N O R Y O N O R OH O N O R H O N O R Y O N O R Y OH CH2=O YH Y -CH2=O OH2 OH2O N O R H + OH OH YH + OH + OH + Figure 2-9 SN2 type of hydrolysis of NANAOCAM-Y with Yor carbamate as the leaving group. O N O R Y O N O R O N O R OH2 O N O R H OH2 -CH2=O Y OH2 + YH + OH+ + Figure 2-10 SN1 type of hydrolysis of NANAOCAM-Y.

PAGE 81

65 A SN1 type of pathway is another possibili ty (Figure 2-10). The lone pair of electrons on nitrogen can donate its electrons to stabilize an incipient carb ocation with Yleaving. This carbocation can subsequently react with water to give hydroxymethyl-Nalkylcarbamic acid alkyl ester which then fa lls apart to give N-alkylcarbamic acid alkyl ester. However none of the data plotted in Figures 2-6 and 2-7 provi des an insight into the mechanism of hydrolysis since both SN1 and SN2 mediated hydrolys is are dependent on the leaving group ability of th e ionized parent molecule, Y-. However the fact that the hydrolyses were pH indepe ndent strongly favors a SN1 mechanism. There is also some precedent in literature suggesting the mechanism of hydrolysis to be SN1. NAlkylamidomethyl esters of carboxylic acid hydrolyze by a SN1 mechanism (Moreira et al., 1994) (Figure 2-11). O N O R O R' O N R R'COO O N OH R Ph Ph + + Ph H O 2 Figure 2-11 Hydrolysis of Nalkylamidomethylcarboxylic acid esters in aqueous buffers. Hydrolysis proceeds by donation of electrons from the amide nitrogen resulting in a carbocation intermediate with the carboxylate group leavi ng. In the subsequent step, the carbocation reacts with wate r to give the hydroxymethylamide. When R’ was naproxen and R was methyl the t 1/2 for hydrolysis of the prodrug wa s 14 min at pH 7.4 and 32 C, while when R was –CH2COOEt t 1/2 for hydrolysis was 2520 min. Thus, introducing an electron withdrawing group on the amide nitr ogen diminishes its ability to donate electrons and slows down reac tion rate, thereby suggesting the mechanism of hydrolysis to be SN1.

PAGE 82

66 The hydrolysis of O-Imidomethyl derivatives of phenols has been reported by Getz et al., (1992). Figure 2-12 above compares the halflives of imidomethyl derivatives of pnitrophenol (PNP) with (N-Methyl-N-methoxycarbonyl) aminomethyl-PNP. The proposed hydrolytic mechanism for the im idomethyl derivatives where saccharin, phthalimide and succinimide served as the imide, was SN2. Thus, presence of two electron withdrawing groups on th e nitrogen disfavors SN1, (as observed in case of N-alkylamido methyl esters of carboxylic ac ids) to the extent that SN2 becomes the major pathway. t 1/2(min) 0.75(25C) 4.9(25C) 7.43(25C) 22(46C) NO2O N S O O O NO2O N O O NO2O N O NO2O OH NO2 NO2O N O O Figure 2-12 Hydrolysis of Nimidomethyl derivatives of phenols in aqueous buffers. Values are shown for t1/2 at th e temperature (in parenthesis).

PAGE 83

67 Implications of NANAOCAM-phenol Hydrolysis on Prodrug Design. NANAOCAM-phenol conjugates are chemica lly stable and the suggestion that their mechanism of hydrolysis is SN1 is based on past literature We were interested in designing derivatives which would be labile under physiological conditions and which would illustrate that the mechanism of hydrolysis was SN1. Hence the syntheses of NANAOCaminoethylidene (ROCO-NRÂ’CH(CH3)), NANAOCaminobenzylidene (ROCO-NRÂ’CH(Ph)) and NArNAOCAM (ROCO-N(Ph)CH2) promoieties were examined. NANAOCaminoethylidene and NANAOCaminobenzylidene derivatives of phenols will stabilize positive charge on the methine carbon leading to a more labile derivative if hydrolysis follows a SN1 type of pathway and will hinder an SN2 type path by creating more steric hinderance. Also if the mechanism is SN2, the rate of hydrolysis should be accelerated by an electron withdrawing group on an N-aryl group where a positive charge on CH2 in N-CH2-O is increased making the CH2 a better target for nucleophilic attack. On the other hand, if the mechanism is SN1, the rate of hydrolysis should be decelerated by an electron withdrawing group on an N-aryl group where a positive charge on CH2 in N-CH2-O is destabilized. Unfortunately our attempts to synthesize NANAOCaminoethylidene and NANAO C-aminobenzylidene derivatives of phenols were unsuccessful. The synthetic prot ocol used is shown in Figure 2-13. An aldehyde was reacted with aqueous methyl amine at room temperature to give Nmethylimines which were reacted with methyl chloroformate under nitrogen at -78C to 0C. The major product obtained was the pa rent aldehyde, N-methyl carbamic acid methyl ester and 2-3% of the desired alkylating agent (NANAOCaminoethylidene chloride or NANAOCaminobenz ylidene chloride) based on 1H-NMR.

PAGE 84

68 Y O N O Y O N X O O O X = H, OCH3, COOEt Y O N O X O X = H, OCH3, COOEt NANAOaminoethylidene-Phenol NANAOaminobenzylidene-Phenol NArNAOCAM-Phenol Attempts to isolate the alkylating agent by column chromatography led to complete decomposition of the alkylating agent. Re action of phenols with the crude reaction mixture containing the alkylating agent also lead to hydrolysis of the alkylating agent to the parent aldehyde and N-methyl carbam ic acid methyl ester quantitatively. RCHO + CH3NH2 RCH=NHCH3 Cl N COOCH3 CH3OCOCl RCHO CH3NHCOOCH3 RCH + + major minor R = Methyl or Aryl NANAOC amino ethylidene chloride or NANAOC amino benzylidene chloride Figure 2-13 Attempts at the synthesis of NANAOCaminoethylidene chloride and NANAOCaminobenzylidene chloride. However our attempts to make NAr NAOCAM conjugates of phenols were successful. NArNAOCAM chlorides (RÂ’OCONRCH2-, R = aryl, RÂ’= CH3) were synthesized and used to alkylate pnitrophenol (Figure 2-3, Table 2-1, 8-10). The halflives of hydrolyses clearly illustrate that when an N-alkyl group (Table 2-3, t 1/2 for 7 = 22 min) was replaced by N-aryl (Table 2-5, t 1/2 for 8 and 10 = 400-520 min while t 1/2 for

PAGE 85

69 9 > 24 h) the rates of hydrolyses were slow er. An electron donati ng substituent on the Naryl ring like –OCH3 stabilizes the SN1 transition state more than an electron withdrawing group like –COOC2H5 or -H and the half lives are in the order of 9 >10 >8. Thus, hydrolysis of NArNAOCAM and N-aryl-N-alk yloxycarbonylaminomethyl derivative of phenols follow a SN1 type of mechanism. Howe ver, although NANAOCAM-phenols hydrolyse by a SN1 type of pathway, they are chemically too stable to revert to the parent drug at a sufficient rate to be effec tive prodrugs based on chemical hydrolysis. Hydrolyses of NANAOCAM-phenols does occur enzymatically. When 1 was used in diffusion cell experiments where hairless m ouse skin was the membrane, about 23% of acetaminophen was regenerated from the prodr ug during its passage through the skin. We will talk more about diffusion cell experiments in Chapter 3. Thus NANAOCAM-phenols represent a novel class of prodrugs which are sufficiently chemically stable to allow form ulation but may be sufficiently enzymatically labile to revert to the pa rent drug at a useful rate. Shorter chain NANAOCAM prodrugs of phenols have both higher lipid and water solubilities compared to ACOM and AOCOM phenolic conjugates (solubililities sh own in Chapter 3). Thus, replacing the oxygen atom in O-CH2 of ROCOOCH2 with a substituted nitrogen (N-R, R = alkyl) makes it possible to increase solubility in a membrane which should increase permeability across a biological barrier such as skin and oral absorption across the GI tract.

PAGE 86

70 O O2N N O O X Table 2-5 N-aryl-N-alkyloxycarbonylaminomet hyl Derivatives of p-nitrophenol (PNP) a. Compd X log kobsv(sec-1) t (min.) 8 OMe -4.54 400 9 COOEt >24 h 10 H -4.65 519 a Hydrolysis experiments in pH 8.8 and 46C Implications of NArNAOCAM-carboxylic acid Hydrolysis on Prodrug Design Besides NANAOCAM derivatives of phe nols, NANAOCAM derivatives of carboxylic acids have also been examined. Ca rboxylic acids, like phenols, are polar and being ionized at physiological pH penetrate membranes with difficulty. By transiently masking the negative charge it has been s hown that it is possible to increase the permeation of carboxylic acids across biological barriers (Boni na et al., 1995 and Rautio et al., 1999). N-Alkyl-N-al kyloxycarbonylaminomethyl derivatives of carboxylic acids are chemically too unstable to serve as usef ul derivatives (compared to phenols which were chemically stable) because the leaving group has a pKa of ~3-5 (compared to 6.89.5), e.g. 15, t 1/2 ~ 6 min at pH 7.1 and 39C. Th us, the goal here was to design derivatives which were chemically more stable so that they could be easily formulated but would hydrolyze independent of enzymes. It has been previously shown that Oimidomethyl derivatives of phenol s like estradiol were enzymatically stable but had a half life of 100 min in pH 7.4 buffe r (Patel et al., 1994 and 1995). This derivative was about 8 times more potent than estradiol alone when given orally. Thus prodrugs whose

PAGE 87

71 hydrolysis is not dependent on enzymes are us eful because their hydrolysis isnÂ’t limited by enzymatic variability and st rict substrate specificity of some esterases in the blood which could lead to incomplete hydrolysis to the parent drug (B eamount et al., 2003). Here, naproxen was chosen as a model ca rboxylic acid drug for which to prepare NANAOCAM prodrugs (Table 2-2). The NANAOCAM deriva tives of naproxen were made by alkylating naproxen with NANAOCAM-Cl in presence of triethyl amine in dichloromethane (Figure 2-3). 11, Table 2-2 had a half life of 2 min at pH 7.1 and 39 C which would make pharmaceutical formulation difficult. Si nce the mechanism of hydrolysis of NANAOCAM esters of carboxylic acids should also be SN1, the NArNAOCAM promoiety should lead to a chemically more stable and useful prodrug of a carboxylic acid since the phenol NArNAOCAM derivatives were much more stable. NArNAOCAM ester derivatives of naproxen (12-14, Table 2-2) were synthesi zed by alkylating naproxen with NArNAOCAM-Cl (Figure 2-4) and their rate s of hydrolysis determined (Table 2-6). As expected the substituted Naryl derivatives increased the half-lives of the hydrolyses from 2 min to over 150 min. As in the case of phenols the order of rates of hydrolyses was, -COOEt < -H < -OMe which is consistent with a SN1 type of hydrolysis. All three NArNAOCAM naproxen derivatives represent potentially useful prodrugs which are reasonably stable but should be sufficientl y labile enough to rel ease the drug at an appropriate rate to express its pharmacological activity. Moreover, since 14 would eventually release p-aminobenzoic acid, a co mponent of sunscreens, it may be the most attractive candidate.

PAGE 88

72 CH3O O O N O O X Table 2-6 Hydrolysis of N-aryl-N-al kyloxycarbonylaminomethyl Derivatives of Naproxen.b Compd X log kobsv.(sec-1). t (min.) 12 H -4.015 119.62 13 OMe -3.63 49.83 14 COOEt -4.13 156.46 b Hydrolysis experiments in pH 7.1 and 39C The kinetics of decomposition of N-(4Â’-ethyloxycarbonylphenyl)-N-methyloxycarbonylaminomethyl ester of naproxen, 14, was also studied in aqueous buffers at 39 C over a wide pH range (Figure 2-3). Table 2-7 Effect of pH of Bu ffer on Rates of Hydrolysis of 14. pH t1/2 (min) log k 4 166.76 -4.16 6 156.29 -4.13 7 156.46 -4.13 8.25 152.72 -4.12 9.2 35.45 -3.49 The rates of hydrolysis were independent of the pH of bu ffers from pH 4.0-8.25 so the mechanism of hydrolyses along this pH range is believed to be SN1 (Moreira et al., 1996 and Iley et al., 1997). At pH 9.2, a sharp increase in rates of hydrolysis was observed. This is probably due to a change in the mechanism of hydrolysis from SN1 to addition-elimination where nucleophilic addi tion occurs on the ca rbonyl functional group of the ester followed by elim ination of carboxylate ion to form hydroxymethyl-N-methyl

PAGE 89

73 carbamic acid methyl ester (Bundgard et al., 1991). N-Methyl carbamic acid methyl ester is formed in the subsequent step by the loss of formaldehyde (Figure 2-14). Attempts to make NArNAOCAM-naproxen derivatives with enhanced water solubility are currently under investigation. NArNAOCAM-naproxen co njugates are thus prodrugs of carboxylic acids which unlike ACOM or AOCOM esters are not dependent on enzymatic hydrolysis and have sufficient chemical stability to be formulated. O H N O COOEt O H N O COOEt O N O COOEt O O O R -CH2=O RCOOH OH + Figure 2-14 Hydrolysis of NArNAOCAM-carboxylicacid conjugates at pH 9.2.

PAGE 90

74 CHAPTER 3 SYNTHESIS AND TOPICAL DELIVERY OF N-ALKYL-NALKYLOXYCARBONYLAMINOMETHYL PRODRUGS OF A MODEL PHENOLIC DRUG:ACETAMINOPHEN Topical approaches to masking a phenolic functional group have been limited to morphine (Drustrup et al., 1991) naltrexone (Stinchcomb et al ., 2002; Pillai et al., 2004; Valiveti et al., 2005; Hammell et al., 2005; Vaddi et al., 2005, Valivetti et al., 2005a), buprenorphine (Stinchcomb et al., 19 96), nalbuphine (Sung et al., 1998) and acetaminophen (Wasdo and Sloan, 2004). Only three promoieties, alkylcarbonyl (AC), alkyloxycarbonyl (AOC) and alkylaminocarbonyl ( AAC), have been utilized but no soft alkylated derivatives of phenol have been used for topica l delivery. Both AC and AOC derivatives of phenols rely on esterase catalysed hydrolysis or addition-elimination type mechanism to revert to the parent drug.Th e ACOM and AOCOM soft alkyl approach is presently being investigated in our lab to improve topical deli very of phenols. Insertion of ‘N(R’)CH2’ into the AOC (ROCO-) promoiety gives the NANAOCAM (ROCONR’CH2-) promoiety which should act as a so ft alkyl promoiety of phenols since it did for 6MP (Siver and Sloan, 1990). It is of interest to see if this insertion improves the biphasic properties and fl ux through skin of phenols sinc e addition of heteroatoms into the promoiety improves aque ous solubility of drugs. In order to test this hypothesis acetaminophen (APAP) was chosen as the model phenolic drug. The performance of the prodrug series in diffusion cell experiments will be compared with that predic ted by the Roberts-Sloan (RS) equation. The results will be

PAGE 91

75 added to the Sloan and coworkers database to generate new coefficients for the parameters in the equation and give a more robust RS equation. Experimental Procedure Materials and Methods Isopropyl myristate (IPM) was obtained from Givaudan Corp (Clifton, NJ). Theophylline (ThH) was purchased from Sigma Chemical Co. (St. Louis, MO, USA); all other reagent chemicals were from Aldric h Chemical Co. (Milwaukee, WI, USA). The water was obtained from a Millipore Milli-Q water ultra filteration system. Ultraviolet spectra were recorded on a Shimadzu UV2501 PC spectrophotometer. A radiometer pH meter 26 was used to determine pH of soluti ons.The vertical, Fran z type diffusion cells were from Crown Glass (Somerville, NJ, USA) (surface area 4.9cm2, 20 mL receptor phase volume, 15 mL donor phase volume). Th e diffusion cells were maintained at 32C with a Fischer (Pittsburgh, PA, USA) circ ulating water bath model 25. The female hairless mice (SKH-hr-1) were from Charle s River (Boston, MA, USA). Statistical analyses were carried out by using SAS 9.0. All animal sacrifices and preparation of membranes were carried out by Professor Ke n Sloan using IACAC approved protocols. Synthesis of Prodrug Derivatives Synthesis begins by the al kylation of APAP with N-alkyl-N-alkyloxycarbonyl aminomethyl chloride (NANAOCAM-Cl) in presence of a base like triethylamine and CH2Cl2 as the solvent. In every case it was necessary to synthesize the corresponding alkylating agent, NANAOCAM-Cl.

PAGE 92

76 N O O Cl R CH2O CH3NH2 NN N Cl O OR R= CH3 and C2H5 DCM + NaOH Figure 3-1 Synthesis of NAN AOCAM-Cl from 1, 3, 5trialkylhexahydrotriazine. Synthesis of NANAOCAM-Cl (Figure 3-1) a) 1, 3, 5-Trimethylhexahydrotriazine was s ynthesized from equimolar equivalents of aqueous formaldehyde, methyl amine and NaOH according to the protocol originally developed by Graymore et al., (1932) and m odified by Siver et al ., (1990). Methyl amine (0.4 mol of 40% aqueous) was placed in an ice bath and an equivalent of 37% aqueous formaldehyde was added dropwise over a peri od of 10 min. The solution was allowed to equilibrate to room temperature and stirred for one hour, then an equivalent of NaOH was added and the contents were stirred for 1.5 h more. The solution was extracted with 4 50 mL CH2Cl2 and the collected CH2Cl2 were dried over Na2SO4, filtered and concentrated to a clear colorless CH2Cl2 solution containing the the hexahydrotriazine derivative. Complete evaporation of CH2Cl2 wasn’t carried out as it resulted in loss of the desired product. For quantif ication purposes the ‘CH2’ of CH2Cl2 at 5.3 and N-CH2-N peak of hexahydrotri azine derivative at 3.2 were used, yield = 82% in CH2Cl2, 1H NMR (CDCl3): 2.3 (s, 9H), 3.2 (s, 6H). Next the 1, 3, 5-trimethylhexahydrot riazine (freshly prepared) in CH2Cl2 was added to an ice cold solution of 3 equivalent of alkyl chloroformate in CH2Cl2 over a period of 10 minutes. The white suspension that was obser ved was allowed to equlilibrate to room temperature and stirred overnight. The suspension was filtered and the filtrate concentrated to an oil. The oil containe d the desired product and corresponding bis(N-

PAGE 93

77 Alkyl-N-alkyloxycarbonyl)ami nomethanes with a ‘CH2’ peak at 4.8.The oil was purified by trituration with hexane overni ght followed by ether overnight. The clear solution was decanted leaving the white resi due (bis derivative ) behind. The clear solution was then concentrated. The peak at 4.8 due to the ‘CH2 ’ of the bis derivative and peak at 5.3 due to the ‘CH2’ of N-alkyl-N-alkyloxycarb onylaminomethyl chloride were used to quantitate the amount of product formed. N-Methyl-N-methyloxycarbonylaminom ethyl chloride: yield = 90%, 1H NMR (CDCl3): 2.9 (s, 3H), 3.75 (s, 3H), 5.3 (s, 2H). N-Methyl-N-ethyloxyoxycarbonylaminom ethyl chloride: yield = 89%, 1H NMR (CDCl3): 1.3 (t, 3H), 2.9 (s, 3H), 4.22 (q, 2H), 5.33 (s, 2H). N-Methyl-N-propyloxycarbonylaminomethyl chloride, N-methyl-N-butyloxycarbonylaminomethyl chloride, N-ethyl-N-methyl oxycarbonylaminomethyl chloride and Nmethyl-N-hexyloxycarbonylaminomethyl chloride were synthesized in an alternate way via a N-alkyl carbamic acid alkyl ester (Figure 3-2). R'NH2 RO O N H R' TMSCl THF (CH2O)n RO O N R' Cl Cl3CO OCCl3 O O Cl RO O Cl RO R' = CH3, C2H5R = C3H7, C4H9, C6H13 + TEA 3 ROH + TEA Figure 3-2 Preperation of alkyl chloroformat e in situ from alcohol and synthesis of NANAOCAM-Cl from alkyl amine.

PAGE 94

78 b) N-Alkyl carbamic acid alkyl ester. To a solution of chloroformate (33 mmol) in 75 mL CH2Cl2 mounted in an ice bath, was added an equivalent of triethylamine and a queous alkyl amine. Th e reaction mixture was allowed to warm to room temperature a nd stirred overnight. Th e clear solution was washed with 3 10 mL brine and the organic layer was dried over Na2SO4 and concentrated to an oil. This oil was purifi ed by trituration with hexane overnight. The suspension that resulted was filtered and th e filtrate evaporated to a yellow oil. N-Methyl carbamic acid pr opyl ester: yield = 60 %, 1H-NMR (CDCl3 ): 0.97 (t, 3H), 1.7 (m, 2H), 2.9 (d, 3H), 4.06 (t, 2H) 4.58 (s, 1H). N-Methyl carbamic acid but yl ester: yield = 49 %, 1H-NMR (CDCl3 ): 0.92 (t, 3H), 1.38 (m, 2H), 1.6 (t, 2H), 2.8 (d, 3H), 4.06 (t, 2H) 4.58 (s, 1H). In the case of N-methyl carbamic acid hexyl ester it was necessary to synthesize the chloroformate by reacting three molar equiva lents of hexanol with an equivalent of triphosgene and three equivale nts of a poorly nucleophilic ba se, triethylamine (Figure 32). A solution of triphosgene (11 mmol) in 100 mL CH2Cl2 was first prepared and placed in an ice bath. To this solution was added a mixture of triethylamine (33 mmol) and the alcohol (33 mmol) in 25 mL CH2Cl2 dropwise over 20 min. The exothermic reaction was allowed to proceed, the contents were allowe d to warm to room temperature and stirred overnight. To this preformed chloroformate wa s added methyl amine and triethylamine as stated above in the synthesis of N-alkyl carbamic acid alkyl esters. N-Methyl carbamic acid he xyl ester: yield = 59%, 1H NMR (CDCl3): 0.89 (t, 3h), 1.31 (m, 6H), 1.58 (m, 2H), 2.79 (d, 3H), 4.04 (t, 2H) 4.69(s, 1H). (c) N-Alkyl-N-alkyloxycarbonylaminomethyl chloride:

PAGE 95

79 A suspension of N-alkyl carbamic acid alkyl ester, 1.7 equivalents of paraformaldehyde and 13 equivalents of trimethylsilyl chloride were refluxed using a CaCl2 drying tube and water condenser for 2.5 h over an oil ba th. The suspension was diluted with CH2Cl2 and filtered to get rid of the unreacted paraformaldehyde. The clear filterate was concentrated with a rotavapor at 40C under reduced pressu re. The yellow oil obtai ned was triturated with hexane overnight, the white suspension obtained was filtered, and the filtrate was concentrated to give the desired alkylating agent. N-Methyl-N-propyloxycarbonylaminom ethyl chloride: yield = 85%, 1H NMR (CDCl3): 0.97 (t, 3H), 1.7 (m, 2H), 3.0 (s, 3H), 4.12 (t, 2H), 5.35 (s, 2H). N-Methyl-N-butyloxycarbonylaminomet hyl chloride yield = 82%, 1H NMR (CDCl3): 0.95 (t, 3H), 1.41 (m, 2H), 1.65 (m, 2H), 3.0 (s, 3H), 4.17 (t, 2H), 5.33 (s, 2H). N-Methyl-N-hexyloxycarbonylaminom ethyl chloride: yield = 83%, 1H NMR (CDCl3): 0.97 (t, 3H), 1.3 (m, 6H), 1.6 (m, 2H), 3.02 (s, 3H), 4.15 (t, 2H), 5.33 (s, 2H). Synthesis of NANAOCAM-APAP prodrugs (Table 3-1): Briefly a suspension of APAP (0.01mol) and triethylamine (0.011mo l) in 25 mL dichloromethane was stirred under reflux conditions for an hour followed by the addition of the alkylating agent, NANAOCAM-Cl (0.011mol). The contents we re stirred overnight and subsequently diluted with 50 mL dichloromethane and washed with water (3 10 mL). The dichloromethane layer was, dried over Na2SO4 and concentrated at 40C over a rotary evaporator. The yellow oil obtained was then purified by column chromatography using ethyl acetate: he xane as eluent.

PAGE 96

80 Table 3-1. NANAOCAM Prodrugs of APAP. O NHCOCH3 N R' O O R Compound N-RÂ’O-R 1 17 18 19 20 CH3 CH3 CH3 CH3 CH3 CH3 C2H5 C3H7 C4H9 C6H13 1 was prepared in 70 % yield from N-methyl-N-methyloxycarbonylaminomethyl chloride, triethylaminie and acetaminophen in CH2Cl2 after recrystallization of the oil obtained using CH2Cl2: hexane 1:3: mp = 86-88C, Rf (0.36,ether), 1H NMR(CDCl3): 7.6(s,1H), 7.39(d,2H), 6.96-6.87(2d,2H), 5.28-5.21(2s,2H), 3.72-3.7(2s,3H), 3.02.97(2s,3H), 2.0(s,3H). Anal calcd for C12H16N2O4 : C, 57.13; H, 6.39; N, 11.1. Found: C, 56.81; H, 6.34; N, 11. 17 was prepared in 65% yield from N-Met hyl-N-ethyloxycarbonyl aminomethyl chloride, triethylamine and acetaminophen in CH2Cl2 after trituration with hexane overnight to give yellow crystals which were recrystalliz ed from ethyl acetate:h exane (1:3) to give white crystals: mp = 75-77C, Rf (0.6, Et2O), 1H NMR(CDCl3): 7.38(d,2H), 7.12(s,1H), 6.96(dd,2H), 5.3(s,2H), 4.15(q,2H), 3.0(s,3H), 2.15(s,3H), 1.25(3H,t). Anal calcd for C13H18N2O4 : C, 58.64; H, 6.81; N, 10.52. Found: C, 58.53; H, 6.93; N, 10.47.

PAGE 97

81 18 was prepared in 76% yield from N-Methyl-N-propyloxycarbonyl aminomethyl chloride, triethylamine and acetaminophen in CH2Cl2 after silica gel column chromatography in ethyl acetate:hexane (3:2) followed by trituration in hexane overnight to give pale white crystals. These crystals were further recrystallized from ethyl acetate:hexane (3:2) to give white crystals: mp = 58-59C, Rf (0.2, ethyl acetate:hexane 3:2), 1H NMR(CDCl3): 7.38(d,2H), 7.14(s,1H), 6.96(dd,2H), 5.3(s,2H), 4.05(t,2H), 3.0(s,3H), 2.15(s,3H), 1.65(2H,m) 0.91(3H,t). Anal calcd for C14H20N2O4 : C, 60.01; H, 5.45; N, 12.73. Found: C, 60.1; H, 5.35; N, 12.58. 19 was prepared in 66 % yield from N-Methyl-N-butyloxy carbonyl aminomethyl chloride, triethylamine and acetaminophen in CH2Cl2 to give a colorless oil which was purified after silica gel column chromatogra phy in ethylacetate:hex ane (3:2) followed by trituration in hexane overnight to give a colorless oil: Rf = 0.13(3:7, Ethyl acetate: hexane), 1H NMR(CDCl3): 7.38(d,2H), 7.14(s,1H), 6.96(dd,2H), 5.3(s,2H), 4.05(t,2H), 3.0(s,3H), 2.15(s,3H), 1.65(2H,m), 1.4(2H,m), 0.91(3H,t). Anal calcd for C15H23N2O4 : C, 61.21; H, 7.53; N, 9.52. Fou nd: C, 61.16; H, 7.65; N, 9.24. 20 was prepared in 63 % yield from N-methyl-N-hexyloxy carbonylaminomethyl chloride, triethylamine and acetaminophen in CH2Cl2 to give a colorless oil which was purified after silica gel column chromatogra phy in ethyl acetate:hex ane (4:1) followed by trituration in hexane overnight to give a colorless oil: Rf (0.36, 7:3, ethyl acetate : hexane), 1H NMR(CDCl3): 7.38(d,2H), 7.29(s,1H), 6.95-6.89(2d,2H), 5.295.23(2s,2H), 4.11-4.07(m,2H), 3.02-2.98(2s,3H), 2.15(s,3H), 1.63(2H,m), 1.3(6H,m), 0.89(3H,m). Anal calcd for C17H27N2O4 : C, 62.33; H, 8.13; N, 8.69. Found: C, 61.96; H, 8.14; N, 8.36.

PAGE 98

82 Determination of Solubilities and Partition Coefficients Molar absorbtivities were determined in tr iplicate for each member of the series in acetonitrile (ACN) and pH 7.1 phosphate buffe r at the maximum absorption wavelength. The molar absorbtivities were calculated using BeerÂ’s law ( = A/c). The so lubilities of the prodrugs in isopropyl myristate (IPM) were determined in triplicate by stirring suspensions of the compound in 3 mL IPM w ith a magnetic stirrer for 24 h at room temperature (23 1C) (Marti n and coworkers, 1985 and Beall et al., 1994). The test tubes containing the suspensions were sealed and thermally insulated from the stirrer. After stirring, the suspensions were filtere d through a 0.45 m nylon membrane filter. An aliquot (~0.1-0.3 mL) was withdrawn from th e clear filtrate of saturated solutions and diluted to 10 mL in a volumet ric flask with ACN. The samples were then analyzed by UV spectroscopy using absorbances determ ined at 240 nm for APAP prodrugs. The solubility in IPM was calculated from the following relationship (equation 3-1): SIPM = (A/ ) (Dilution factor) (3-1) Dilution factor = (Vfinal/Valiquot) (3-2) where A is the absorbance of the sample at 240 nm, is the molar absortivity of the sample at 240 nm in ACN and Valiquot is the volume of the satu rated filterat e aliquot and Vfinal is the final diluted sample volume. In the case of prodrug derivatives which were oils, direct solubility measurements were not possible. Thus partition coefficients between IPM and pH 4.0 buffers were used to estimate SIPM. Solubilities in water (SAQ) were determined by stirri ng suspensions in deionized water for 1 h to limit the extent of hydrolysis of the prodrugs. The samples were filtered through nylon filters diluted with ACN and analyzed by UV spectroscopy. SAQ was determined using the following equation:

PAGE 99

83 SAQ = (A/ ) (Dilution factor) (3-3) Determination of SAQ of compounds which were oils was carried out by stirring the compound in deionized water for 1 h and ensuring that a biphasic solution was present at all times. The test tubes cont aining the biphasic solution we re centrifuged for 2 minutes; an aliquot (~0.1-0.3 mL) was withdrawn from the water layer and diluted to 10 mL with ACN in a volumetric flask. The samples were then analyzed by UV spectroscopy as above. For determination of partition coeffici ents between IPM and pH 4.0 buffer (KIPM: 4.0) for compounds which were solids at room temperature, a measured volume (~0.5 mL1 mL) of the filtered saturated IPM solutions from the lipid solubility experiments were mixed with a measured volume of pH 4.0 acet ate buffer (~1-5 mL) in a 10 mL test tube (Beall et al., 1993). The test tube was capped and vigorously shaken for 10 seconds and subsequently centrifuged for 2 min to allo w the clear separation of two phases. An aliquot (~0.3 mL) was withdraw n from the IPM layer and dilu ted to 10 mL with ACN in a volumetric flask and analyzed by UV spectroscopy. The KIPM: 4.0 was calculated using the following relationship: KIPM: 4.0 = (V4.0/VIPM) AF/ (AI-AF) (3-4) where V4.0 is the volume of pH 4.0 buffer used, VIPM is the volume of IPM used, AI is the initial absorbance of the saturated IP M solution before partitioning and AF is the absorbance of the compound remaining after partitioning. The pH 4.0 buffer solubility was estimated from KIPM: 4.0 using the following equation: S4.0 = SIPM / KIPM: 4.0 (3-5)

PAGE 100

84 Determination of partition coeffi cients between IPM and water (KIPM: AQ) for compounds which were oils at room temper ature was carried out by using a measured volume (~0.5 mL – 1.0 mL) of saturated solution used for determination of SAQ placed in a 10 mL test tube and a measured volume (~1-10 mL) of IPM. The test tubes were centrifuged; an aliquot was ta ken from the water layer and di luted with ACN to 10 mL in a volumetric flask. Samples were subse quently analyzed by UV spectroscopy. KIPM: AQ and SIPM were calculated usi ng the following equations: KIPM: AQ = (V4.0/VIPM) (AI-AF)/AF (3-6) SIPM = KIPM: AQ SAQ (3-7) Solubility ratios (SR) were calculated from the ratio of SIPM / SAQ. The methylene values were calculated using equation 3-8: = (log SR n+m – log SR n)/m (3-8) where n is the number of methylene units in the promoiety of one prodrug (the lowest member of the homologous series) and m is the number of additional units in the promoiety in the higher member of the hom ologous series. Similarly the methylene values (Leo et al., 1971 and Hansch and Leo, 1979) using K values are also reported. = (log K n+m – log K n)/m (3-9) Determination of Flux through Hairless Mice Skins (Beall et al., 1994) The mice were rendered unconscious using CO2 and sacrificed by cervical dislocation. Full thickness skins were removed by dissection; the pieces were scraped to remove excess fat, cut into proper sizes a nd placed dermal side down on the diffusion cells. A Franz diffusion cell consists of two compartments; donor compartment and a receptor compartment. The receptor compartment has a side arm which can be used to sample the compartment. The skins were he ld in place using rubber rings, and the two

PAGE 101

85 halves of the cells were clamped together. The receptor phase wa s maintained at 32C with a circulating water bath. A fully assemb led Franz diffusion is shown in Figure 3-3. The receptor side was filled with 20 mL pH 7.1 buffer containing 0.1% v/v formaldehyde (2.7 mL/37% aqueous formaldehyde) to preven t microbial growth (S loan et al., 1991). No air bubbles were present in the receptor side. A magnetic stir bar was added through the side arm of the receptor compartment a nd suspended over a stir plate to stir the contents throughout the experiment. The mouse sk ins were kept in contact with buffer for 48 h prior to application of the donor phase to condition the membranes; the receptor phase was replaced with fresh buffer at leas t twice to leach out any water soluble UV absorbing material present in the skin which would interfere with the UV quantification of acetaminophen. Pre-applica tion leach periods from 2-120 h were found to have no effect on the subsequent flux of theophy lline from a theophylline/propylene glycol suspension (33 mg/0.5 mL) (Sloan et al., 1986). The vehicle used to deliver the prodrugs was IPM; IPM is widely used in dermatological formulations. It is colorless, odorless and non-irritating to the skin (S tenback and Shubik, 1974 and Fr osch and Kligman, 1977) and has been used extensively in th is lab to create the database. In all cases, the prodrug was applied as a suspension in IPM. These suspensions were prepared by stirring the test compound fo r 24 h in 2 mL IPM at room temperature; the final suspension concentra tion exceeded the compounds solubi lity by at least ten fold. Application of suspensions of test co mpound in the donor phas e ensured that all compounds tested were at the same th ermodyanamic activity (Higuchi, 1960 and Woodford and Barry, 1982) and allowed ca lculation of the maximum flux for a compound.

PAGE 102

86 A 0.5 mL aliquot of a well stirred IPM suspension was evenly applied to the conditioned membrane surface. To obtain a sa mple from the receptor compartment, 5-6 mL of buffer Figure 3-3 A Franz diffusion cell (Reproduced with permission from S.C. Wasdo, Ph.D dissertation, University of Florida, 2005) was removed using a Pasteur pipette from the si de arm of the receptor and placed in a test tube for quantification using UV spectroscopy. In order to maintain sink conditions, the entire receptor contents we re changed each time a sample was removed. Samples were collected after 8 h, 19 h, 22 h, 25 h, 28 h, 31 h and 48 h of the initial application of the donor phase. Receptor phases were analyzed by UV within 24 h of sample collection. After the 48 h diffusion period, the rema ining donor suspension was removed by thoroughly washing the skin with methanol. Methanol wash was found to have minimal effect on the barrier properties of the skin in control studies (Koch et al., 1986). In order to quantify the amount of dermal penetrati on of the prodrug, the skins were kept in contact with buffer for an additional period of 24 h. The length of the post-application leach period was sufficient to remove 8590% of the residual compound in the skin (Siver, 1987). To evaluate the integrity of membra ne, a suspension of 33 mg/0.5 mL of ThH/propylene glcol was applied uniformly to the membrane surface. Samples were

PAGE 103

87 taken after 1 h, 2 h, 3 h, and 4 h and placed in test tubes for analysis. The entire receptor phase was changed with fresh buffer every ti me a sample from the receptor phase was removed. An increase in the flux of ThH comp ared to controls was an indication that the barrier function of the skin had been i rreversibly affected by the drug/vehicle combination (Sloan et al., 1986). Determination of Prodrug Hydrolysis by UV Spectroscopy Absorbance at any wavelength was assumed to be a combination of the absorbances drug and any intact prodrug. Usi ng BeerÂ’s law, the mathematical expression will be A = CPP + CDD (3-10) where A was the absorbance at a particular wavelength, CP concentration of prodrug, CD concentration of drug, P, molar absortivity of prodrug and D molar absortivity of drug at wavelength By measuring absorbances at two wa velengths it was possible to calculate CD and CP. A 1 = CPP 1 + CDD 1 (3-11) A 2 = CPP 2 + CDD 2 (3-12) Simulataneously solving these equations we arrive at CD and CP. CP = (A 1D 2 A 2D 1) / ( P 1D 2 P 2D 1) (3-13) CD = (A 1 CPP 1) / D 1 (3-14) CP and CD were then added to arrive at the to tal species of acetaminophen present. For APAP prodrugs, 1 was 240 nm and 2 was 280 nm. Calculation of Maximum Flux Maximum flux was calculated from the plot of the cumulative amounts of drug species delivered through the skin versus time. The cumulative amount permeated can be

PAGE 104

88 calculated by adding amount permeating thr ough the skin at each time interval or sampling time. The slope of the best fit li ne passing through the steady-state portion divided by the cross sectional ar ea of the diffusion cell (4.9 cm2) gives us the maximum flux, JM.in molecm-2h-1 A typical plot between cumula tive amount of drug species and time is shown in Figure 1-4 of Chapter 1. Physicochemical Properties of NANAOCAM Prodrugs of APAP All prodrugs of APAP had lower melting poi nts than the parent drug (Table 3-2). Only three of the five prodrugs were solids; 1, 17 and 18 while 19 and 20 were oils compared to the AOC-APAP series where al l prodrugs were solids. The melting points decreased as carbon chain lengths increased. Thus insertion of ‘CH2N(R’)’ into the AOC (ROCO-) promoiety to give the NANAOCAM promoiety leads to a greater decrease in crystalline lattice energy and hence melting points compared to AOC-APAP derivatives. Solubilities The SD of the solubilities in isopropyl my ristate and water were all less than 5%. The solubilities of prodrugs in IPM, water a nd solubility in pH 4.0 buffer are shown in Table 3-2 while molar absortivit ies, partition coefficients and values are shown in Table 3-3. Lipid solubilities measur ed in IPM increased as car bon chain lengths increased for the APAP series. The most lipid soluble member of the series 20 was 69 fold more soluble than APAP while the first member of the series 1 was 7.4 fold more soluble than APAP. The IPM solubilities of NANAOCAM-APAP were higher than AOC-APAP prodrugs (Wasdo and Sloan, 2004) probably because th e melting points of these prodrugs were considerably lower. Unlike the AOC-APAP series a regular increase in solubilities in IPM was seen along the series using the NANAOCAM promoiety.

PAGE 105

89 The direct water solubilities or pH 4. 0 of all APAP prodrugs were lower than APAP. The most water soluble member of the series 1 was 2 fold less water soluble than APAP. This is consistent with AOC-APAP pr odrugs where the most soluble member of the series, C1–AOC-APAP, was about 4 fold less so luble than APAP. Thus, the decrease in melting points is not enough to increase the solubility in water, because masking the hydrogen bonding phenolic group with NANAO CAM leads to an increase in the molecular weight to an extent that the decrea se in lattice energy isn’t able to compensate. Regardless of the irregular be havior of the absolute solubilities, the ratios of the solubilities in IPM and AQ (SRIPM: AQ) were reasonably well behaved. The average methylene SR was 0.53 0.05. The values obtained for NANAOCAM-APAP prodrugs closely relate to AOC-APAP se ries which had a value of 0.57. The SD for partition coefficients determined between IPM and pH 4.0 buffer (KIPM: 4.0) were all less than 10%. The average methylene K was 0.56 0.05 for the series. The value obtained is in close agreement w ith with other series of prod rugs (Beall, 2001; Sloan, 2003 and Wasdo, 2004). Thus values obtained using solubility ratios are in close agreement with values obtained using part ition coefficients between IP M and pH 4.0 buffer. This consistency makes values a robust indicator of cons istent homologous series behaviour.

PAGE 106

90 Table 3-2 Molecular Weights (MW), Melting Po ints (mp), Log Solubilities in Isopropyl Myristate (Log SIPM), Log Solubilities in Water (Log SAQ) and Log Solubilities in pH 4.0 Buffer (Log S4.0). Compound MW mp (C) Log SIPM (mM) Log SAQ (mM) Log S4.0 (mM) APAP 151 167-170 0.28 2 1.84 1, R = CH3, RÂ’ = CH3 252 86-88 1.15 1.66 1.61 17, R = C2H5, RÂ’ = CH3 266 75-77 1.3 1.2 1.22 18, R = C3H7, RÂ’ = CH3 280 59-60 1.61 0.95 0.91 19, R = C4H9, RÂ’ = CH3 294 1.83 0.56 nd 20, R = C6H13, RÂ’ = CH3 322 2.12 -0.16 nd Table 3-3 Molar Absortivities in Acetonitrile and Buffer ( ), Log Solubility Ratios between IPM and Water (log SRIPM:AQ ), the Differences Between Log SRIPM:AQ ( SR), the Log of Partition coeffi cients Between IPM and pH 4.0 Buffer (Log KIPM:4.0), and the Differences Between Log KIPM:4.0( K). a Units of 1 104 L/mole. b Molar absortivities measured at 240 nm for compounds APAP, 1, 17-20. c Buffer: pH 7.1 phosphate buffer with 0.11 % formaldehyde. d Molar absortivities measured at 240nm for compounds APAP, 1, 17-20. e Molar absortivities measured at 280nm for compounds APAP, 1, 17-20. Buffer a, c Compound CH3CNa b d e Log SRIPM:AQ SR Log KIPM:4.0 K APAP 1.36 0.8 0.12 -1.72 1 1.43 1.11 0.1 -0.51 -0.45 17 1.42 1.16 0.13 0.1 0.61 0.08 0.53 18 1.45 1.26 0.15 0.66 0.56 0.69 0.61 19 1.44 1.23 0.13 1.27 0.58 20 1.47 1.14 0.14 2.28 0.51

PAGE 107

91 Diffusion Cell Experiments The maximum flux obtained for NANAOCAM prodrugs of APAP are presented in Table 3-4. All JMIPM values were within the 30% variation in J values seen for in vitro hairless mouse skin diffusion cell experiments. Only two derivatives 1 (2.18 ) and 17 (1.22 ) gave higher delivery of tota l species from IPM than APAP itself. 18 in the APAP series gave a flux comparable to APAP alone. All other prodrugs delivered less parent drug + prodrug species thro ugh the skin than APAP. The C6 derivative, 20, performed worse in the series yet it was the mo st soluble in IPM. All prodrugs derivatives had higher lipid solubility than the parent drug however it was the most water soluble derivative that gave highest flux through the skin for each series. 1 was 7.4 times more lipid soluble, 0.46 times as water soluble AP AP but 3 that of other members of the series and gave the greatest flux through sk in. The most lipid soluble member of the series 20 (69 ) higher than APAP had the lowest flux (0.43 ) of APAP. 1 showed the least decrease in water solubility compared to APAP at the same time while exhibiting higher lipid solubility compared to the parent drug and hence performed better than other more lipid soluble members of the series whose water solubility was compromised. NANAOCAM prodrugs of AP AP perform marginally better than AOC-APAP derivatives. The best performing prodrug C1-AOC-APAP was 6.3 fold more lipid soluble, 0.29 fold water soluble and exhibited 1.7 fo ld higher flux than APAP through skin. The marginal increase in flux from C1-NANAOCAM-APAP could be accounted for by the higher aqueous solubility of the prodrug.

PAGE 108

92 Table 3-4 Solubilities in IPM (SIPM), Solubilities in Water (SAQ) and Flux (JMIPM) through in vitro Hairless Mouse Skins from IPM. Compound SIPM (mM) SAQ (mM) JMIPM ( molcm2h-) APAP 1.91 100 0.51 1, R = CH3, RÂ’ = CH3 14.13 45.71 1.11 17, R = C2H5, RÂ’ = CH3 19.95 15.85 0.62 18, R = C3H7, RÂ’ = CH3 40.74 8.91 0.51 19, R = C4H9, RÂ’ = CH3 67.61 3.63 0.43 20, R = C6H13, RÂ’ = CH3 131.83 0.69 0.22 Prodrug Bioconversion to Parent Drug. NANAOCAM prodrugs of APAP are stable to chemical hydrolysis as seen before in Chapter 2. In comparison to chemical hydrolysis, NANAOCAM derivatives of APAP are far more susceptible to enzymatic hydrol ysis. The percentage of intact prodrug found in the receptor phase was indicative of enzyma tic conversion to the parent drug. For the APAP series, the percentages of intact prodrug C1 to C6 (1, 17-20) in the receptor phase were 78, 77, 78, 76 and 80% respectively. Th e percentages of intact prodrug seen in receptor phases for the APAP series are not consistent with AOC-APAP series where the bioconversion of prodrug to APAP increases with increase in carbon chain length while in this case percentage of intact prodrug rema ins constant irrespecti ve of the alkyl side chain. Permeability Coefficients and Solubility Parameter Values When fluxes from IPM (JIPM) were divided by their corresponding solubility in IPM (SIPM) permeability coefficients (PMIPM) were obtained. The log PMIPM values of the prodrugs synthesized are given in Table 3-5.

PAGE 109

93 Table 3-5 Log Permeability Values for the APAP from IPM through Hairless Mouse Skins (log PMIPM) and Solubility Parameter values. Compound Log PMIPM i (cal cm-3)1/2 1 -1.10 12.71 17 -1.51 12.42 18 -1.90 12.17 19 -2.19 11.95 20 -2.78 11.58 PMIPM values decreased along with their resp ective calculated so lubility parameter for the both series. A plot of of log PMIPM vs i values for APAP prodrugs (1, 17-20) gave a positive slope (Figure 3-4). Such dependen ce has been seen before for lipophilic drugs of polar heterocyclic drugs like 5-FU and phe nolic drugs like APAP (Roberts and Sloan, 1999; Wasdo and Sloan, 2004). y = 1.4798x 19.901 R2 = 0.9994 -3 -2.5 -2 -1.5 -1 -0.5 0 11.51212.513 Solubility ParameterLog PMIPM 1 17 18 19 20 Figure 3-4 Plot of solubil ity parameter versus log PMIPM for 1, 17-20. For NANAOCAM-APAP prodrugs, log PMIPM decreases linearly with increasing size of the promoiety. This decrease in PMIP M can be explained by the increase in SIPM and decrease of JMIPM with increasing alkyl chain length. Residual Amounts in Skin The residual skin concentrations (Crs) for NANAOCAM prodrugs of APAP are presented in Table 3-6.

PAGE 110

94 Table 3-6. Residual Skin Concentrations of Total APAP (Crs) and Ratios of Dermal versus Transdermal Fluxes (D/T). Compound Crs SD ( mol) D/T 1 4.793 1.8 4.32 17 1.574 0.14 2.53 18 5.442 0.66 10.67 19 5.79 0.94 13.47 20 6.215 0.92 28.27 The most lipid soluble member of the series, 20, gave highest dermal delivery into the skin and also had the hi ghest D/T ratio. D/T ratios increase with increase in lipophilicity of the prodrugs except for 17. Compounds exhibiting higher JMIPM (trandermal flux) were not as effective in delivering total APAP species into the skin (dermal flux, Crs) and had lower D/T ratios. Thus if delivered appropriately these prodrugs can act as a reser voir of phenolic drugs. Second Application Fluxes. The second applicati on theophylline flux (JJIPM) values are presented in Table 3-7. Table 3-7 Second Application Theophylline Flux (JJIPM) data for Flux of Theophylline from Propylene Glycol. Compound JJIPM SD ( molcm2h-1) APAP 0.74 1 0.49 0.09 17 0.87 0.16 18 0.52 0.24 19 1.23 0.1 20 0.49 0.09 Skin penetration by theophylline from pr opylene glycol, was approximately the same or lower than control for skins treated with NANAOCA M prodrugs of APAP. Normalization of the JMIPM values by the respective JJIPM values did not change the rank order of the performances of APAP, 1, 17-20. Thus, the differences in JJIPM must be due

PAGE 111

95 to differences in the abilities of prodrugs to deliver APAP and not damage to the skin barrier. Modelling the Flux of NANAOCAM prodrugs of APAP through Hairless Mouse Skin from IPM using the RS equation. Application of the Roberts-Sloan (RS) equation to analysis of flux obtained from NANAOCAM prodrugs data allows quantification of the effect of IPM solubility, water solubility and molecular wei ght on flux through hairless mice skin.The first attempt to predict flux of the APAP prodr ugs was carried out using the RS equation derived from n = 63 compound database (e quation 3-15; Wasdo, 2005). Log JMIPM = -0.502 + 0.517 log SIPM + (1 0.517) log SAQ – 0.00266 MW (3-15) The inclusion of five APAP prodrugs (1, 17-20) revised the coefficients of the RS model as shown in equation 3-16. Log JMIPM = -0.356 + 0.53 log SIPM + (1 0.53) log SAQ – 0.00336 MW (r2 = 0.91, n = 68) (3-16) NANAOCAM prodrugs of APAP underperformed regardless of the model used. The residual values (exp log JMIPM – predicted log JMIPM = log JMIPM) for compounds 1, 1720 was 0.27 log units according to equation 3.15 and 0.23 log units using equation 3.16. A plot experimental vs calculated log JMIPM through hairless mouse skins is shown using equation 3.15 and 3.16 in Figures 3-6 and 3-7, respectively. The residual values for the entire database of n = 68 was ~0.16 using e quations 3.15 and 3.16 respectively. Both the models correctly identified the best perf orming members of the APAP series and rank order of the performance of prodrugs was also correctly identified.The experimental flux, calculated flux and error in predicting flux ( log JMIPM) using each model is given in Table 3-8.

PAGE 112

96 Table 3-8 Experimental Flux (log JMIPM), Calculated Flux (Calc. log JMIPM) and Error in Predicting Flux ( log JMIPM) for Compounds 1, 17-26 through Hairless Mouse Skins from IPM. -2.5 -1.5 -0.5 0.5 1.5 2.5 3.5 -2.5-0.51.53.5 Calculated log JMIPMlog JMIPM = 0.502+0.517 log SIPM + (1-0.517) log SAQ0.00266 MWExperimental log JMIPM n = 63 database NANAOCAM prodrugs of APAP Figure 3-5 Experimental vers us calculated log maximum flux values through hairless mouse skin from IPM using equation 3.15. Compound log JMIPM Calc.log JMIPM Eq 3.15 log JMIPM Eq 3.15 Calc. Log JMIPM Eq 3.16 log JMIPM Eq 3.16 1 0.05 0.22 0.18 0.18 0.14 17 -0.21 0.04 0.25 0.0009 0.21 18 -0.29 0.04 0.34 0.002 0.29 19 -0.37 -0.07 0.29 -0.11 0.26 20 -0.66 -0.34 0.32 -0.39 0.27

PAGE 113

97 -2.5 -1.5 -0.5 0.5 1.5 2.5 3.5 -2.5-0.51.53.5 Calculated log JMIPMlog JMIPM = 0.356+0.53 log SIPM + (1-0.53) log SAQ 0.00336 MWExperimental log JMIPM n = 63 database NANAOCAM prodrugs of APAP Figure 3-6 Experimental vers us calculated log maximum flux values through hairless mouse skin from IPM using equation 3.16. Conclusions NANAOCAM derivatives of acetaminophen were found to be enzymatically labile. The slow hydrolysis rates of NANAOCAM de rivatives of phenol containing drugs makes them potentially useful candidates to prev ent their premature metabolism during oral drug delivery in addition to increasing their topical delivery. The delivery of the total acetaminophen (A PAP) containing speci es by its N-alkylN-alkyloxycarbonylaminomethyl (NANAOCAM) de rivatives from IPM is enhanced compared to the parent drug molecule. The more water soluble member of the more lipid soluble series was the most effective at e nhancing the delivery of total APAP species through mouse skin from IPM. Insertion of ‘CH2(NR’)’ into the AOC promoiety increases biphasic solubility and flux of phenolic drugs and the NANAOCAM-APAP derivatives perform better than AOC-APAP

PAGE 114

98 The amount of drug species delivered could be reasonably predicted by the RS equation. The addition of 5 new compounds to the database of exis ting 63 compounds for the RS equation gave new coefficients which we re not significantly different from the previous coefficients. The result reinforces the importance of biphasi c solubility to flux. The y coefficient in the RS equation (3.16) suggests the c ontribution of solubility in a lipid phase (SLIPID) to be 0.53 and solubility in a water phase (SPOLAR) to be 0.47 on a scale of 1. The dependence of flux on water so lubility holds regardle ss of the degree of hydration in vitro or in vivo, human skin or mouse skin, mineral oil, propylene glycol, water or IPM (Sloan and Wasdo, 2004).

PAGE 115

99 CHAPTER 4 SYNTHESIS AND TOPICAL DELIVERY OF N-ALKYL-NALKYLOXYCARBONYLAMINOMETHYL PRODRUGS OF AN IMIDE CONTAINING DRUG: THEOPHYLLINE Just like the phenolic func tional group, an imide (aci dic –NH group) is commonly found in drug molecules; examples include 6MP, 5-FU, 5-Iodo-2-deoxycytosine and 5fluorocytosine. We chose theophylline (ThH ) as our model imide containing drug. Theophylline is commonly used to treat as thma and is partially effective when given orally or topically to treat psoriasis (Iancu et al., 1979 and Berenbein et al., 1979). It inhibits phosphodiesterases and leads to an increase in cAMP leve ls whose levels in psoriatic skin are severely compromised (Bour ne et al., 1974; Voorh ees and Duell, 1971). Since theophylline exhibits a narrow therapeu tic range increasing the oral dose to treat psoriasis is not a viable op tion. Topical delivery of the ophylline is limited by its poor solubility in the skin. Prod rugs of theophylline could potentially solve this problem enabling this old drug to be topically effectiv e in treating psoriasis. Sloan and coworkers have reported ACOM-Th, AOC-Th, AC-Th and mannich bases of ThH as useful prodrugs of ThH (Sloan and Bodor, 1982; Sloa n et al., 1984; Kerr et al., 1998 and Sloan et al., 2000). AOC, AC and Mannich bases (-CH2NRR’) of ThH rely on chemical hydrolysis to give ThH. While AOC and AC hydrolyse by addition elimination type of mechanism, Mannich bases hydrolyse by a SN1 type of mechanism. Since the leaving group in both cases is Thwhere ThH has a pKa of 8.6, the rates of hydrolysis are very fast, e.g. 7-AOC-Th has a t 1/2 = 1-3 min, 7-AC-Th has a t 1/2 <1 min and Mannich bases of ThH have t 1/2 <1 min. Prodrugs with such short half lives make pharmaceutical

PAGE 116

100 formulation very difficult. The sh elf life of compounds with su ch short half-lives in water means they require anhydrous st orage conditions. This makes it difficult to keep costs of marketing the compound down. The NANAOCA M promoiety, where nitrogen is built into the promoiety that is conjugated to ThH, should have improved stability because it doesnÂ’t rely on an addition elimin ation type of pathway but on a SN1 type of mechanism that is more stable than Mannich bases. The ability of the nitrogen to donate electrons and stabilize the carbocation formed in the transition state is diminished in NANAOCAM derivatives (ROCONRCH2-, R = alkyl) compared to Mannich bases (RRÂ’N-CH2-, R = RÂ’ = alkyl). This is because of presence of an electron withdrawing alkyloxycarbonyl (ROCO) attached dir ectly to the NRÂ’-CH2 nitrogen in NANAOCAM compared to Mannich bases where the nitroge n is attached to electron don ating alkyl groups which aid in stabilizing the ca rbocation and increase its rate of hydrolysis over the NANAOCAM promoiety (Figure 4-1). Introduc tion of a nitrogen hetero atom in the promoiety should also give better biphasic so lubility because nitrogen is basic and can act as hydrogen bond acceptor and, hence improve flux through the skin. Th N R R' Th N COOR R' N R R' N COOR R' + + Th Th + + more stable less stable Figure 4-1 Carbocations formed as interm ediates by hydrolysis of Mannich bases of theophylline and NANAOCAM-Th. In Chapter 2, we looked at the chemi cal hydrolysis of NANAOCAM-X conjugates in aqueous buffers, where X was phenol, carboxylic acid or 6MP. In the preceding

PAGE 117

101 chapter we looked at the topical deliver y of NANAOCAM prodrugs of acetaminophen, here we extend the use of the NANAOCAM promoiety to imide containing drugs like theophylline. In this chapter we will i nvestigate the synthesis, physicochemical characterization and diffusi on of NANAOCAM derivatives of theophylline through in vitro hairless mouse skins from IPM. The performance of the NANAOCAM prodr ugs of theophylline in diffusion cell experiments were compared with that predicted by the Roberts Sloan (RS) equation. The results were added to the Sloan database to generate new coefficients for the parameters in the equation and give a more robust RS equation. Experimental Procedure Materials and Methods Isopropyl myristate (IPM) was obtained from Givaudan Corp (Clifton, NJ). Theophylline (ThH) was purchased from Sigma Chemical Co. (St. Louis, MO, USA); all other reagent chemicals were from Aldric h Chemical Co. (Milwaukee, WI, USA). The water was obtained from a Millipore Milli-Q water ultra filteration system. Ultraviolet spectra were recorded on a Shimadzu UV2501 PC spectrophotometer. A radiometer pH meter 26 was used to determine pH of soluti ons. The vertical, Franz type diffusion cells were from Crown Glass (Somerville NJ, USA) (surface area 4.9cm2, 20 mL receptor phase volume, 15 mL donor phase volume). Th e diffusion cells were maintained at 32C with a Fischer (Pittsburgh, PA, USA) circ ulating water bath model 25. The female hairless mice (SKH-hr-1) were from Charle s River (Boston, MA, USA). Statistical analyses were carried out by using SAS 9.0. The animal research adhered to the NIH “Principles of Laboratory Animal Care.”

PAGE 118

102 Synthesis of Prodrug Derivatives Synthesis begins by the alkylation of ThH with N-alkyl-N-alkyloxycarbonylaminomethyl chloride (NANAOCAM-Cl) in pr esence of a base like triethylamine and CH2Cl2 as the solvent. In every case it was necessary to synthesize the corresponding alkylating agent, NANAOCAM-Cl. N O O Cl R CH2O CH3NH2 NN N Cl O OR R= CH3 and C2H5 DCM + NaOH Figure 4-2 Synthesis of NAN AOCAM-Cl from 1, 3, 5trialkylhexahydrotriazine. Synthesis of NANAOCAM-Cl a) 1, 3, 5-Trimethylhexahydrotriazine was s ynthesized from equimolar equivalents of aqueous formaldehyde, methyl amine and NaOH according to the protocol originally developed by Graymore et al., (1932) and m odified by Siver et al ., (1990). Methyl amine (0.4 mol of 40% aqueous) was placed in an ice bath and an equivalent of 37% aqueous formaldehyde was added dropwise over a peri od of 10 min. The solution was allowed to equilibrate to room temperature and stirred for one hour, then an equivalent of NaOH was added and the contents were stirred for 1.5 h more. The solution was extracted with 4 50 mL CH2Cl2, dried over Na2SO4, filtered and concentrated to a clear colorless CH2Cl2 solution containing the the hexahydrotriazine derivative. Complete evaporation of CH2Cl2 wasn’t carried out as it resulted in lo ss of the desired product. For quantification purposes the ‘CH2’ of CH2Cl2 at 5.3 and N-CH2-N peak of hexahydr otriazine derivative at 3.2 were used, yield = 82% in CH2Cl2, 1H NMR (CDCl3): 2.3 (s, 9H), 3.2 (s, 6H).

PAGE 119

103 Next the 1, 3, 5-trimethylhexahydrot riazine (freshly prepared) in CH2Cl2 was added to an ice cold solution of 3 equivalent of alkyl chloroformate in CH2Cl2 over a period of 10 minutes. The white suspension that was observe d was allowed to equlilibrate to room temperature and stirred overnight. The suspension was filtered and the filtrate concentrated to oil. The oil contained th e desired product and co rresponding bis(N-AlkylN-alkyloxycarbonyl)aminomethanes with a ‘CH2’ peak at 4.8.The oil was purified by trituration with hexane overnight followed by ether overnight. The clear solution was decanted leaving the white residue (bis deri vative) behind. The clear solution was then concentrated. The peak at 4.8 due to the ‘CH2 ’ of the bis derivative and the peak at 5.3 due to the ‘CH2’ of N-alkyl-N-alkyloxycarbonylaminom ethyl chloride were used to quantitate the amount of product formed. N-Methyl-N-methyloxycarbonylaminom ethyl chloride: yield = 90%, 1H NMR (CDCl3): 2.9 (s, 3H), 3.75 (s, 3H), 5.3 (s, 2H). N-Methyl-N-ethyloxyoxycarbonylaminom ethyl chloride: yield = 89%, 1H NMR (CDCl3): 1.3 (t, 3H), 2.9 (s, 3H), 4.22 (q, 2H), 5.33 (s, 2H). N-Methyl-N-propyloxycarbonylaminom ethyl chloride, N-methyl-Nbutyloxycarbonyl aminomethyl chloride, Nethyl-N-methyloxycarbonylaminomethyl chloride and N-methyl-N-hexyloxycarbonylaminom ethyl chloride were synthesized in an alternate way via a N-alkyl carbamic acid alkyl ester (Figure 4-3A).

PAGE 120

104 R'NH2 RO O N H R' TMSCl THF (CH2O)n RO O N R' Cl Cl3CO OCCl3 O O Cl RO O Cl RO R' = CH3, C2H5R = C3H7, C4H9, C6H13 + TEA 3 ROH + TEA Figure 4-3 Synthesis of NANAO CAM-Cl from alkyl amine and preperation of alkyl chloroformate in situ from alcohol b) N-Alkyl carbamic acid alkyl ester. To a solution of chloroformate (33 mmol) in 75 mL CH2Cl2 mounted in an ice bath, was added an equivalent of triethylamine and a queous alkyl amine. Th e reaction mixture was allowed to warm to room temperature and stirred overnight. The clear solution was washed with 3 10 mL brine and the organic layer was dried over Na2SO4 and concentrated to an oil. This oil was purifi ed by trituration with hexane overnight. The suspension that resulted was filtered and th e filtrate evaporated to a yellow oil. N-Methyl carbamic acid propyl ester: yield = 60 %, 1H-NMR (CDCl3 ): 0.97 (t, 3H), 1.7 (m, 2H), 2.9 (d, 3H), 4.06 (t, 2H) 4.58 (s, 1H). N-Methyl carbamic acid but yl ester: yield = 49 %, 1H-NMR (CDCl3 ): 0.92 (t, 3H), 1.38 (m, 2H), 1.6 (t, 2H), 2.8 (d, 3H), 4.06 (t, 2H) 4.58 (s, 1H). N-Ethyl carbamic acid met hyl ester: yield = 69 %, 1H-NMR (CDCl3): 1.05 (t, 3H), 3.1 (q, 2H), 3.7 (s, 2H), 4.58 (s, 1H). In the case of N-methyl carbamic acid hexyl ester it was necessary to synthesize the chloroformate by reacting three molar equiva lents of hexanol with an equivalent of

PAGE 121

105 triphosgene and three equivalents of a poorly nucleophilic base, trie thylamine (Figure 43B). A solution of triphosgene (11 mmol) in 100 mL CH2Cl2 was first prepared and placed in an ice bath. To this solution wa s added a mixture of triethylamine (33 mmol) and the alcohol (33 mmol) in 25 mL CH2Cl2 dropwise over 20 min. The exothermic reaction was allowed to proceed, the contents were allowed to warm to room temperature and stirred overnight. To this preformed chloroformate was added methyl amine and triethylamine as stated above in the synthe sis of N-alkyl carbamic acid alkyl ester. N-Methyl carbamic acid hexyl ester: yield = 59%, 1H NMR (CDCl3): 0.89 (t, 3h), 1.31 (m, 6H), 1.58 (m, 2H), 2.79 (d, 3H), 4.04 (t, 2H) 4.69(s, 1H). (c) N-Alkyl-N-alkyloxycarbonylaminomethyl chloride: A suspension of N-alkyl carbamic acid alkyl ester, 1.7 equivalents of paraformaldehyde and 13 equivalents of trimethylsilyl chlo ride was refluxed fo r 2.5 h using a CaCl2 drying tube, water condenser and an oil bath The suspension was diluted with CH2Cl2 and filtered to get rid of the unr eacted paraformaldehyde. The clear filterate was concentrated with rotavapor at 40C under reduced pressu re. The yellow oil obtai ned was triturated with hexane overnight. The white suspension obtained was filtered, and the filtrate was concentrated to give the desired alkylating agent. N-Methyl-N-propyloxycarbonylaminom ethyl chloride: yield = 85%, 1H NMR (CDCl3): 0.97 (t, 3H), 1.7 (m, 2H), 3.0 (s, 3H), 4.12 (t, 2H), 5.35 (s, 2H). N-Methyl-N-butyloxycarbonylaminomet hyl chloride yield = 82%, 1H NMR (CDCl3): 0.95 (t, 3H), 1.41 (m, 2H), 1.65 (m, 2H), 3.0 (s, 3H), 4.17 (t, 2H), 5.33 (s, 2H). N-Methyl-N-hexyloxycarbonylaminom ethyl chloride: yield = 83%, 1H NMR (CDCl3): 0.97 (t, 3H), 1.3 (m, 6H), 1.6 (m, 2H), 3.02 (s, 3H), 4.15 (t, 2H), 5.33 (s, 2H).

PAGE 122

106 N-Ethyl-N-methyloxyoxycarbonylaminomet hyl chloride: yield = 85%, 1H NMR (CDCl3): 1.25 (t, 3H), 3.45 (s, 2H), 3.8 (s, 3H), 5.33 (s, 2H). (d) Synthesis of NANAOCAM-Th (Figure 4-4). Theophylline (5.5 mol, 1 equivalent), triethylamine (1.1 equivalents) and 20 mL CH2Cl2 were stirred for 20 min. A white suspensi on formed. The alkylating agent (N-alkyl-Nalkyloxycarbonylaminomethyl chloride) was then added to the suspension. An exothermic reaction occured and the suspen sion cleared to give a solution which was stirred overnight at room temperature. Th e reaction mixture was diluted with 50 mL CH2Cl2 and extracted with 10 mL 1N HCl, 20 mL NaHCO3 solution and 3 50 mL water. The CH2Cl2 layer was then dried over Na2SO4 and concentrated to give a white solid which was crystallized from CH2Cl2: ether (1:4) to give pur e white crystals of N7(Nalkyl-N-alkyloxycarbonyl)amin omethyl theophyllines. N N N N O O N O OR R' N N N N H O O N R' Cl O OR TEA, DCM,Rt. Overnight Figure 4-4 Alkylation of th eophylline with NANAOCAM-Cl.

PAGE 123

107 Table 4-1 NANAOCAM Prodrugs of Theophylline. N N N N O O N R' O O R Compound N-RÂ’ O-R 21 22 23 24 25 26 CH3 CH3 CH3 CH3 CH3 C2H5 CH3 C2H5 C3H7 C4H9 C6H13 CH3 21: N7-(N-Methyl-N-methyloxycarbonyl)aminomethyl theophylline. Purified using CH2Cl2: hexane (1:4) to give white crys tals, yield = 73%, mp = 165-166C, Rf (0.49, EtOAc), 1H NMR(CDCl3): 7.98-7.75 (2s,1H), 5.8 (s,2H), 3.8-3.75 (2s,3H), 3.61 (s,3H), 3.43 (s, 3H). Anal calcd for C11H15N5O4 : C, 46.97, H, 5.38; N, 24.9. Found: C, 46.88; H, 5.22; N, 24.77. 22: N7-(N-Methyl-N-ethyloxycarbonyl)aminomet hyl theophylline. Purified using CH2Cl2: ether (1:4) followed by CH2Cl2: hexane (1:4) to give pur e white crystals, yield = 84%, Rf (0.21, ether), mp = 115-117 C, 1H NMR (CDCl3): 7.98-7.78 (2s,1H), 5.8 (s,2H), 4.2-4.11 (2q,3H), 3.61 (s,3H), 3.43 (s,3H), 3.1 (s,3H), 1.36-1.28 (2t,3H). Anal calcd for C12H17N5O4 : C, 48.81, H, 5.8; N, 23.72. Found: C, 48.8; H, 5.84; N, 23.67. 23: N7-(N-Methyl-N-propyloxycarbonyl)aminomethyl theophylline. Purified using CH2Cl2: hexane (1:4) followed by hot EtOH to gi ve pure white crystals, yield = 80%, mp = 128-129C, Rf (0.24, ether), 1H NMR (CDCl3): 7.98-7.75 (2s, 1H), 5.8 (s, 2H),

PAGE 124

108 4.17-4.08 (2t, 2H), 3.6 (s, 3H), 3.43 (s, 3H), 3.1 (s, 3H), 1.74-1.61 (m, 2H), 1-0.92 (2t, 3H). Anal calcd for C13H19N5O4 : C, 50.48; H, 6.19; N, 22.64. Found: C, 50.60; H, 6.29; N, 22.62. 24: N7-(N-Methyl-N-butyloxycarbonyl)aminomethyl theophylline. Purified using CH2Cl2:hexane (1:3) to give white crys tals, yield = 71%, mp = 103C, Rf (0.29, ether); 1H NMR (CDCl3): 7.98-7.78 (2s, 1H), 5.8 (s, 2H), 4.2-4.12 (2t, 2H), 3.61 (s, 3H), 3.42 (s, 3H), 3.01 (s, 3H), 1.7-1.63 (m, 2H), 1.4-1.35 (m, 2H), 0.94 (2t, 3H). Anal calcd for C14H21N5O4 : C, 52.0; H, 6.55; N, 21.66. Found: C, 52.17; H, 6.73; N, 21.68. 25: N7-(N-Methyl-N-hexyloxycarbonyl)aminomethyl theophylline. Purified using CH2Cl2:petroleum ether (1:3) to give white cr ystals, yield = 59%, mp = 74-75C, Rf (0.22, ether); 1H NMR (CDCl3): 7.99-7.74 (2s, 1H), 5.8 (s, 2H), 4.2-4.12 (2t, 2H), 3.61 (s, 3H), 3.42 (s, 3H), 3.01 (s, 3H), 1.63 (m, 2H), 1.31 (m, 6H), 0.89 (m, 3H). Anal calcd for C16H25N5O4 : C, 54.69,H, 7.17; N, 19.93. Found: C, 54.18; H, 7.13; N, 19.53. 26: N7-(N-Ethyl-N-methyloxycarbonyl)aminomet hyl theophylline. Purified using CH2Cl2: hexane (1:4) followed by CH2Cl2: hexane (1:4) to give pure white crystals, yield = 80%, Rf (0.14, ether), mp = 145-147C, 1H NMR(CDCl3): 8.03-7.78 (2s,1H), 5.81 (s,2H), 3.84-3.77 (2s,2H), 3.61 (s,3H), 3.53 (q, 2H), 3.43 (s,3H), 1.13 (2t,3H). Anal calcd for C12H17N5O4 : C, 48.81, H, 5.8; N, 23.72. Found: C, 48.78; H, 5.74; N, 23.77. Determination of Solubilities and Partition Coefficients Molar absorbtivities were determined in tr iplicate for each member of the series in acetonitrile (ACN) and pH 7.1 phosphate buffer at the absorption wavelength. The molar absorbtivities were calculated using BeerÂ’s law ( = A/c). The solubilities of the prodrugs

PAGE 125

109 in isopropyl myristate (IPM) were determined in triplicate by stirring suspensions of the compound in 3 mL IPM with a magnetic stirre r for 24 h at room temperature (23 1C) (Beall et al., 1994). The test tubes containing the suspensions were sealed and thermally insulated from the stirrer. After stirring, the suspensions were filtered through a 0.45 m nylon membrane filter. An aliquot (~0.1-0.3 mL) was withdrawn from the clear filtrate of saturated solutions and dilute d to 10 mL in a volumetric flask with ACN. The samples were then analyzed by UV spectroscopy and absorbances determined at 273 nm. The solubility in IPM was calculated from the following relationship (equation 4-1): SIPM = (A/ ) (Dilution factor) (4-1) Dilution factor = (Vfinal/Valiquot) (4-2) where A is the absorbance of the sample at 273 nm, is the molar absortivity of the sample at 273 nm in ACN and Valiquot is the volume of the satu rated filterat e aliquot and Vfinal is the final diluted sample volume. Solubilities in water (SAQ) were determined by stirri ng suspensions in deionized water for 1 h to limit the extent of hydrolysis of the prodrugs. The samples were filtered through nylon filters and analyzed by UV spectroscopy. SAQ values were determined using the following equation: SAQ = (A/ ) (Dilution factor) (4-3) For determination of partition coeffici ents between IPM and pH 4.0 buffer (KIPM: 4.0) a measured volume (~0.5 mL-1 mL) of the filtered saturated IPM solutions from the lipid solubility experiments were mixed w ith measured volume of pH 4.0 acetate buffer (~1-5 mL) in a 10 mL test tube (Beall et al., 1993). The test tube was capped and vigorously shaken for 10 seconds and subsequently centrifuged for 2 min to allow the

PAGE 126

110 clear separation of two phases. An aliquot (~0.3 mL) was withdrawn from the IPM layer and diluted to 10 mL with ACN in a volum etric flask and analyzed by UV spectroscopy as above. The KIPM: 4.0 was calculated using the following relationship: KIPM: 4.0 = (V4.0/VIPM) AF/ (AI-AF) (4-4) where V4.0 is the volume of pH 4.0 buffer used, VIPM is the volume of IPM used, AI is the initial absorbance of the saturated IP M solution before partitioning and AF is the absorbance of the compound remaining after partitioning. The pH 4.0 buffer solubility can be estimated from KIPM: 4.0 using the following equation: S4.0 = SIPM / KIPM: 4.0 (4-5) Solubility ratios (SR) were calculated from the ratio of SIPM / SAQ. The methylene values were calculated using equation: = (log SR n+m –log SR n)/m (4-6) where n is the number of methylene units in the promoiety of one prodrug (the lowest member of the homologous series) and m is the number of additional units in the promoiety in the higher member of the hom ologous series. Similarly the methylene values (Leo et al., 1971 and Hansch and Leo, 1979) using K values are also reported. = (log K n+m –log K n)/m (4-7) Determination of Flux through Hairless Mice Skins (Beall et al., 1994) The mice were rendered unconscious using CO2 and sacrificed by cervical dislocation. Full thickness skins were removed by dissection; the pieces were scraped to remove excess fat, cut into proper sizes a nd placed dermal side down on the diffusion cells. A Franz diffusion cell consists of two compartments; donor compartment and a receptor compartment. The receptor compartment has a side arm which can be used to inject and eject sample into or out of the compartment. The skins were held in place using

PAGE 127

111 round rubber rings and the two halves of the cells were clamped together. The receptor phase was maintained at 32C with a circul ating water bath. A fully assembled Franz diffusion is shown in Figure 3-3 in the previous chapter. The receptor side was filled with 20 mL pH 7.1 buffer containing 0.1% v/v formaldehyde (2.7 mL/ L of 37% aqueous formaldehyde) to prevent microbial growth (Sloan et al., 1991) ensuring that no air bubbles were present in the re ceptor side. A magnetic stir bar was added through the side arm of the receptor compartment and suspende d over a stir plate to stir the contents throughout the experiment. The mouse skins were kept in contact with buffer for 48 h prior to application of the donor phase to condition the memb ranes; the receptor phase was replaced with fresh buffer at least twic e to leach out any water soluble UV absorbing material present in the skin which would inte rfere with the quantification of theophylline. Pre-application leach periods from 2-120 h were found to have no effect on the subsequent flux of theophylline from a th eophylline/propylene glycol suspension (33 mg/0.5 mL) (Sloan et al., 1986). The vehicle used to deliver the prodrugs was IPM; IPM is widely used in dermatological formulati ons. It is colo rless, odorless an d non-irritating to the skin (Stenback and Shubik, 1974 and Frosch and Kligman, 1977) and has been used extensively in this lab to create the database. In all cases, the prodrug was applied as a suspension in IPM. These suspensions were prepared by stirring the test compound fo r 24 h in 2 mL IPM at room temperature; the final suspension concentration exceeded the compounds solubility by at least ten fold. Application of suspensions of test co mpound in the donor phase ensures that all compounds tested are at the same thermodya namic activity (Higuchi, 1960 and Woodford and Barry, 1982) and allows calculation of the maximum flux for a compound.

PAGE 128

112 A 0.5 mL aliquot of a well stirred IPM suspension was evenly applied to the conditioned membrane surface. To obtain a sa mple from the receptor compartment, 5-6 mL of buffer was removed using a Pasteur pi pette from the side ar m of the receptor and placed in a test tube for quantification usi ng UV spectroscopy. In order to maintain sink conditions, the entire receptor contents we re emptied and filled with fresh buffer. Samples were collected after 8 h, 19 h, 22 h, 25 h, 28 h, 31 h and 48 h of the initial application of the donor phase. Receptor pha ses were analyzed by UV within 24 h of sample collection using the molar absortivities as in Table 4-4. After the 48 h diffusion period, the rema ining donor suspension was removed by thoroughly washing the skin with methanol. Methanol wash was found to have minimal effect on the barrier properties of the skin in control studies (Koch et al., 1986). In order to quantify the amount of dermal penetrati on of the prodrug, the skins were kept in contact with buffer for an additional period of 24 h. The length of the post-application leach period was sufficient to remove 8590% of the residual compound in the skin (Siver, 1987). To evaluate the integrity of membra ne, a suspension of 33 mg/0.5 mL of ThH/propylene glycol was applied uniformly to the membrane surface. Samples were taken after 1 h, 2 h, 3 h, 4 h and placed in test tubes for further analysis. The entire receptor phase was changed with fresh buffer every time a sample from the receptor phase was taken. An increase in the flux of theophylline compared to controls was an indication that the barrier function of the skin had been irreversibly affected by the drug/vehicle combination (Sloan et al., 1986).

PAGE 129

113 Determination of Prodrug Hydrolysis by UV Spectroscopy Absorbance at any wavelength was assumed to be a combination of the absorbances drug and any intact prodrug. Usi ng BeerÂ’s law, the mathematical expression will be: A = CPP + CDD (4-8) where A was the absorbance at a particular wavelength, CP concentration of prodrug, CD concentration of drug, P, molar absortivity of prodrug and D molar absortivity of drug at wavelength By measuring absorbances at two wa velengths it was possible to calculate CD and CP. A 1 = CPP 1 + CDD 1 (4-9) A 2 = CPP 2 + CDD 2 (4-10) Simultaneously solving these equations we arrive at CD and CP. CP = (A 1D 2 A 2D 1) / ( P 1D 2 P 2D 1) (4-11) CD = (A 1 CPP 1) / D 1 (4-12) CP and CD were then added to arrive at the tota l species of theophylline present. For ThH prodrugs, 1 was 260 nm and 2 was 284 nm. Calculation of Maximum Flux Maximum flux was calculated from the pl ot of the cumulative amounts of drug species permeated through the skin versus time. The cumulative amount permeated can be calculated by adding amount permeating th rough the skin at each time interval or sampling time. The slope of the best fit li ne passing through the steady-state portion divided by the cross sectional ar ea of the diffusion cell (4.9 cm2) gave us the maximum flux, JM.in mole cm-2 h-1 A typical plot between cumula tive amount of drug species and time is shown in Figure 1-4 of Chapter 1.

PAGE 130

114 Physicochemical Properties of NANAO CAM Prodrugs of Theophylline Melting Point Behaviour of NANAOCAM Prodrugs of Theophylline All prodrugs of ThH had lower melting point s than the parent drug (Table 4-2 and 4-3). All prodrugs of the ho mologous series were solids at room temperature. All prodrugs had lower melting points than ThH itself. The decrease in melting points with increasing carbon chain lengths along the series were co mparable to the AOC-Th and ACOM-Th prodrug derivatives (Table 4-2). Th e even carbon chain member of the series had higher melting points than the odd carbon ch ain length member. Th is alternate rise and fall of melting points on addition of a methylene unit to the alkyl chain of a promoiety along a series has previously been observed for several series of homologous series (Chikos, 2001 and Wasdo and Sloan, 2004). Solubilities The SD of the solubilities in isopropyl my ristate and water were all less than 5%. The solubilities of prodrugs in IPM, water a nd solubility in pH 4.0 buffer are shown in Table 4-3 while molar absortivities, partition coefficients and values are shown in Table 4-4. NANAOCAM prodrugs in the theophylline series showed an irregular increase and decrease in IPM solubility was seen. All prodr ugs were more lipid soluble than ThH. IPM solubility generally increases as melting point decreases along the se ries. IPM solubility

PAGE 131

115 Table 4-2 Melting Point Comparisons of NANAOCAM-Th with AOC-Th and ACOMTh. N N N N O O O O R N N N N O O O O R N N N N O O N R' O O R NANAOCAM-Th AOC-Th ACOM-Th NANAOCAM-Th AOC-Th ACOM-Th R RÂ’ mp (C) R mp (C) R mp (C) CH3 C2H5 C3H7 C4H9 C6H13 CH3 CH3 CH3 CH3 CH3 CH3 C2H5 165-166 115-117 128-129 103 75-77 145-147 CH3 C2H5 C3H7 C4H9 C6H13 174-175 139-141 85.5-87.5 80.5-82 77-78.5 CH3 C2H5 C3H7 C4H9 C5H11 163.5-165 146-147 104-105 86-87 58-60 decreases after C2 member of the series. The initial in crease in solubility occurs because the promoiety masks a polar N-H group and th e short alkyl chains decrease crystal packing efficiency. As the alkyl chain grow s longer van der Waal s interaction between chains become stronger, leading to a decrea se in crystal lattice energy and decrease solubility (Yalkowsky, 1977). ACOM-6MP a nd 1-alkylcarbonyl 5FU prodrugs show similar behaviour (Waranis and Sloan, 1988; Beall et al., 1996 and Patrick et al., 1997). All NANAOCAM prodrugs of ThH were less water soluble than the parent drug. Water solubility increased on going from 21 to 22 and decreased thereafter from 23 to 25. The most lipid member of the series,22, was 2.5 fold less soluble than ThH. In comparison to AOC-Th or ACOM-Th, the solubilities in both IPM and water of NANAOCAM derivatives were lower.

PAGE 132

116 Regardless of the irregular behavior of the absolute solubilities, the ratios of the solubilities in IPM and AQ (SRIPM: AQ) (Table 4-3) were reas onably well behaved. The average methylene SR was 0.55 0.09. The SD for part ition coefficients determined between IPM and pH 4.0 buffer (KIPM: 4.0) were all less than 10%. The average methylene K was 0.5 0.08. The value obtained is in close agreement with with other series of prodrugs (Beall, 2001; Sloan, 2003 and Wasdo, 2004). Thus values obtained using solubility ratios are in close agreement with values obtained using partition coefficients between IPM and pH 4.0 buffer. This consistency makes values a robust indicator of homologous series behaviour. Diffusion Cell Experiments The maximum flux obtained for NANAO CAM prodrugs of theophylline are presented in Table 4-5. All JMIPM values were within the 30% variation in J values seen for in in vitro hairless mouse skin diffusion cell experiments. Only 22 (1.58) performed better than ThH at delivering total species from IPM. All other prodrugs delivered less parent drug + prodrug species through the skin than by ThH itself. All prodrugs derivatives had higher lipid solubility than the parent drug however it was the most water soluble derivative that gave highest flux through the skin. 22 was 0.4 fold as water soluble as ThH while being 64 times more lipid soluble than ThH and gave the highest deliver y of total species through the skin. 22 was the most lipid soluble and water soluble member of the series and hence gave the highest delivery through skin. 25 was the second most lipid sol uble member of the series (58) but its water solubility was lowest in the series (0.003) of ThH and it ended up giving the lowest delivery through skin (0.06) of ThH. In the 7-AOC-Th series (Sloan et al., 2000), the best performing prodrug was 133 fold fold more lipid soluble and 0.77 as water

PAGE 133

117 Table 4-3 Molecular Weights (MW), Melting Po ints (mp), Log Solubilities in Isopropyl Myristate (Log SIPM) Log Solubilities in Water (Log SAQ) and Estimated Log Solubilities in pH 4.0 Buffer (Log S4.0). Compound MW mp (C) log SIPM (mM) log SAQ (mM) log S4.0 (mM) ThH 179 270-274 -0.47 1.66 1.66 21, R = CH3, R’ = CH3 280 165-166 0.43 0.97 0.93 22, R = C2H5, R’ = CH3 294 115-117 1.34 1.25 1.38 23, R = C3H7, R’ = CH3 308 128-129 0.88 0.44 0.47 24, R = C4H9, R’ = CH3 322 103 1.04 -0.08 -0.02 25, R = C6H13, R’ = CH3 350 75-77 1.27 -0.88 -0.7 26, R = CH3, R’ = C2H5 294 145-147 0.95 0.99 1.17 soluble as ThH and gave 3 fold increase of flux across the skin. Unfortunately insertion of ‘N(R’)CH2’ moiety into the AOC moiety doesn’t increase lipid solubility by much and leads to a drastic reduc tion of water solubility in compar ison to ThH. As a result, increase in delivery of total ThH containing species due to the AOC moiety was twice compared to the NANAOCAM moiety. Similarly in the 7-ACOM-Th series, the best performing prodrug C3 was about 0.65 times as water soluble as ThH and 41 times more lipid soluble than ThH and improved flux by about 4 fold. The decreased biphasic solubility of the NANAOCAM-Th prodrugs leads to decr eased flux across hairless mouse skins compared to the 7-ACOM-Th prodrugs..

PAGE 134

118 Table 4-4 Molar Absortivities in Acetonitrile and Buffer ( ), Log Solubility Ratios between IPM and Water (log SRIPM:AQ ), the Differences Between Log SRIPM:AQ ( SR), the Log of Partition Coefficients Between IPM and pH 4.0 Buffer (Log KIPM:4.0), and the Differences Between Log KIPM:4.0( K). a Units of 1 104 L/mole. b Molar absortivities measured at 273 nm. c Buffer: pH 7.1 phosphate buffer with 0.11 % formaldehyde. d Molar absortivities measured at 260 nm.. e Molar absortivities measured at 284 nm. Prodrug Bioconversion to Parent Drug. NANAOCAM prodrugs of ThH are stable to chemical hydrolys is as seen before in Chapter 2. In comparison to chemical hydr olysis, NANAOCAM deri vatives of ThH are far more susceptible to enzymatic hydrolys is based on the amount of intact prodrug obtained after their passage through hairless mouse skins. The percentage of intact prodrug found in the receptor pha se was indicative of enzyma tic conversion to the parent drug. The percentages of intact prodrug C1 to C6 (21-26) in the receptor phase were 37, 26, 34, 28, 29 and 32% respectively. Irrespec tive of chain length the prodrug conversion for the members of the series remains constant. This particular trend was also seen in the NANAOCAM-APAP series, the am ount of intact prodrug seen for the ThH series were lower than for the APAP series which could be accounted by the lower pKa of ThH (8.6) Buffer a, c Compound CH3CNa b d e Log SRIPM:AQ SR Log KIPM:4.0 K ThH 0.8 0.62 0.51 -2.12 -2.14 21 0.79 0.45 0.63 -0.55 0.4-0.51 22 0.78 0.47 0.63 -0.11 0.5-0.04 0.47 23 0.83 0.49 0.65 0.44 0.60.41 0.45 24 0.82 0.44 0.6 1.12 0.51.07 0.61 25 0.79 0.47 0.55 2.16 1.97 0.45 26 0.84 0.53 0.64 -0.04 -0.13 0.38

PAGE 135

119 Table 4-5 Solubilities in IPM (SIPM), Solubilities in Water (SAQ) and Flux (JMIPM) through in vitro Hairless Mouse Skins from IPM. Compound SIPM (mM) SAQ (mM) JMIPM ( molcm2h-1) ThH 0.34 45.71 0.48 21, R = CH3, RÂ’ = CH3 2.69 9.33 0.16 22, R = C2H5, RÂ’ = CH3 21.88 17.78 0.76 23, R = C3H7, RÂ’ = CH3 7.59 2.75 0.09 24, R = C4H9, RÂ’ = CH3 10.96 0.83 0.07 25, R = C6H13, RÂ’ = CH3 18.62 0.13 0.03 26, R = CH3, RÂ’ = C2H5 8.91 9.77 0.34 versus APAP (9.5).The more acidic leaving gr oup is more sensitive to esterase catalysed hydrolysis (Barton et al., 1994) Conversion of prodrug to drug rarely has a substantial influence on the transdermal delivery through skin of total species from a prodrug because flux depends on the solubility of the drug in the skin (Bando et al., 1997 and Sloan, unpublished results). Carbamates are generally considered to be stable to enzymatic hydrolysis compared to simple esters or carbonates but there is precedence in literature that hydrolysis does occur. The marketed drug bambuterol is a carbamate derivative of phenol which undergoes enzymatic hydrolysis by non specific c holinesterases to yield the parent drug (Tunek et al., 1988). O O OH N H O N O N Bambuterol

PAGE 136

120 Permeability Coefficients and Solubility Parameter Values When flux from IPM (JIPM) was divided by their corres ponding solubility in IPM (SIPM) permeability coefficient (PMIPM) was obtained. The log PMIPM values of the prodrugs synthesized here are given in Table 4-6. Table 4-6 Log Permeability values for Theophylline Prodrugs from IPM through Hairless Mouse Skins (log PMIPM) and Solubility Parameter Values. Compound Log PMIPM i (cal cm-3)1/2 21 -1.23 13.49 22 -1.46 13.15 23 -1.93 12.85 24 -2.19 12.59 25 -2.79 12.36 26 -1.42 13.15 PMIPM values decreased along with their respec tive calculated sol ubility parameter A plot of of log PMIPM versus i values for compounds 21-26 gave a positive slope (Figure 4-5). Such dependence has been seen before for lipophilic drugs of polar heterocyclic drugs like 5-FU and phenolic drugs like APAP (Roberts and Sloan, 1999; Wasdo and Sloan, 2004). y = 1.3893x 19.802 R2 = 0.9504 -3 -2.5 -2 -1.5 -1 -0.5 0 1212.51313.514 Solubility ParameterLog PMIPM 21 22 26 23 24 25 Figure 4-5 Plot of solubil ity parameter versus log PMIPM for 21-26.

PAGE 137

121 For NANAOCAM-Th prodrugs, log PMIPM decreases linearly with increasing size of the promoiety. This decrease in PMIPM can be explained by the increase in SIPM and decrease of JMIPM with increasing al kyl chain length. 22 however had a higher JMIPM and SIPM than 21 but a lower PMIPM because the increase in JMIPM on going from 21 to 22 is lower than the increase in SIPM from 21 to 22 and hence the ratio JMIPM/SIPM is lower. Residual Amounts in Skin The residual skin concentrations (Crs) for NANAOCAM prodrugs of theophylline are presented in Table 4-7. 22 gave the highest Crs values. This deriva tive also gave the highest flux through the skin. 22 thus had the most solubility in the skin and gave higher transdermal and dermal flux. The best perf orming prodrug giving both higher transdermal and dermal delivery has been observed be fore in the AOC-APAP series (Wasdo and Sloan, 2004). The other prodrugs in the seri es deliver similar amounts of theophylline into the skin and no regular trend was seen. Table 4-7. Residual Skin Concentrat ions of Total Theophylline Species (Crs). Compound Crs SD ( mol) 21 1.358 0.42 22 2.064 0.41 23 1.19 0.26 24 1.204 0.14 25 1.3 0.15 26 2.31 0.23 Second Application Fluxes The second application theophylline flux (JJIPM) values are presented in Table 4-7.

PAGE 138

122 Table 4-8: Second Application Theophylline Flux (JJIPM) data for Flux of Theophylline from Propylene Glycol. Compound JJIPM SD ( molc m 2h-1 ) ThH 0.810.08 21 0.71 0.13 22 0.59 0.16 23 0.82 0.27 24 0.6 0.19 25 0.52 0.11 26 0.69 0.003 Skin penetration by theophylline from pr opylene glycol, was approximately the same or lower for skins treated w ith NANAOCAM prodrugs of theophylline. Normalization of the JMIPM values by the respective JJIPM values did not change the rank order of the performances of theophylline and its prodrugs 21-26. Thus, the differences in JJIPM must be due to differences in the abilities of prodrugs to deliver theophylline and not damage to the skin barrier. Modelling the Flux of NANAOCAM Prodru gs of Acetaminophen and Theophylline through Hairless Mouse Skin from IPM using the RS equation. Application of the Roberts-Sloan (RS) equation to analysis of flux obtained from NANAOCAM prodrugs data allows quantification of the effect of IPM solubility, water solubility and molecular weight on flux through hairless mice skin. First, the flux of the theophyllin e prodrugs was predicted using the RS equation derived from n = 63 compound database (equa tion 3-15; Wasdo and Sloan, 2005). Log JMIPM = -0.502 + 0.517 log SIPM + (1 0.517) log SAQ – 0.00266 MW (3-15) Inclusion of NANAOCAM-Th prodrugs (21-26) revised the coefficients of the RS model as shown in equation 4-13. Log JMIPM = -0.232 + 0.549 log SIPM + (1 0.549) log SAQ – 0.00389 MW (r2 = 0.91, n = 69) (4-13)

PAGE 139

123 Finally, all NANAOCAM prodrugs of acetaminophen and theophyl line were added to the n = 63 database from IPM to give the final form of the RS equation as shown by equation 4-14. Log JMIPM = -0.166 + 0.555 log SIPM + (1 0.555) log SAQ – 0.00422 MW (r2 = 0.90, n = 74) (4-14) NANAOCAM prodrugs of theophylline underperformed regardless of RS equation used in each case. The residual values (experimental log JMIPM – predicted log JMIPM = log JMIPM) for compounds 21-26 was 0.28 log units according to equation 3-15 and 0.16 log units using equation 4-13.The av erage error in calculating log JMIPM was 0.17 log units for the entire n = 69 database which is comparable to the log JMIPM of 0.16 log units obtained for the n = 63 database. A plot experimental vs calculated log JMIPM through hairless mouse skins is shown usi ng equation 4-13 in Figure 4-6. The residual values when all new prodrugs of APAP and ThH (NANAOCAM database) were fit to equation 4-14 for n = 74 database was found to be 0.2 log units and 0.13 log units respectively. The av erage error of predicting log JMIPM values for entire database (n = 74) is ~ 0. 16 log units using equation 414 A plot experimental vs calculated log JMIPM through hairless mouse skins for th e n = 74 database using equation 4-14 is shown in Figure 4-7 The behaviour of these prodrugs is similar to AOC-APAP prodrugs which also underperfor med based on their predicted log JMIPM of 0.27 log units. Both the models correctly identified the best performing members of the ThH series and rank order of the performance of prodrugs is also correctly identified.All the models correctly identified the best perf orming members of the APAP and ThH series

PAGE 140

124 and rank order of the performance of prodr ugs was also correctly identified.The experimental flux, calculated flux and error in predicting flux ( log JMIPM) using each model is given in Table 4-9. Table 4-9 Experimental Flux (log JMIPM), Calculated Flux (Calc. log JMIPM) and Error in Predicting Flux ( log JMIPM) for Compounds 1, 17-26 through Hairless Mouse Skins from IPM. Compound Log JMIPM Calc. log JMIPM Eq 3.15 log JMIPM Eq 3.15 Calc. log JMIPM Eq 4.13 log JMIPM Eq 4.13 Calc. log JMIPM Eq 4.14 log JMIPM Eq 4.14 1 0.05 0.15 0.10 17 -0.21 -0.033 0.17 18 -0.29 -0.031 0.26 19 -0.37 -0.14 0.22 20 -0.66 -0.42 0.24 21 -0.79 -0.56 0.24 -0.65 0.15 -0.68 0.119 22 -0.12 0.012 0.13 -0.076 0.0042 -0.11 0.0124 23 -1.05 -0.66 0.39 -0.75 0.297 -0.78 0.26 24 -1.15 -0.86 0.29 -0.95 0.205 -0.98 0.17 25 -1.52 -1.20 0.32 -1.29 0.23 -1.33 0.19 26 -0.46 -0.31 0.15 -0.41 0.061 -0.44 0.00297 -2.5 -1.5 -0.5 0.5 1.5 2.5 3.5 -2.5-0.51.53.5 Calculated log JMIPMlog JMIPM = 0.232+0.549 log SIPM+ (1-0.549) log SAQ 0.00389 MWExperimental log JMIPM n = 63 database NANAOCAM prodrugs of ThH Figure 4-6 Experimental versus calculated log maximum flux values through hairless mouse skin from IPM using equation 4.13 for n = 69.

PAGE 141

125 -2.5 -1.5 -0.5 0.5 1.5 2.5 3.5 -2.5-0.51.53.5Calculated log JMIPMlog JMIPM = -0.502 +0.517 log SIPM + (1-0.517) log SAQ 0.00266 MWExperimental log JMIPM n = 63 database NANAOCAM prodrugs of APAP and ThH Figure 4-7 Experimental versus calculated log maximum flux values through hairless mouse skin from IPM using equation 4.14 for n = 74. Conclusions For the 7-NANAOCAM-Th series of prodrugs the most lipid soluble and the most water soluble member of the series was the most effective in delivering total ThH species from IPM through hairless mouse skins. The 7-NANAOCAM-Th prodrugs were generally less effective than both 7-AOC a nd 7-ACOM prodrugs of ThH because they were generally less soluble in lipid (IPM) and water. However, they were more stable than their AOC counterparts which should make formulation more feasible. The addition of the 7-NANAOCAM-Th series to the RS database and subsequent fitting of n = 69 to the RS equation gave new coefficients that were not different from the previous parameter estimates. The flux of NANAOCAM prodrugs is accurately predicted by the RS equation which reflects the fact that solubility in lipid and water are important criteria for optimizing flux through the skin.

PAGE 142

126 CHAPTER 5 SUMMARY OF RESULTS OBTAIN ED AND FUTURE WORK There were four major objectives to this work. The first objective was the synthesis and characterization of N-alkyl-N-al kyloxycarbonylaminomethyl (NANAOCAM) derivatives of phenols, carboxylic acids, th eophylline, 6-mercaptopurine. The second was; synthesis and charac terization of N-aryl-N-a lkyloxycarbonylaminomethyl (NArNAOCAM) derivatives of phenols and carb oxylic acids followed by investigation of the mechanism of chemical hydrolysis of their NANAOCAM and NArNAOCAM derivatives in aqueous buffers.The third object ive was to evaluate the delivery of total APAP and ThH by their NANAOCAM derivativ es across hairless mouse skins from IPM. The fourth objective was to combine this new data with previously synthesized prodrugs series of 5-FU and 6MP and fit th e combined database to the Roberts Sloan equation (RS) to predict flux. All the objectiv es of the study were completed. An optimized synthesis of NANAOCAM de rivatives of phenols, carboxylic acids, theophylline and 6-mercaptopurine has been deve loped. The method gave better yields of the desired product and isolation of pure product was less cumbersome than reported before (Siver and Sloan, 1990). A new synt hetic route was deve loped to synthesize NArNAOCAM derivatives of phenols and carbo xylic acids in moderate yields. The new compounds synthesized were completely ch aracterized by TLC, UV, NMR and elemental analysis. The mechanism of hydrolysis of NANAOCAM and NArNAOCAM drug derivatives was found to be SN1 type. The hydrolysis was dependent on pKa of the

PAGE 143

127 leaving group and independent of pH of buffe r. Carboxylic acid prodrugs had higher rates of hydrolysis than phenols. NANAOCAM-carboxy lic acids were too uns table to serve as useful prodrugs and were stabalised by making their analogous NArNAOCAM promoiety. Both NANAOCAM and NArNAOCAM derivatives served as chemically labile derivatives of carboxylic acids so they do not rely on enzymatic processes to release the active drug molecule. A mo re water soluble NArNAOCAM prodrug of carboxylic acid would be predicted to increa se membrane permeability if given orally. NANAOCAM-phenol derivatives are chemica lly stable but enzymatically labile derivatives. The ThH prodrugs were more la bile enzymatically than APAP prodrugs. Both NANAOCAM and NArNAOCAM promoieties if conjugated to phenolic drugs like morphine and estradiol shoul d limit their first pass metabolism and increase oral bioavailability. The physicochemical properties of NANAO CAM prodrugs of AP AP and ThH such as melting points, solubilities in IPM and wa ter, partition coeffici ents between IPM and pH 4.0 acetate buffers were determined. NANAOCAM derivatives of both acetaminophen and theophylline increased flux across the skin by 2.1 and 1.6 fold respectively. The derivative gi ving the highest flux was more lip id soluble than the parent drug and exhibited the highest water solubility in the series. The addition of APAP and ThH prodrugs to the IPM database reinforced the conclusion reached previously. The relative importance of biphasic solubility on fl ux is exemplified by the coefficients of y in RS equation. The flux of both APAP and ThH prodrugs was accurately predicted by the RS equation. A slightly increased dependence on lipid solubility and less dependence on water solubility was seen compared to older RS parameter estimates. The error in

PAGE 144

128 predicting log J values and st andard deviations were low. All parameter estimates were statistically significant and had p-values less than 0.05. As the sample size increases, the precision of parameter estimates will increase and the estimations will get closer to the true parameter valu e of the population.

PAGE 145

129 LIST OF REFERENCES Ahmad, N.; Agarwal, R.; and Mukhtar, H. Cytochrome P-450 and Drug Development for Skin Diseases. Skin. Pharmacol. 1996, 9(4), 231-241. Ahmad, N.; and Mukhtar, H. Cytochro me P-450 and Drug Development for Skin Diseases. J. Invest. Dermatol. 2004, 123(3), 417-25. Albert, A. Chemical Aspects of Selective Toxicity. Nature 1958, 182, 421-422. Anderson, D.C.; Nichols, E.; Manger, R.; W oodle, D.; Barry, M.; and Fretzberg, A.R. Tumor Cell Retention of Antibody Fab Fragments Is Enhanced by an Attached HIV TAT Protein-Derived Peptide. Biochem. Biophys. Res. Commun. 1993, 194, 876-884. Bando, H.; Mohri, S.; Yamashita, F.; Takakura, Y.; and Hashida, M. Effects of Skin Metabolism on Percutaneous Pene tration of Lipophilic Drugs. J. Pharm. Sci. 1997, 86(6), 759-761. Barry, B. W. Penetration Enhancer s: Mode of action in Human Skin. Pharmacology and the Skin, Vol 1 (B Shroot, H Schaefer, eds.). Karger, Basel (1987) pp. 121-137. Barton, P.; Laws, A.P.; and Page, M.I. Struct ure-Activity Relationships in the Esterase Catalysed Hydrolysis and Transester ification of Esters and Lactones. J. Chem. Soc. Perkin. Trans.2 1994, 2021-2029. Beall, H.; Getz, J.J.; and Sloan, K.B. The Estimation of Water Solubility for Prodrugs unstable in Water. Int. J. Pharm 1993, 93, 37-47. Beall, H.; Prankerd, R.; and Sloan, K.B. Transdermal delivery of 5-Fluorouracil (5-FU) through Hairless Mice Skin by 1-al kyloxycarbonyl-5-FU Prodrugs: Physicochemical Characterization of Prodr ugs and Correlations with Transdermal Delivery. Int. J. Pharm. 1994, 111, 223-233. Beall, H.D.; Prankerd, R.; and Sloan, K.B. Transdermal delivery of 5-Fluorouracil (5-FU) by 1-Alkylcarbonyl-5-FU prodrugs. Int. J. Pharm. 1996, 129, 203-210. Beall, H.D.; and Sloan, K.B. Topical Delivery of 5-Fluorouracil (5-FU) by 3Alkylcarbonyl-5-FU prodrugs. Int. J. Pharm. 2001, 217, 127-137.

PAGE 146

130 Beamount, K.; Webster, R.; Gardner, I.; a nd Dack, K. Design of Ester Prodrugs to Enhance Oral Absorption of Poorly Permeable Compounds: Challenges to the Discovery Scientist. Current Drug Metabolism, 2003, 4, 461-485. Berenbein, B.A.; Mikhailov, G.S.; and Shumay N.T. Immediate results of Psoriasis treatment with Phosphodiesterase Inhibitors. Vestn. Dermatol. Venerol. 1979, 5, 21-25. Berti, J. J.; and Lipsky, J. L. Transcut aneous Drug Delivery: A practical Review. Mayo Clin. Proc. 1995, 70, 581-586. Bisset, D. L. Anatomy and Biochemistry of the skin. Transdermal Delivery of Drugs, Vol. I (A. F. Kydonieus and B. Berner eds.). Marcel Dekker, New York (1992) pp. 179-204. Block, L.H. Medicated Topicals. Remington, The Science and Practice of Pharmacy, 20th edition. (Gennaro, A.R. ed.). Lippincott Williams & Wilkins, Philadelphia (2000) pp 836-857. Bonina, F.P.; Montenegro, L.; De Caprariis, P. ; Palagiano, F.; Trapani, G.; and Liso, G. In vitro and in vivo evaluation of Pol yoxyethyleneIndomethacin Esters as Dermal Prodrugs. J. Controlled. Release. 1995, 34, 223-232. Bonina, F.P.; Puglia, C.; Barbuzzi, T.; De Capr ariis, P.; Palagiano, F.; Rimoli, M.G.; and Saija, A. In vitro and in vivo evaluati on of Polyoxyethylene Esters as Dermal Prodrugs of Ketoprofen, Naproxen and Diclofenac. Eur. J. Pharm. Sci. 2001, 14, 123-134. Bourne. H.R.; Lichtenstein, L.M.; Melmon, K.L.; Henney, C.S.; Weinstein, Y.; and Shearer, G.M. Modulation of Inflammation and Immunity by cyclic AMP. Science 1974, 184, 19-28. Bundgaard, H.; and Johansen, M. Prodrugs as Drug Delivery systems VIII. Bioreversible Derivatization of Hydant oins by N-hydroxylation. Int. J. Pharm. 1980, 5, 67-77. Bundgaard, H.; Klixbull, U.; and Falch, E. Prodrugs as Drug Delivery Systems. 43. OAcyloxymethyl salicylamide N-Mannich bases as double prodrug forms for Amines. Int. J. Pharm. 1986, 29, 19-28. Bundgaard, H. Design and Application of Prodrugs. In A Textbook of Drug Design and Development; Krogsgaard-Larsen, P., Bundgaard, H., Eds.; Harwood: Reading, U.K., 1991; pp113-91. Bundgaard, H.; and Rasmussen, G.J. Prodrugs of Peptides 11. Chemical and Enzymatic Hydrolysis Kinetics of N-Acyloxymethyl Derivatives of Peptide-like bond Pharm. Res. 1991, 8, 1238-1242.

PAGE 147

131 Buur, A.; Bundgaard, H.; and Falch, E. Pr odrugs of 5-fluorouracil. IV. Hydrolysis Kinetics, Bioactivation and Physic ochemical Properties of various NAcyloxymethyl derivatives of 5-Fluorouracil. Int. J. Pharm. 1985, 24, 43-60. Buur, A.; and Bundgaard, H. Prodrugs of 5-Fluorouracil: V. 1-Alkyloxycarbonyl Derivatives as potential Prodrug forms for improved Rectal or Oral delivery of 5Fluorouracil. J. Pharm. Sci. 1986, 75, 522-527. Chikos, J.S.; and Nichols, G. Simple Re lationships for the Estimation of Melting Temperatures of Homologous Series. J. Chem. Eng. Data. 2001, 46, 562-573. Clement, B. Reduction of N-Hydr oxylated compounds: Amidoximes (NHydroxyamidines) as Prodrugs of Amidines. Drug Metab. Rev. 2002, 34, 565-579. Clement, B.; and Lopian, K. Ximelagatr an, an Amidoximean Ester-Prodrug. Drug. Metab. Disp. 2003, 31, 645-651. Cohen, H.; and Turnbull, D. Molecular transport in Liquids and Gases, J. Chem. Phys. 1959, 31, 1164-1169. Conover, C.D. Campothecan Delivery Systems: The utility of Aminoacids Spacers for the Conjugation of Campothecan with Poly ethylene Glycol to create Prodrugs. Anti-Cancer Drug Design 1999, 14, 459-506. De Mario, M.D.; and Ratain, M.J. Oral Chem otherapy: Rationale and Future Directions. J. Clin. Oncol. 1998, 16, 2557–2567. Dillaha, C. J.; Jansen, G.T.; Honeycutt, W.M.; and Holt, C.A. Further Studies with Topical 5-Fluorouracil. Arch. Dermatol. 1965, 92, 410-417. Drustrup, J.; Fullerton, L.; Christrup, L.; and Bundgaard, H. Utilization of Prodrugs to Enhance the Transdermal Absorption of Morphine. Int. J. Pharm. 1991, 71, 105116. Ettmayer, P.; Amidon, G.; Clement, B.; and Testa, B. Lessons Learned from Marketed and Investigational Prodrugs. J. Med. Chem. 2004, 47, 2393-2404. Fedors, R. F.A Method of Estimating both the Solubility Parameters and Molar Volumes of Liquids. Polym. Eng. Sci. 1974, 14, 147-154. Fick, A. Ann. Physik. Chem 1855, 94, 59. Fleisher, D.; Bong, R.; and Stewart, B. H. Improved Oral Drug Delivery: Solubility Limitations overcome by the use of Prod rugs. Advanced Drug Delivery Reviews, 1996, 19, 115-130.

PAGE 148

132 Flynn, G. L. Physicochemical Determinates of Skin Absorption. In J.R.Gerrity and C. J. Henry (eds.), Principles of Route-to Route Ex trapolation for Risk assessment. Elsevier, Amsterdam, 1990, pp 93-127. Frosch, P.J.; and Kligman, A.M. “The Cham ber-Scarification test for assessing and the Irritancy of Topically Applied Substances in Cutaneous Toxicity, eds., V.A. Drill and R. Lazer, Academic Press, New York, pp. 127-154 (1977). Fuchs, S.M.; and Raines, R.T. Pathway for Polyarginine entry into Mammalian Cell. Biochemistry 2004, 43, 2438-2444. Getz, J.J.; Prankerd, R.; and Sloan, K.B. Mechanism of Hydrolysis of O-Imidomethyl Derivatives of Phenols. J.Org.Chem. 1993, 58, 4913-4918 Graymore, J. Reduction products of cer tain Cyclic Methyleneamines II J. Chem. Soc., 1932, 1353-1357. Green, M.; and Loevenstein, P.M. Autonom ous functional domain s of Chemically Synthesized Human Imm unodeficiency Virus Tat Trans-Activator Protein. Cell. 1988, 55, 1179-1188. Greenwald, R.B.; Gilbert, C.; Pendri, A.; C onover, C.; Yang, R.; Xia, J.; and Martinez, A. Drug Delivery Systems: Water Soluble Taxol 2’-Poly (ethylene glycol) Ester Prodrugs-Design and in Vivo Effectiveness. J. Med. Chem. 1996, 39, 424-431. Greenwald, R.B.; Pendri, A.; Conover, C.; G ilbert, C.; Yang, R.; and Xia, J. Drug Delivery Systems. 2. Campothecan 20-O-Po ly (ethylene glycol) Ester Transport Forms. J. Med. Chem. 1996, 39, 1938-1940. Greenwald, R.B.; Pendri, A.; Conover, C.; Zh ao, H.; Choe, Y.H.; Martinez, A.; Shum, K.; and Guan, S. Drug Delivery Systems Employing 1,4or 1,6-Elimination: Poly (ethylene glycol) Prodrugs of Amine-Containing Compounds. J. Med. Chem. 1999, 42, 3657-3667. Greenwald, R.B.; Choe, Y.H.; Conover, C.; Shum, K.; Wu, D.; and Royzen, M. Drug Delivery Systems Based on Trimethyl Lock Lactonization: Poly (ethylene glycol) Prodrugs of Amine-Containing Compounds. J. Med. Chem. 2000, 43, 475-487. Greenwald, R.B.; PEG drugs: An Overview. J. Controlled. Release. 2001, 159-171. Greenwald, R.B.; Choe, Y. H.; McGuire, J.; and Conover, C. D. Effective Drug Delivery by PEGylated Drug Conjugates. Adv. Drug. Del. Rev. 2003, 55, 217-250. Gustafsson, D.; Nystrom, J.E.; Carlsson, S.; Bredberg, U.; Eriksson, U.; Gyzander, E.; Elg, M.; Antonsson, T.; Hoffmann, K.J.; Ungell, A.L.; Sorenson, H.; Nagard, S.; Abrahamsson, A.; and Bylund, R. The Dire ct Thrombin Inhibitor Melagatran and its Oral Prodrug H 376/95: Intestinal Absorption Properties, Biochemical and Pharmacodyanamic Effects. Thrombosis Research. 2001, 101, 171-181.

PAGE 149

133 Guy, R.H.; and Hadgraft, J. Physicochemical Aspects of Percutaneous Penetration and its Enhancement. Pharm Res. 1988, 5, 753-758. Hammell, D.C.; Stolarczyk, E.I.; Klausner, M.; Valiveti, S.; Hamad, M.O.; Crooks, P.A.; and Stinchcomb, A.L. Bioconversion of Naltrexone and its 3-O-Alkyl-Ester Prodrugs in a Human Skin Equivalent. J.Pharm. Sci. 2005, 94(4), 828-836. Hansch, C.; and Leo, A. Substituent Constants for Correlation analysis in Chemistry and Biology. Wiley & sons, 1979. Henry, S.; McAllistair, D.V.; Allen, M.G. ; and Prausnitz, M.R. Microfabricated Microneedles: A Novel Approach to Transdermal Drug Delivery. J.Pharm.Sci. 1998, 87, 922-926. Higuchi, T. Physical Chemical Analysis of Percutaneous Absorption Process from Creams and Ointments. J. Soc. Cosmet. Chem. 1960, 11, 85-97. Higuchi, T. Design of Chemical Stru cture for Optimal Dermal Delivery. Curr. Probl. Dermatol, vol 7 (G. A. Simon et al .Â’ ed.). Karger, Basel (1978) pp. 121-131. Hou, S.Y.E., Mitra, A.K., White, S.H., Menon, G.K., Ghadially, R.; and Elias, P.M. Membrane Structures in Normal and Essential Fatty Acid-Deficient Stratum Corneum: Characterization by Ruthen ium Tetroxide Staining and X-Ray Diffraction J. Invest. Dermatol. 1991, 96, 215-223. Iancu, L.; Shneur, A.; and Cohen, H. Trials with Xanthine derivatives in Systemic treatment of Psoriasis. Dermatologica 1979, 159, 55-61. Iley, J.; Moreira, R.; Calheiros, T.; and Mendes, E.Acyloxymethyl as a Drug Protecting Group: Part 4. The Hydrolysis of Te rtiary Amidomethyl Ester Prodrugs of Carboxylic acid agents. Pharm. Res. 1997, 14, 1634-1639. Iley, J.; Calheiros, T; and Moreira, R. Phthalimidomethyl as a Drug Pro-moiety. Probing its Reactivity. Bioorg. Med. Chem. Lett. 1998, 8, 955-958. Jain, R.; Majumdar, S.; Nashed, Y.; Pal, D.; and Mitra, A. Circumventing PGlycoprotein-Mediated Cellu lar Efflux of Quinidine by Prodrug Derivatization. Mol.Pharm. 2004, 1(4), 290-299. Jansen, A.B.A.; and Russell, T.J. Some Novel Penicillin Derivatives. J. Chem. Soc. 1965, 2127-2132. Johansen, M.; and Bundgard, H. Prodrugs as drug delivery systems.VI. Kinetics and Mechanism of the Decomposition of N-Hydr oxy methylated Amides and Imides in Aqueous Solution and Assessment of thei r Suitability as Possible Prodrugs. Arch. Pharm. Chem., Sci. Ed. 1979, 7, 175-192.

PAGE 150

134 Jona, J.A.; Dittert, L.W.; Crooks, P.A.; Milo sovich, S.M.; and Hussain, A.A. Design of Novel Prodrugs for the Enhancement of the Transdermal Penetration of Indomethacin. Int. J. Pharm. 1995, 123, 127-136. Kadir, R. Delivery of Theophylline into Excised Human Skin from Alkanoic acid Solutions: a Push-Pull mechanism. J. Pharm. Sci., 1987, 76, 774. Kerr, D.; Roberts, W.J.; Tebbett, I.; and Sl oan, K.B. 7-Alkylcar bonyloxymethyl Prodrugs of Theophylline: Topical Delivery of Theophylline Int. J. Pharm. 1998, 167, 37-48. Kim, D.T.; Introduction of Soluble Proteins into MHC class I pathway by Conjugation to an HIV-Tat Peptide. J. Immunol. 1997,159, 1666-1668. Kirschberg, T. A.; VanDeusen, C. L.; Rothbard, J. B.; Yang, M.; and Wender, P. A. Arginine-Based Molecular Transporters: The Synthesis and Chemical Evaluation of Releasable Taxol-Transporter Conjugates Org. Lett. 2003, 5, 3459-3462. Koch, S.A.M. Prodrug Approaches to Increas e the Topical Delivery of 5-Fluorocytosine and 5-Iodo-2Â’Deoxycytidine. PhD dissertation, University of Florida, Gainesville, FL (1986). Krise, J.P.; Zygmunt, J.; Georg, G.I.; a nd Stella, V.J. Novel Prodrug Approach for Tertiary Amines: Synthesis and Prelim inary Evaluation of N-Phosphonooxymethyl Prodrugs. J.Med.Chem. 1999, 42, 3094-3100. Labaree, D. C.; Reynolds, T. Y.; and Hochbe rg, R. B. Estradiol-16-Carboxylic Acid Esters As Locally Active Estrogens. J. Med. Chem. 2001, 44, 1802-1814. Labaree, D. C.; Zhang, J.; Harris, H.; OÂ’ Connor, C.; Reynolds, T. Y.; and Hochberg, R. B. Synthesis and Evaluation of B-, C-, and D-Ring-Substituted Estradiol Carboxylic Acid Esters As Locally Active Estrogens. J. Med. Chem. 2003, 46, 1886-1904. Laupeze, B.; Amiot, L.; Bertho, N.; Grosset, J.M.; Lehne, G.; Fauchet, R.; and Fardel, O. Differential Expression of the Efflux Pumps P-glycoprotein and Multidrug Resistance-Associated Protein in Hu man Monocyte-Derived Dendritic Cells.Hum. Immunol. 2001, 62(10), 1073-80. Leo, A.; Hansch, C.; and Elkins, D. Pa rtition Coefficients and Their Uses. Chem. Rev. 1971, 71, 525-616. Leweke, U.H.; and Lippold, B.C. Abso rption of Sunscreens and Other Compounds Through Human Skin in Vivo: Derivation of a method to Predict Maximum Fluxes. Pharm. Res. 1995, 12, 1354-1360. Lindgren, M.; Hallbrink, M.; Prochiantz, A. and Langel, U.; Cell-Penetrating Peptides. Trends Pharmacol. Sci. 2000, 21, 99-103.

PAGE 151

135 Luedtke, N.W.; Carmichael, P.; and Tor, Y. Cellular Uptake of Aminoglycosides, Guanidinoglycosides and Polyarginine. J. Am. Chem. Soc. 2003, 125, 1237412375. Maeda, H.; Seymour, L.; and Miyamoto, Y. Conjugates of Anti-cancer Agents and Polymers: Advantages of Macromolecular Therapeutics in Vivo. Bioconjugate Chem. 1992, 3, 351-362. Majumdar, S.; Thomas, J.D.; Wasdo, S.C.; and Sloan, K.B. The Effect of Water Solubility of Solutes on Their Flux th rough Human Skin: The Extended and Edited Flynn Database. The AAPS Journal 2004, 6(4), Abstract W4250. Majumdar, S.; and Sloan, K.B. Predicti on of the Flux of Sunscreens and Other Compounds from PG/water through Human s Skin in vivo using the Roberts Sloan equation. The AAPS Journal 2005, 7(S2), Abstract R6075. Mann, D.A.; and Frankel, A.D. Endocyto sis and Targeting of exogenous HIV-1 Tat protein. EMBO J. 1991, 10, 1733-1739. Martin, A.; Wu, P. L.; and Valesquez, T. Extended Hildebrand Solubility Approach: Sulfonamides in Binary and Ternary Solvents. J. Pharm. Sci. 1985, 74, 277-282. Martin, A.; Wu, P. L.; Liron, Z.; and Cohen, S. Dependence of Solute Solubility Parameters on Solvent Polarity. J. Pharm. Sci. 1985a, 74, 638-642. McKenzie, A.W.; and Atkinson, R.M. Topical Activities of Betamethasone Esters in Man. Arch. Dermatol. 1964, 89, 741-746. Michaels, A.S.; Chandrasekaran, S.K.; and Shaw, J.E. Drug Permeation through Human Skin: Theory and in vitro Experimental Measurement. Am. Inst. Chem. Eng. 1975, 21, 985-996. Milosovich, S.; Hussain, A.; Dittert, L.; A ungst, B.; and Hussain, M. Testosteronyl-4dimethylaminobutyrate HCl: A Prodrug w ith improved Skin Permeation Rate. J. Pharm. Sci. 1993, 82, 227-228. Mitchell, D.J.; Kim, D.T.; Steinman, L.; Fathman, C.G.; and Rothbarg, J.B. Polyarginine enters cell more efficiently than other Polycationic Homopolymers. J. Pep. Res. 2000, 56, 318-325. Mitragotri, S.; Blankschtein, D.; and Lange r, R. Transdermal Drug Delivery using Low Frequency Sonophoresis. Pharm. Res. 1996, 13, 411-420. Mitragotri, S.; and Kost, J. Low Frequency Sonophoresis. A review Adv.Drug.Deliv .Rev. 2004, 56(5), 589-601.

PAGE 152

136 Moreira, R.; Mendes, E.; Calheiros, T.; Bacelo, M.J.; and Iley, J. A New Direct Synthesis of Tertiary N-Acyloxymethylamide Pr odrugs of Carboxylic Acid Drugs. Tetrahedron Lett. 1994, 35, 7107 Moreira, R.; Calheiros, T.; Cabrita, J. ; Mendes, E.; Pimentel, M.; and Iley, J. Acyloxymethyl as Drug Protecting Group. Part 3. Tertiary O-Amidomethyl Esters of Penicillin G: Chemical hydrol ysis and Antibacterial Activity. Pharm. Res. 1996, 13, 70-75. Nagahara, H. Transduction of full-length Ta t Fusion proteins into Mammalian cells: Tatp27kip/ induces cell migration. Nature Med 1998, 4, 1449-1452. Pannatier, A.; Testa, B.; and Etter, J. Enzymatic Hydrolysis by Mouse Skin Homogenates: Structure-Metabolism Relatio nships of Para-Nitrobenzoate Esters. Int. J. Pharm. 1981, 8, 167-171. Patel, J.; Prankerd, R.; and Sloan, K.B. A Prodrug Approach to Increasing the Oral Potency of a Phenolic Drug. Part 1. Synthe sis, Characterization and Stability of an O-Imidomethyl Derivati ve of 17-Estradiol. J. Pharm. Sci. 1994, 83,1477-1481. Patel, J.; Katovich, M.J.; Sloan, K.B.; Curr y, S.H.; and Prankerd, R. A Prodrug Approach to Increasing the Oral Potenc y of a Phenolic Drug. Part 2. Pharmacodyanamics and Preliminary Bioavailability of an Orally Administered O-Imidomethyl Derivative of 17-Estradiol. J. Pharm. Sci. 1995, 84,174-178. Patrick, A.I.; Beall, H.D.; Gilroy, P.; and Sloan, K.B. Effect of Vehicles on Topical Delivery of 5-Fluorouracil (5-F U) by 1-Acyl-5-FU prodrugs. Int. J. Pharm. 1997, 154, 39-48. Paudel, K.S.; Nalluri, B.N.; Hammell, D.C. ; Valiveti, S.; Kiptoo, P.; Hamad, M.O.; Crooks, P.A.; and Stinchcomb, A.L. Tran sdermal delivery of Naltrexone and its Active Metabolite 6-b-Naltrexol in Human Sk in in vitro and Guinea Pigs in vivo. J.Pharm. Sci. 2005, In press. Perrin, D.; and Dempsey, B. Buffers for pH and Metal ion control; J. Wiley and Sons, New York, 1974. Pillai, O.; Hamad, M.O.; Crooks, P.A.; and Stinchcomb, A.L. Physicochemical Evaluation, In Vitro Human Skin Diffusion and Concurrent Biot ransformation of 3O-Alkyl Carbonate Prodrugs of Naltrexone. Pharm. Res. 2004, 21(7), 1146-1152. Ponec, M.; Kempencar, J.; Shroot, B.; and Caro n, J. C. Glucocortico ids: Binding Affinity and Lipophilicity. J. Pharm. Sci. 1986, 75, 973-975. Poulsen, B. J., Design of topica l drug products: Biopharmaceutics, in Drug Design : Medicinal Chemistry, ed., E. J. Ariens, Academic Press, New York, Vol. 4, pp. 149-192 (1971).

PAGE 153

137 Potts, R.O.; and Guy, R.H. Predicting Skin Permeability Pharm Res. 1992, 9, 663-669. Prausnitz, M.; Mitragotri, S.; and Langer, R. Current Status and Future Potential of Transdermal Drug Delivery. Nat. Rev. Drug Discovery 2004, 3, 115-124. Riviere, J.E.; and Heit, M.C. Electrica lly Assited Transdermal Drug Delivery. Pharm. Res. 1997, 14(6), 687-697. Rautio, J.; Nevalainen, T.; Taipale, N.; Vepsalainen, J.; Gynther, J.; Pederson.; and Jarvinen, J. Synthesis and in vitro evaluation of aminoacy loxyalkyl esters of 2-(6methoxy-2-naphthyl)propionic acid as nove l naproxen prodrugs for dermal drug delivery. Pharm. Res. 1999, 16, 1172-1178. Roberts, W.J.; and Sloan, K.B. Correlation of Aqueous and Lipid Solubilities with Flux for Prodrugs of 5-Fluorouracil, Theophylline and 6-Mercaptopurine: a Potts-Guy approach. J. Pharm. Sci. 1999, 88, 515-532. Roberts, W.J.; and Sloan, K.B. Application of the Transformed Potts-Guy Equation to In Vivo Human Skin Data J. Pharm. Sci. 2001, 90, 1318-1323. Rothbard J. B.; Garlington, S.; Lin, Q.; Ki rschberg, T.; Kreider, E.; McGrane, P. L.; Wender, P. A.; and Khavari. P. A. Conjugation of Arginine Oligomers to Cyclosporin A Facilitates Topical De livery and inhibition of inflammation Nature Medicine 2000, 6, 1253-1257. Rothbard, J.B.; Kreider, E.; VanDeusen, C. L.; Wright, L.; Wylie, B.L.; and Wender, P. A. Arginine-Rich Molecular Transporters for Drug Delivery: Role of Backbone Spacing in Cellular Uptake J. Med. Chem. 2002, 3612-3618 Rothbard, J. B.; Jessop, T. C.; and Wender, P. A. Adaptive Translocation: The Role of Hydrogen Bonding and Membrane Potential in the Uptake of Guanidinium-rich Transporters into Cells Adv. Drug Delivery Rev. 2005, 57, 495-504. Saab, A.N.; and Sloan, K.B. The Eff ect of Aminomethyl (N-Mannich base) Derivatization on the Ability of S6-Pivaloyloxymethyl-6-M ercaptopurine Prodrug to deliver 6-Mercaptopurine through Hairless Mouse Skin. Int. J. Pharm. 1989, 57, 253-261. Saab, A.N.; Sloan, K.B.; Beall, H.; and Villanueva, R. Effect of Aminomethyl (NMannich base) Derivatiza tion on the Ability of S6-Acetyloxymethyl-6Mercaptopurine prodrug to deliver 6-Merc aptopurine through Hairless Mouse Skin. J. Pharm. Sci. 1990, 79, 1099-1104. Safadi, M.; Oliyai, R.; and Stella, V. J. Phosphoryloxymethyl Carbamates and Carbonates-Novel Water-Soluble Prodrugs for Amines and Hindered Alcohols. Pharm.Res. 1993, 10, 1350-1355.

PAGE 154

138 Schaefer, H.; and Redelmeier, T. E. Skin Barrier : Principles of percutaneous absorption. New York: S. Karger AG Publishers; 1996 Scheuplein, R. J.; and Blank, I. H. Permeability of the Skin. Physiological Rev. 1971, 51, 702-747. Schinkel, A.H. The Physiological functi on of Drug-transporting P-glycoproteins. Sem. Cancer Biol. 1997, 8, 161–170. Seki, H.; Kawaguchi, T.; and Higuchi, T. Sp ecificity of Esterases and Structure of Prodrug Esters: Reactivity of various Acylated Acetaminophen compounds and Acetylaminobenzoated compounds. J. Pharm. Sci. 1988, 77,855. Sherertz, E. F.; Sloan, K. B.; and McTier nan, R.G. Use of Theoritical Partition Coefficients determined from Solubility Parameters to predict Permeability Coefficients for 5-Fluorouracil. J. Invest. Dermatol. 1987, 89, 147-151. Shoemaker, D.P.; Garland, C.W.; and Stinfeld, J.I. In Experiments in Physical Chemistry: R.H.Summersgill, J.Stern, M. Eichberg, Eds., Mcgraw-Hill, New York, 1974, pg 283-286. Shindo, Y.; Witt, E.; Han, D.; Epstein, W.; and Packer, L. Enzymatic and Non-Enzymatic Antioxidants in Epidermis and Dermis of Human Skin. J. Invest. Dermatol. 1994, 102, 122-124. Siver, K.G. Prodrug Approaches to Increase the Topical Delivery of 6-Mercaptopurine PhD dissertation, University of Florida, Gainesville, FL (1987). Siver, K.G.; and Sloan, K.B. The Effect of Structure of Mannich base Prodrugs of 6Mercaptopurine on their Ability to De liver 6-Mercaptopurine through Hairless Mice Skin. Int. J. Pharm. 1988, 48, 195-206. Siver, K.; and Sloan, K.B. Alkylation of 6-Mercaptopurine (6MP) with N-Alkyl-NAlkyloxycarbonylaminomethyl Chlorides: S6-(N-Alkyl-N-Alkyloxycarbonylaminomethyl-6MP Prodrug Structure Eff ect on the Dermal Delivery of 6MP. J. Pharm. Sci. 1990, 79, 66-73. Sloan, K.B. Prodrugs for Dermal Delivery. Adv. Drug. Del. Rev. 1989, 3, 67-101. Sloan, K. B. Functional Group Consider ations in the Development of Prodrug Approaches to Solving Topical Delivery Problems. Topical and Ocular Drug Delivery (K. B. Sloan, ed.). Me rkel Dekker, New York (1992) pp. 17-116. Sloan, K. B. Use of Solubility Parameters from Regular Theory to Describe PartitioningDriven Processes. Topical and Ocular Drug Delivery (K. B. Sloan, ed.). Merkel Dekker, New York (1992) pp. 179-204.

PAGE 155

139 Sloan, K.B.; and Bodor, N. Hydroxym ethyl and Acyloxymethyl Prodrugs of Theophylline: Enhanced Delivery of Polar Drugs through Skin. Int. J. Pharm. 1982, 12, 299-313. Sloan, K.B.; Hashida, M.; Alexander, J.; Bodor, N.; Higuchi, T. Prodrugs of 6Thiopurines : Enhanced Delivery through the Skin. J. Pharm. Sci. 1983, 72, 372377. Sloan, K.B.; and Koch, S.A.M. Effect of Nucleophilicity and Leaving Group Ability on the SN2 Reactions of Amines with (Acyloxy) alkyl -Halides: A Product Distribution Study J. Org. Chem. 1983a, 48, 635-640. Sloan, K.B.; and Koch, S.A.M. Reactions of (Acyloxy) alkyl -Halides with Phenols : Effect of Nucleofugacity and Nu cleophilicity on Product Distribution J. Org. Chem. 1983b, 48, 3777-3783. Sloan, K.B.; Koch, S.A.M.; and Siver, K.G. Mannich base derivatives of Theophylline and 5-Fluorouracil: Synthesis, Properties and Topical Delivery Characterstics. Int. J. Pharm. 1984, 21, 251-264. Sloan, K. B.; Siver, K. G. S.; and Koch, S. A. M. The Effect of Vehicle on the Diffusion of Salicyclic Acid through Hairless Mouse Skin. J. Pharm. Sci. 1986, 75, 744-749. Sloan, K. B.; Koch, S. A. M.; Siver, K. G. S.; and Flowers, F. B. Use of Solubility Parameters to Predict Flux through Skin. J. Invest. Dermatol. 1986a, 87, 244-252. Sloan, K.B.; Beall, H.D.; Weimer, W.R.; a nd Villaneuva, R. The Effect of Receptor Phase Composition on the Permeability of Hairless Mouse Skin in Diffusion Cell Experiments. Int. J. Pharm. 1991, 73(2), 97-104. Sloan, K.B.; Della Vecchia, S.A.; Estes, J.V.; and Roberts, W. J. 7-Alkylcarbonyl and 7Alkyloxycarbonyl Prodrugs of Theophylline Int. J. Pharm. 2000, 205, 53-63. Sloan, K.B.; and Wasdo, S. Designing for Topical Delivery: Prodrugs can make a the Difference.Medicinal Research Reviews. 2003, 23, 763-793. Sloan, K.B.; Wasdo, S.C.; Ezike, U.M.; Murray, T.; Nickels, D.; Singh, S.; Shanks, T.; Tovar, J.; Ulmer, K.; and Waranis, R. Topical Delivery of 5-Fluorouracil and 6Mercaptopurine by their Alkyloxycar bonyloxymethyl prodrugs from Water: Vehicle Effects on Design of Prodrugs. Pharm. Res. 2003a, 20(4), 639-645. Sloan, K.B.; and Wasdo, S.C. “Chapter 4: The Role of Prodrugs in Penetration Enhancement.” Percutaneous Penetration Enhancers(Smith and Maibach, ed.) 2nd edn(2004), pp.51-64. Stella, V.J.; Charman, W.N.A.and Naringr ekar, V.H.; Prodrugs. Do they have Advantages in Clinical Practice? Drugs 1985, 29, 455-473.

PAGE 156

140 Stella, V.J. A Case for Prodrugs. Fosphenyltoin. Adv. Drug.Deliv.Rev 1996, 19(2), 331330. Stenback, F.; and Shubik, P. Lack of Toxi city and Carcinogenicity of some commonly used Cutaneous Agents. Toxicol. Appl. Pharmacol. 1974, 30, 7-13. Stigbrand, T.; and Fishman, W.H. Human Alkaline Phosphatases; Alan R.Liss: New York, 1984, Chapter 1. Stinchcomb, A.L.; Paliwal, R.; Dua, R.; Imoto, H.; Woodard, R.W.; and Flynn, G.L. Permeation of Buprenorphine and its 3Alkyl-Ester Prodrugs through Human Skin. Pharm. Res. 1996, 13, 1519-1523. Stinchcomb, A.L.; Swaan, P.; Ekabo, O.; Harris, K.K.; Browe, J.; Hammell, P.D.; Cooperman, T.A.; and Pearsall, M. Strai ght –chain Naltrexone Ester Prodrugs: Diffusion and Concurrent Esterase Biotransformation in Human Skin. J. Pharm. Sci. 2002, 91, 2571-2578. Stuttgen, G. Drug absorption by intact a nd damaged skin.Dermal and Transdermal absorption (R. Brandau and B. H. Lipphold, eds.). Wissenschalftliche Verlagsgesellschaft mbH, Stuttgart (1982) pp. 27-40. Swanson, H. I. Cytochrome P450 expre ssion in Human Kerati nocytes: An Aryl Hydrocarbon Receptor Perspective.Chem Biol Interact. 2004,149 (2-3):69-79. Sung, K.C.; Han, R.Y.; Hu, O.Y.P.; and Hs u, L.R. Controlled Release of Nalbuphine Prodrugs from Biodegradable Polymeri c Matrices: Influences of Prodrug Hydrophilicity and Polymer Composition. Int. J. Pharm. 1998, 172, 17-25. Vaddi, H.K.; Hamad, M.O.; Chen, J.; Banks, S.L.; Crooks, P.A.; and Stinchcomb, A.L. Human Skin Permeation for Branched Chain 3-O-Alkyl Ester and Carbonate Prodrugs of Naltrexone. Pharm. Res. 2005, 22(5), 758-765. Valiveti, S.; Hammell, D.C.; Paudel, K.S.; Hamad, M.O.; Crooks, P.A.; and Stinchcomb, A.L. In Vivo Evaluation of 3-O-Alkyl Es ter Transdermal Prodrugs of Naltrexone in Hairless Guinea Pigs. J. Controlled. Release. 2005, 102(2), 509-520. Valiveti, S.; Paudel, K.S.; Hammell, D.C.; Hamad, M.O.; Chen, J.; Crooks, P.A.; and Stinchcomb, A.L. In Vitro/In Vivo Co rrelation of Transdermal Naltrexone Prodrugs in Hairless Guinea Pigs. Pharm.Res. 2005a, 22(6), 981-989. van Hal D.A., Jeremiasse, E., Junginger, H.E., Spies, F.; and Bouwstra, J.A. Structure of Fully Hydrated Human Stratum Corneu m: A Freeze-Fracture Electron Microscopy Study. J. Invest. Dermatol. 1996, 106, 89-95. Vives, E. Cellular Uptake of the Tat Peptid e: an Endocytosis Mechanism following Ionic Interactions. J. Mol. Recog. 2003, 16, 265-271.

PAGE 157

141 Voorhees, J.J.; and Duell, E.A. Psoriasis as a Possible Defect of the Adenyl CyclaseCyclic AMP Cascade. Arch. Dermatol. 1971,104, 352-358. Taylor, H.E.; and Sloan, K.B. 1-Alkylcarbonyloxymet hyl Prodrugs of 5-Fluorouracil (5FU): Synthesis, Physicochemical Prope rties and Topical Delivery of 5-FU. J. Pharm. Sci. 1998, 87, 15-20. Thiele, J.J.; Dreher, F.; and Packer, L. Antioxidant Defense systems in skin. J. Toxicol. Cut& ocular Toxicol. 2002, 21, 1960. Tsuji, T.; and Sugai, T. Topically Ad ministered 5-Fluorour acil in Psoriarsis. Arch. Dermatol. 1972,105, 208-212. Tunek, A.; Levin, E.; and Svensson, L.A. Hydrolysis of 3H-Bambuterol, a Carbamate Prodrug of Terbutaline, in Blood from Hu mans and Laboratory Animals in vitro. Biochem. Pharmacol. 1988, 37, 3867-3876. Wacher,V.J.; Salphati, L.; and Benet, L.Z., Active Secretion and Enterocytic Drug Metabolism Barriers to Drug Absorption. Adv. Drug Deliv. Rev. 1996, 20, 99–112. Waranis, R.P.; and Sloan, K.B. The Effects of Vehicles and Prodrug Properties and their Interactions on the Delivery of 6-Mercaptopurine, thr ough Skin: Bisacyloxymethyl6Mercaptopurine prodrugs. J. Pharm. Sci. 1987, 76, 587. Waranis, R.P.; and Sloan, K.B. The Effects of Vehicles and Prodrug Properties and their interactions on the Delivery of 6-Mercaptopurine, through Skin: S6Acyloxymethyl-6Mercaptopurine prodrugs. J. Pharm. Sci. 1988, 77, 210-215. Wasdo, S.C.; and Sloan, K.B. Topical delivery of a Model Phenolic Drug: Alkyloxycarbonyl Prodrugs of Acetaminophen. Pharm. Res. 2004, 21, 940-946. Wasdo, S.C. Topical delivery of a Model Phenolic Com pound: Alkyloxycarbonyl Prodrugs of Acetaminophen Ph.D. Disseratati on University of Florida, Gainesville (2005). Watkins, P.B. Drug Metabolism by Cytochromes P450 in the Liver and Small Bowel. Gastrointest. Pharmacol. 1992, 21, 511–526. Wender, P. A.; Mitchell, D. J.; Pattabiraman, K.; Pelkey, E. T.; Steinman, L.; and Rothbard, J. B. The Design, Synthesis, a nd Evaluation of Molecules That Enable or Enhance Cellular Uptake: Pept oid Molecular Transporters Proc. Natl. Acad. Sci. U. S. A. 2000, 97, 13003-13008 Wenkers, B.P.; and Lippold, B.C. Skin Pene tration of Nonsteroidal Anti-Inflammatory Drugs out of a Lipophilic Vehicle: In fluence of the Viable Epidermis. J. Pharm. Sci. 1999, 88, 1326-1331.

PAGE 158

142 Windholz, M.; Budavari, S.; Blumetti, R. F.; and Otterbein, E. S. In The Merck Index, Merck & Co., INC. Rahway, N.J., USA, 1983, Tenth Edition. Woodford, R.; and Barry, B.W. Optimization of Bioavailability of Topical Steroids: Thermodyanamic Control. J. Invest. Dermatol. 1982, 79, 388-391. Yalkowsky, S.H. In Design of Biopharmaceutical Pr operties through Prodrugs and Analogues; Roche, E.B., Ed.; American Phar maceutical Association Academy of Pharmaceutical Sciences: Washington, D.C., 1977; pp 392-408. Yamaoka, T.; Tabata, Y.; and Ikeda, Y. Distri bution and Tissue Uptake of Poly (ethylene glycol) with Different Molecular Weight s after Intraveneous Administration to Mice. J. Pharm. Sci. 1994, 83, 601-606.

PAGE 159

143 BIOGRAPHICAL SKETCH Susruta Majumdar was born in Calcutta, India, on 21st October 1977 to Dipak and Subhra Majumdar. He and his family lived all over Northern India from Rishikesh (in the foothills of the Himalayas), to desert like Gurgaon and finally the nationÂ’s capital, New Delhi. Susruta went to the University of Delhi and got a bachelorÂ’s and masterÂ’s in chemistry in 2000. After a brief research stin t with Professor J.M.K hurana at Delhi where he worked on reduction of chalcones, he arri ved at Brigham Young University in Provo, UT, in 2001. He transferred to the University of Florida to pursue a Ph.D. in medicinal chemistry in 2002 under the superv ision of Professor Ken Sloa n. He married Antara, who is a graduate student in bi ostatistics at the University at Buffalo, NY, in 2004.