Structure-activity studies of polyamine analogues as antineoplastics

MISSING IMAGE

Material Information

Title:
Structure-activity studies of polyamine analogues as antineoplastics
Physical Description:
xiv, 188 leaves : ill. ; 29 cm.
Language:
English
Creator:
Feng, Yang, 1969-
Publication Date:

Subjects

Subjects / Keywords:
Polyamines -- analogs & derivatives   ( mesh )
Polyamines -- chemistry   ( mesh )
Polyamines -- metabolism   ( mesh )
Polyamines -- pharmacology   ( mesh )
Antineoplastic Agents   ( mesh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1996.
Bibliography:
Includes bibliographical references (leaves 174-187).
Statement of Responsibility:
by Yang Feng.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 49348419
ocm49348419
System ID:
AA00011178:00001

Full Text










STRUCTURE-ACTIVITY STUDIES OF POLYAMINE ANALOGUES
AS ANTINEOPLASTICS














By


YANG FENG


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


1996














ACKNOWLEDGMENTS


I would like to express my deepest and sincere gratitude to the chairman of my

committee, Dr. Raymond J. Bergeron, for providing me with the opportunity to complete

my graduate studies. I shall always be indebted to him for the guidance, encouragement,

and support he has provided throughout every aspect of my dissertation research and in

other areas pertinent to my scientific development. He has provided an interdisciplinary

and enriched scientific environment by allowing me to investigate in many aspects of drug

discovery.

For four years, I have experienced organic synthesis, primary activity screening in

cell culture, biochemistry studies, toxicity study in animal models and computer-aided

molecular modeling. I have also shared the excitement of promising news of our

polyamine analogues on patients with clinical doctors. It has been a truly genuine pleasure

and my good fortune to have been associated with Dr. Bergeron and his laboratory.

I would like to thank the members of my committee, Dr. Margaret 0. James, Dr.

John H. Perrin and Dr. Charles A. Sninsky, for their contributions at the inception and

culmination of this project as well as their encouragement and support throughout the

procedure.

I am also indebted to Dr. James S. McManis for his mentoring in organic synthesis.

Under his supervision, I initiated my very first organic reaction in the University of

Florida. I also would like to express my deepest appreciation to the knowledgeable Dr.

William R. Weimar, for his patience and countless hours in guiding enzymatic studies and

for answering my questions. Further, I would like to send my special thanks to Dr. Otto








Phanstiel IV and Dr. Fenglan Gao, for the encouragement, advice and friendship not only

during their stay at the University of Florida, but also in the following year.

I would also like to acknowledge and thank Hua Yao, Hristina Dimova and Brian

Raisler for their help in cell culture, HPLC analysis and animal studies. Without their

assistance, it would have been impossible to finish my dissertation.

There are other people to whom I am grateful for their help through these years in

Dr. Bergeron's group: Dr. Yao, Sam Algee, Rick Smith and Tim Vinson, and my fellow

graduate students Meiguo and Jeffery.

To my parents, Professors Chaowu Feng and Xijing Dong, I would like to express

my love and thanks for their unconditional love and guidance throughout my life. I am

forever grateful for their encouragement and support of my academic career.

Finally, I would like to express my deep love and appreciation to my husband,

Xiaodong Zhang. His love, encouragement and belief in me have carried me through the

completion of this work. I thank him for his patient understanding and the sacrifices he has

made so that I could pursue my doctorate degree and initiate my postdoctoral research.














TABLE OF CONTENTS



ACKNOW LEDGMENTS........................................................................... ii

LIST OF FIGURES ................................................................................ vi

LIST OF TABLES ................................................................................ ix

LIST OF ABBREVIATIONS..................................................................... x

ABSTRACT ........................................................................................ xiii

CHAPTERS

1 INTRODUCTION
Polyamines-General .................................................................... 1
Polyamine Biosynthesis and Catabolism........................................... 8
Polyamine Transport ................................................................ 13
Polyamine and Cancer...............................................................15

2 TRIAMINES AS ANTIPROLIFERATIVE AGENTS
Design Concept ......................................................................28
QSAR Study of Polyamine Analogues............................................34

3 SYNTHESIS
Previous Studies on Polyamine Synthesis.........................................50
Synthetic Methods....................................................................54

4 MATERIALS AND METHODS
Synthetic Methods....................................................................75
Cell Culture and Biological Assays.............................................. 100

5 BIOLOGICAL ACTIVITIES
Biological Evaluation of Tetraamine Analogues................................ 108
Biological Evaluation of Triamine Analogues.................................. 115

6 METABOLISM OF POLYAMINE ANALOGUES
Introduction......................................................................... 138
Experimental........................................................................ 141
Results .............................................................................. 145
Discussion .......................................................................... 147

7 CONCLUSION ........................................................................ 166

REFERENCES.................................................................................. 174








BIOGRAPHICAL SKETCH.................................................................... 188













LIST OF FIGURES


Eigum page

1-1. Structure of natural polyamines ....................................................... 21

1-2. Formation of hypusine ................................................................. 22

1-3. Polyamine biosynthesis and the interconversion pathway of catabolism..........23

1-4. Metabolites of polyamines formed in the terminal polyamine catabolism
pathway through oxidative deamination.............................................24

1-5. Structure of paraquat, a kind of herbicide transported by polyamine
transport apparatus..................................................................... 25

1-6. Structure of ODC inhibitor and AdoMetDC inhibitors............................26

2-1. Structure of unusual triamines found in plants and bacteria......................40

2-2. Oxidative deamination of spermidine and spermine by serum amine oxidase.
Ammonia, hydrogenperoxide, acrolein and aldehydes were generated............41

2-3. Measured versed calculated activity values for polyamine analogues.
The values presented here were expressed in the form of -logKi, directly
used in the molecular modeling. Each point represented a polyamine
analogue, and 99 polyamine analogues were included in the database ...........42

3-1. Humora and Quick's modification of spermidine at N with tert-
butoxycarbonyl group (BOC) as the protecting group............................60

3-2. Ganem's masked polyamines as cyclic urea or hexahydropyrimidines............61

3-3. Modification of spermidine and its homologues at the secondary nitrogen.
(a) Acrylonitrile (excess); (b) H2/W-2 Raney nickel/NaOH/EtOH or LiAlH4;
(c) (f) 4-Chlorobutyronitrile/butanol/Na2CO3; (d) (e) Acrylonitrile (1 equiv.).. .62

3-4. Synthesis of triprotected triamines. (A) Eugster's eight-step synthesis of
triprotected SPD. TOS = tosyl. (B) Bergeron's five-step synthesis of
triprotected NSPD (n = 3, x = 3); SPD (n = 3, x = 4) and HSPD
(n = 4, x = 4) ........................................................................... 63

3-5. Synthesis of triprotected putrescine reagent and its application in synthesis of
asymmetric spermine homologues. X = Cl or Br. (a) 4-Chlorobutyro-
nitrile/NaH/DMF; (b) H2/Ra Ni/NH3/CH3OH; (c) Mesitylenesulfonyl








chloride/NaOH(aq)/CH2Cl2; (d) NaH/DMF/2; (e) TFA/CH2CI2;
(f) NaH/DMF/3; (g) HBr in HOAc (30%)/PhOH, HC1........................... 64
3-6. Synthesis of linear tetraamine analogues. (a) (c) mesitylenesulfonyl chloride/
1 N NaOH(aq)/CH2Cl2; (b) NaH/DMF/iodomethane or iodoethane.
(d) NaH/DMF/wo-haloalkylsulfonamide (65, 66, 67 or 71);
(e) HBr in HOAc (30%)/PhOH/CH2C12, then HC1............................... 65
3-7. Synthesis of polyamine segment reagents. (a) Mesitylenesulfonyl
chloride/NaOH(aq)/CH2Cl2; (b) NaH/DMF/a,(o-dihaloalkane ...................66

3-8. Synthesis of DECroNSPM and DECroSPM. (a) EtNH2/50% NaOH/CH2Cl2;
(b) H2/Ra Ni/NH3/CH3OH; (d) NaH/DMF/a,wo-dibromoalkane;
(e) HBr in HOAc (30%)/PhOH/CH2Cl2, then HCI...............................67
3-9. Synthesis of NSPD, HSPD 4,5- and 5,5-triamine analogues.
(a) 4-Bromobutyronitrile (1 equiv)/NaHIDMF; (b) 5-Bromovaleronitrile
(1 equiv); (c) NaH/DMF; (d) H2/Ra Ni/NH3/CH30OH; (e) Mesitylenesulfonyl
chloride/NaOH(aq)/CH2Cl2; (f) NaH/DMF/haloalkane; (g) HBr in HOAc
(30%)/PhOH/CH2Cl2, then HCl; (h) NaH/DMF/4-bromobutyronitrile
(2 equiv) or 5-bromovaleronitrile (2 equiv); (i) same as (e); (j) same as (g) ......68
3-10. Synthesis of DE(4,5). (a) Oxalyl chloride/tert-butyl alcohol; (b) NaH/
DMF/66; (c) TFA/CH2CI2; (d) NaH/DMF/67; (e) HBr in HOAc
(30%)/PhOH/CH2Cl2, then HC1.....................................................69
3-11. Synthesis of monopropyl SPD and HSPD analogues. (a) Ph3CCI/CH2Cl2;
(b) Mesitylenesulfonyl chloride/1 N NaOH(aq)/CH2Cl2; (c) NaH/DMF/
68 or 69; (d) HBr in HOAc(30%)/PhOH/CH2CI2, then HCl...................70
3-12. Synthesis of homospermidine by previous methods.
(A) Okada's synthesis of HSPD. (B) Bergeron's synthesis of HSPD...........71
3-13. Synthesis of homospermidine in this study. (a) N-(4-bromobutyl)-
phthalimide/NaH/DMF; (b) (H2N)2, H20/EtOH; (c) HBr in HOAc
(30%)/PhOH/CH2CI2, then HCl.....................................................72
5-1. 96 h cell growth vs. DE(4,5) concentration.
(a) Mice leukemia L1210 cells, doubling time = 12 h.
(b) Human melanoma MALME-3M cells, doubling time = 48-72 h........... 124
6-1. Structures of alkylated polyamines involved in dealkylation
biotransform ation .................................................................... 150
6-2. HPLC chromatogram of the metabolism study of DPNSPD.
(a) Sample was taken from L1210 cells treated with DPNSPD (100 tLM).
(b) Sample of (a) spiked with MPNSPD (849 pmole/100 pL)................... 151
6-3. HPLC chromatogram of the metabolism study of DPSPD.
(a) Sample was taken from L1210 cells treated with DPSPD (100 pM)
(b) Samples of (a) spiked with MPSPD(N1) (782 pmole/100 lL)








(c) Sample of 500 jiM DPSPD treated L1210 cells spiked with MPSPD(N8)
(600 pm ole/100 gL) ................................................................. 152

6-4. HPLC chromatogram of the metabolism study of DPHSPD.
(a) Sample was taken from L1210 cells treated with DPHSPD (500 pM)
(b) Sample of (a) spiked with MPHSPD (600 pmole/100 pL)................. 154

6-5. HPLC chromatogram of the metabolism study of DP(4,5).
Sample was taken from L1210 cells treated with DP(4,5) (100 jPM).......... 155

6-6. HPLC chromatogram of the metabolism study of MPNSPD.
(a) Sample was taken from L1210 cells treated with MPNSPD (100 PM)
(b) Sample of (a) spiked with NSPD (900 pmole/100 pL) ..................... 156

6-7. HPLC chromatogram of the metabolism study of DP(4,5).
The sample was taken from L1210 cells treated with MPHSPD (100 pM)..... 157

6-8. Purification of rat liver polyamine oxidase........................................ 158

6-9. Metabolism of N1-acetylspermine and DPNSPD by partially purified
polyamine oxidase in 12 h period. The reaction product spermidine or
MPNSPD was detected by HPLC ................................................. 159

6-10. Kinetics of the metabolism of different substrates catalyzed by partially
purified polyamine oxidase ......................................................... 160

6-11. Metabolism of polyamine oxidase (PAO) catalyzed reactions.
(A) Deaminopropylation of NI-acetylspermine by PAO. (B) Depropylation
of DPNSPD by PAO ................................................................ 161

6-12. Structures of the synthetic polyamine analogues that were demonstrated
to be substrates of polyamine oxidase. The structures of NI-acetylspermine
and NI-acetylspermidine, the natural substrates of PAO, were also
included for comparison. The nitrogens attacked by PAO were
illustrated in the block ............................................................... 162

7-1. The structure-activity relationship between the polyamine analogues and
IC50 values. (a) Tetraamine analogues. (b) Triamine analogues ............... 172

7-2. The structure-activity relationship between the polyamine analogues and Ki
values. (a) Tetraamine analogues. (b) Triamine analogues..................... 173














LIST OF TABLES


Table page

1-1. Different polyamine substrates of mammalian amine oxidases....................27

2-1. Comparison of actual and calculated Ki values of polyamine analogues.........43

3-1. Polyamine analogues synthesized in this study....................................73

5-1. Tetraamine analogues structures, abbreviations, L1210 growth inhibition,
and transport.......................................................................... 125

5-2. Impact of tetraamine analogues on polyamine pools............................. 127

5-3. Effect of tetraamine analogues on omithine decarboxylase (ODC),
S-Adenosyl-L-methionine decarboxylase (AdoMet), and Spermidine/
Spermine N1-Acetyltransferase (SSAT) in L1210 cells......................... 129

5-4. Triamine analogue structures, abbreviations, L1210 growth inhibition, and
transport............................................................................... 130

5-5. Impact of triamine analogues on polyamine pools ............................... 132

5-6. Effect of triamine analogues on omithine decarboxylase (ODC),
S-Adenosyl-L-methionine decarboxylase (AdoMet), and
Spermidine/Spermine N1-Acetyltransferase (SSAT) in L 1210 cells........... 134

5-7. Comparison of the acute and chronic toxicity of tetraamine and triamine
analogues on m ice...................................................................... 136

5-8. Summation of intracellular levels of analogues and polyamines analyzed for
amine equivalence after exposure to polyamine analogues...................... 137

6-1. Metabolites detected in triamine analogue treated L1210 cells.................. 163

6-2. Metabolism of DPNSPD in different culture systems ........................... 164

6-3. Substrate study of rat liver polyamine oxidase ................................... 165













LIST OF ABBREVIATIONS


(20H)DEHSPM(R,R)
Acetyl CoA
AdoMet
AdoMetDC
AIDS
AO
BOC
BOC-ON
BSAO
CoMFA
DAO
DE
DE(3,4,4)
DE(4,3,4)
DE(4,5)
DE(4,5,4)
DE(5,4,5)
DE(5,5)
DECroNSPM
DECroSPM
DEHSPD
DEHSPM
DENSPD
DENSPM
DEPUTY
DESPD
DESPM
dcAdoMet
DFMO
DIPHSPM
DIPNSPM
DM
DM(4,5)
DM(5,5)
DMNSPD
DMNSPM
DMSPD
DMSPM
DP
DP(4,5)


N1 ,N4-Diethyl-(3R),(12R)-dihydroxylhomospermine
Acetyl coenzyme A
S-Adenosyl-L-methionine
S-Adenosyl-L-methionine decarboxylase
Acquired immunodeficiency syndrome
Amine oxidase
tert-Butoxycarbonyl
2-(tert-Butoxycarbonyloxyimino)-2-phenylacetonitrile
Bovine serum amine oxidase
Comparative molecular field analysis
Diamine oxidase
Diethyl
N1 ,N13-Diethyl(3,4,4)tetraamine
N1 ,N13-Diethyl(4,3,4)tetraamine
N1 ,N10-Diethyl(4,5)triamine
N1 ,N15-Diethyl(4,5,4)tetraamine
N1,N16-Diethyl(5,4,5)tetraamine
N1,N1 1-Diethyl(5,5)triamine
N1 ,N1 I-Diethylcrotylnorspermine
N1 ,N2-Diethylcrotylspermine
N1,N9-Diethylhomospermidine
N1 ,N14-Diethylhomospermine
N1 ,N7-Diethylnorspermidine
N1,N1 l-Diethylnorspermine
N,N'-diethylputrescine
N1,N8-Diethylspermidine
N1 ,N12-Diethylspermine
Decarboxylated S-adenosyl-L-methionine
a-Difluromethylomithine
N1 ,N14-Diisopropylhomospermine
N1,N1 l-Diisopropylnorspermine
Dimethyl
N1 ,N1 -Dimethyl(4,5)triamine
N1,N1 1-Dimethyl(5,5)triamine
N1,N7-Dimethylnorspermidine
N1 ,N1 1Dimethylnorspermine
N ,N8-Dimethylspermidine
N1,N12-Dimethylspermine
Dipropyl
N1 ,N10-Dipropyl(4,5)triamine








DP(5,5)
DPNSPD
DPNSPM
DPSPD
DPSPM
DTBHSPM
elF-5A
ETBHSPM
FAD
FBS
FDES
GSH
GI tract
HEPES
HSPD
HSPM
IC50
KB cells
ME
MEHSPM
MENSPD
MENSPM
MESPD(N1)
MESPD(N8)
MESPM
MGBG
MOP
MR
MPNSPD
MPSPD(N1)
MPSPD(N8)
MPHSPD
NADH
NADP
NADPH

NMDA
cSAT
nSAT
NSPD
NSPM
ODC
PAO
PhOH
PIP(3,3,3)
PIP(3,4,3)
PIP(4,4,4)
PIP(5,4,5)
PLS
PLP
PUT


N1,N ll-Dipropyl(5,5)triamine
N1,N7-Dipropylnorspermidine
NI,N1 l-Dipropylnorspermine
N1,N8-Dipropylspermidine
N1,N12-Dipropylspermine
N1 ,N4-Di(tert-butyl)homospermine
Eukaryotic translation initiation factor 5A
N1-(tert-Butyl)-N14-ethylhomospermine
Flavin adenine dinucleotide
Fetal bovine serum
N1 ,N12-Bis(2,2,2-trifluoroethyl)spermine
Glutathione in reduced form
Gastrointestinal tract
4-(2-Hydroxyethyl)-1 -piperazineethanesulfonic acid
Homospermidine
Homospermine
Concentration of drug required to reach 50% of inhibition
A line of human epidermoid carcinoma cells
Monoethyl
N1-Monoethylhomospermine
N1-Monoethylnorspermidine
N1-Monoethylnorspermine
N1-Monoethylspermidine
N8-Monoethylspermidine
N1-Monoethylspermine
Methylglyoxal bis(guanylhydrazone)
3-(N-Morpholino)-propanesulfonic acid
Multiple regression
N1-Monopropylnorspermidine
N1-Monopropylspermidine
NS-Monopropylspermidine
N1 -Monopropylhomospermidine
Nicotinamide adenine dinucleotide in reduced form
Nicotinamide adenine dinucleotide phosphate
Nicotinamide adenine dinucleotide phosphate in reduced
form
N-Methyl-D-aspartate
cytosolic Spermidine/Spermine acetyltransferase
nuclear Spermidine/Spermine acetyltransferase
Norspermidine
Norspermine
Omithine decarboxylase
Polyamine oxidase
Phenol
N,N'-Bis(4-piperidinyl)-1,3-diaminopropane
N,N'-Bis(4-piperidinyl)- 1,4-diaminobutane
N,N-Bis(4-piperidinylmethyl)- 1,4-diaminobutane
N,N-Bis[2-(4-piperidinylethyl)]-1,4-diaminobutane
Partial least squares
Pyridoxal 5'-phosphate
Putrescine








PYP(3,3,3)
PYP(3,4,3)
PYP(3,4,4)
PYP(4,4,4)
PYP(5,4,5)
QSAR
SAO
SPD
SPM
SSAO
SSAT
TFA
t.i.d.


N,N"-(4-pyridyl)- 1,3-diaminopropane
N,N"-(4-pyridyl)- 1,4-diaminobutane
N-(4-pyridyl)-N"-(4-pyridylmethyl)- 1,4-diaminobutane
N,N'-Bis(4-pyridylmethyl)-1,4-diaminobutane
N,N'-Bis { 2-(4-pyridylethyl) }-1,4-diaminobutane
Quantitative structure-activity relationship
Serum oxidase
Spermidine
Spermine
Swine serum amine oxidase
Spermidine/spermine NI-acetyltransferase
Trifluoroacetic acid
three injections per day









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


STRUCTURE-ACTIVITY STUDIES OF POLYAMINE ANALOGUES AS
ANTINEOPLASTICS


By

Yang Feng

December, 1996



Chairman: Raymond J. Bergeron
Major Department: Medicinal Chemistry


Triamine analogues have been selected as candidates for anticancer agents for

high activity and low host toxicity. The research presented outlines the design and

synthesis of a series of triamine analogues, their antiproliferative activities in neoplastic

cell lines in vitro and in vivo toxicities in animal models.

Computer-assisted molecular modeling is used to quantify the relationship

between structure and activity, based on a library of polyamine analogues with vast

diversity. A model with high predictive ability is obtained and used to guide the design

of triamine analogues. The triamines synthesized in this study differ in the length and

spacing of their methylene backbone, and also the size of the terminal alkyl groups.

All of the compounds were screened against murine L1210 ascites leukemia in

vitro by IC50 assay at 48 and 96 h. In an attempt to correlate this activity with some

aspect of polyamine metabolism, each compound was tested for its ability to compete

with spermidine for the polyamine uptake apparatus, its impact on the polyamine

biosynthetic enzymes ornithine decarboxylase (ODC) and S-adenosylmethionine

decarboxylase (AdoMetDC), and its effect on the polyamine catabolizing enzyme

spermidine/spermine Ni-acetyltransferase (SSAT) and on the polyamine pools.








The triamine analogues show highly antiproliferative activities, especially diethyl

4,5-triamine [DE(4,5)], which is highly active in both mice leukemia (L1210) and human

melanoma (MALME-3M) cell lines. Acute and chronic toxicity studies showed that

triamine analogues are distinctively less toxic than tetraamine analogues.

The metabolism of polyamine analogues in L1210 cells was investigated.

Polyamine oxidase is responsible for the significant N-depropylation of a series of

dipropyl triamines.













CHAPTER 1
INTRODUCTION


Polyamines-General


Polyamines are ubiquitous cell components essential for normal cell proliferation

and differentiation (Jinn8 et al. 1991; Pegg & McCann 1982). They are also recognized in

malignant transformation. Understanding of their biological function and the mechanism is

important in the regulation of abnormal cell growth. During the last thirty years, a lot of

attention has been focused on the naturally occurring polyamines, putrescine (1,4-

diaminobutane), spermidine [N-(3-aminopropyl)-1,4-butanediamine] and spermine [N,N'-

bis(3-aminopropyl)-1,4-butanediamine] (Figure 1-1). All of them are found in eukaryotes,

while spermine is absent from most prokaryotes (Morgan 1987). Two other polyamines,

cadaverine (1,5-pentanediamine) and 1,3-propanediamine (Figure 1-1) also occur in nature,

but they are only occasionally found in animal tissues (Jinne et al. 1978).

Functionally, spermidine and spermine are the most important polyamines in higher

eukaryotes (Pegg & McCann 1988). Polyamines have been shown to be essential for cell

proliferation and differentiation in numerous studies (Janne et al. 1991). In studies in

which specific polyamine biosynthesis inhibitors partly deplete polyamines and in which

polyamine-deficient mutant cells are used, cell proliferation and differentiation is halted.

Structurally, polyamines are simple aliphatic molecules composed of a linear

backbone chain containing amine moieties, usually separated by three or four carbon

methylene bridges. Putrescine, cadaverine and 1,3-propanediamine are diamines, and

spermidine and spermine are triamine and tetraamine respectively. Because they are

protonated at physiological pH (pKa's of primary amine = 8-11, pKa's of secondary amine

= 9-10), polyamines are believed to behaved as polybases.






2
The concentrations of polyamines vary considerably between different tissues and

cells (Tabor & Tabor 1964). Rapid proliferating cells, such as intestinal mucosal cells,

bone marrow and tumor cells, have a large demand for polyamines. Cells involved in

polynucleotide and protein synthesis, such as proliferating cells or protein-secreting cells,

contain spermidine and spermine in the millimolar range. Their concentrations also vary

with the cell cycle. One of the first events in cell proliferation is induction of polyamine

biosynthesis. This precedes both protein and nucleic acid synthesis (Williams-Ashman &

Canellakis 1980). Under steady-state, intracellular polyamine pools are maintained within

a relatively constant range by the collective effects of biosynthesis, degradation and

transport, all of which are biologically specific for and elaborately regulated by the

polyamines themselves (Porter et al. 1992).

Biological Roles of Intracellular Polyamines


Extensive studies show polyamines play very important biological roles, but their

exact functions have not been understood very clearly because of the complexity. As

polycharged molecules, polyamines provide endogenous cations and participate in

regulation of the intracellular pH (Canellakis 1989). Polyamines play many biological roles

in the course of their interaction with a variety of negatively charged molecules and cellular

structures which include: DNA, RNA, ribosomes, proteins, membranes, etc. The

following studies represent these major identified biological roles of polyamines.

Interaction of polyamines with DNA

Polyamines are believed to be important candidates in regulating DNA

conformation, and for controlling DNA replication and transcription by the transformation

of acetylation (Matthew 1993). In vitro, the interaction of polyamines with DNA was first

shown by the ability of spermidine and spermine to precipitate DNA. Polyamines have

been shown to stabilize DNA against thermal and alkaline denaturation, enzymatic

degradation and shear-induced strain (Tabor 1962, Cavanaugh et al. 1984, Bachrach &






3
Eilon 1969, Hung et al. 1983). The interaction of polyamine-DNA complexes was seen by

X-ray diffraction studies (Stevens 1970, Feuerstein et al. 1991). The stabilizing effects

were attributed to neutralization of the negative charges on the phosphate groups and the

consequent decrease in the level of stacking energy. At physiological concentration,

polyamines can condense DNA and chromatin and promote B to Z DNA transitions

(Gosule & Schellman 1978). It has been found that the structure of polyamines plays an

important role in the binding of DNA and chromatin. The presence of the charges and a

specific chain length of methylene (-CH2-) group is required for optimal binding (Vertino

et. al 1987, Balasundaram & Tyagi 1991). Acetylation of polyamines is an important mode

of regulating polyamine-chromatin interaction, by diminishing the number of charges on

polyamine molecules and as a result changing the helical twist of DNA caused by

polyamine binding in nucleosomes.

Interaction of polyamines with RNA

Intracellular levels of polyamines and of ornithine decarboxylase markedly increase

when RNA synthesis is stimulated. A dramatic example is the l,(XX)-fold stimulation of

ornithine decarboxylase in kidney cells after androgens are administrated to castrated mice,

accompanied with a substantial increase in RNA levels and a reduction in DNA content

(Henningsson et. al 1978). Spermine can bind to specific sites on the rRNA molecule, and

be involved both in stabilizing the conformation of t-RNA and in facilitating t-RNA splicing

(Sakai & Cohen 1973, Tabor & Tabor 1984).

Effect of polyamines on protein biosynthesis

Many papers have been published on stimulation of in vitro protein biosynthesis by

polyamines, and a few papers have addressed the effects of polyamines in vivo. The

addition of polyamines results in a qualitative difference in the polypeptides synthesized in

a cell-free system (Atkins et al. 1975). A combination of Mg2+ and polyamine in vitro

results in higher incorporation of amino acids than does optimal Mg2+ concentration alone

(Tabor & Tabor 1976, Abraham & Phil 1981). Polyamines can facilitate the association of








ribosome subunits, increase the fidelity of translation and facilitate chain termination

(Rosano et al. 1983, Thompson & Kaim 1982, Hryniewicz & Vonder Haar 1983). The

effects of polyamines on polypeptide chain initiation and elongation has been studied since

the 1970s (Jorstad & Morris 1974, Igarashi et al. 1978, Sander et al. 1978), but there was

not a clear explanation of the process until hypusine, a post-translational modified amino

acid, was found.

Hypusine serves as a unique and essential amino acid constituent of the eukaryotic

translation initiation factor eIF-5A (Park et al. 1993). It occurs at a single position per

protein. A spermidine derivative, hypusine is formed as a post-translational modification

of the elF-5A precursor protein in the two-step process (Figure 1-2). The first step is the

formation of deoxyhypusine, which involves the transfer of the 4-aminobutyl group from

spermidine to lysine-50 in the eIF-5A precursor protein by deoxyhypusine synthase. The

second step is the hydroxylation of deoxyhypusine by deoxyhypusine hydroxylase.

The enzymes involved in hypusine biosynthesis, especially deoxyhypusine

synthase has attracted many interests as a target for growth inhibition. A couple of diamine

and polyamine analogues have been reported to be good inhibitors of deoxyhypusine

synthase. The spermidine binding site of this enzyme has also been defined through

inhibitor studies by Jakus et al. (1993). The result provided a basis for potential control of

protein biosynthesis and cell proliferation. By using guanyl diamine as an inhibitor of

deoxyhypusine synthase, Park et al. postulated that the growth inhibition of Chinese

hamster ovary cells was not mediated through an interference with polyamine metabolism,

but through inhibition of hypusine (1994). A recent study on the cytostasis induced by S-

adenosyl-L-methionine decarboxylase inhibitor (AbeAdo) demonstrated the growth

inhibition was due to the depletion of hypusine-containing form of eIF-5A, which is

secondary to the depletion of spermidine by inhibition of S-adenosyl-L-methionine

decarboxylase (Byers et al. 1994). All of these observations suggest that hypusine is one

of the links between polyamines and growth regulation.








Functions of Extracellular Polyamines


Besides the important biological functions of intracellular polyamines, endogenous

polyamines are found to have multiple effects in the central nervous system and have been

suggested to be neurotransmitters or neuromodulators. One of the effects is to regulate the

activity of the N-methyl-D-aspartate (NMDA) receptor channel, a subtype of glutamate

receptor channels (Williams et al. 1991, Scatton 1993).

Polyamines and NMDA receptor

NMDA receptors are modulated by several endogenous ligands, including glycine,

hydrogen ions, and the divalent cations Mg2+ and Zn2+. Ransom and Stec first observed

that the polyamines, spermine and spermidine, caused an increase in the binding to NMDA

receptors of open channel blocker MK-801, suggesting that polyamines enhance NMDA

receptor activity (1988). Because spermine does not activate NMDA receptors in the

absence of glutamate and glycine, it may act on an allosteric site independent from the

binding sites for these co-agonists (Ransom & Stec 1988, McGure et al. 1990).

Another action of polyamines at NMDA receptors is found as inhibitory effect.

Spermine acts as either an agonist or antagonist depending on its concentration. Recent

studies implied that potentiation and blocks are separate processes (Rock & Macdonald

1992, Benveniste & Mayer 1993).

The complexity of the effects of polyamines on NMDA receptors suggests that there

may be more than one polyamine-binding site on the receptor channel, and at least one

specific polyamine binding site has been demonstrated on the receptor complex (Benveniste

& Mayer 1993, Williams et al. 1989). Eletrophysiological studies have shown that

polyamines enhance NMDA receptor currents by increasing channel opening frequency and

by increasing the affinity of the receptor for glycine (Rock & Macdonald 1995). On the

other hand, polyamines reduce NMDA receptor currents by producing voltage-dependent

reduction of single-channel amplitudes and/or by producing an open channel block. This

finding represents that polyamine specific NMDA receptors could be novel therapeutic






6
target for the treatment of ischemia-induced neurotoxicity, epilepsy, and neurodegenerative

diseases, which are believed at least partly to be caused by overstimulation of NMDA

receptors.

Polyamine and gastrointestinal tract

Polyamines have relatively high concentrations in the gastrointestinal (GI) tract

(Tabor & Tabor 1964). Polyamines absorbed from GI tract are a major source of

exogenous polyamine, which is either acquired from diet or synthesized in the gut lumen

by a high population of microbial flora (Saydjari et al. 1989). It has been found that the GI

tract is a major source of polyamine for tumor growth, especially when the cellular

polyamine's biosynthesis is terminated by enzyme inhibitors (Sarhan et. al 1989, Hessels et

al. 1989).

Polyamines not only are essential for the regulation of intestinal growth as non-

peptide growth promoting compounds (Johnson 1988), they also play a very important role

in the GI tract--an inhibition effect on GI tract motility. In 1967, De Meis first reported his

findings on natural polyamine's relaxing effect on intact and glycerol-treated muscle. In his

study, spermidine and spermine induced relaxation of smooth muscle (guinea pig ileum and

tanenia coli) which were made to contract by acetylcholine, histamine, nicotine, caffeine

and excess of potassium. The inhibitory effect of synthetic polyamine analogues on gastric

emptying was first found in studies of the commercial polyethyleneimines, widely used as

fungicides, bactericides, antiviral and antitumor agents (Melamed et al. 1977). Further

studies showed the effect of other low molecular-weight polyamine analogues on gastric

emptying. Interestingly, although both spermidine and spermine very actively inhibited

gastric emptying, their analogues, norspermidine and norspermine, were inactive.

Triamine N-aminoethyl-1,4-diaminobutane and tetraamine N, N'-bis(3-aminopropyl)

piperazine showed very low activity in inhibition of gastric emptying (Balair et al. 1981).

It was clear that polyamines had a profound influence on gastric emptying and that








endogenouss spermidine and spermine may have some unrecognized GI secretomotor

activity" (page 347, Balair et al. 1981).

Polyamine pharmacophore has been further investigated in Bergeron's group as an

excellent candidate for construction of antitransit, antidiarrheal drugs. From a structure-

activity perspective, it is obvious that very small changes in the polyamine's structure could

completely eradicate the molecule's ability to inhibit gastric emptying. Among the synthetic

analogs of polyamines, diethylhomospermine (DEHSPM) profoundly inhibits

gastrointestinal motility in rats and this inhibition was reversed with the co-administration

of bethanecol, a cholinergic agonist, but not with other pharmacologic antagonist or agonist

(Sninsky et al. 1993). This drug is now studied in clinical trial phase II as a new class of

antidiarrheal agents, especially in treating AIDS-related infectious diarrhea. The latest

study showed that by introducing two hydroxyl groups into the aminobutyl segments of

DEHSPM, a novel class of antidiarreal drug was discovered with high activity and much

lower toxicity than DEHSPD (Bergeron et al. 1996).

The mechanism involved in the inhibition of GI tract motility is not quite clear at

this point. Studies on other smooth muscles suggested that polyamines inhibit smooth

muscle contraction at the plasma membrane which decreases the influx of calcium (Martin

& Tansy 1986, Chideckel et al. 1985). In study of isolated segments of the rat small

intestine, the inhibition of spermidine was sensitive to increased calcium (Chideckel et al.

1985). This phenomenon was also observed in permeabilized smooth muscle of guinea pig

ileum (Sward et al. 1994).

Localization of Polyamines

Along with the investigation of polyamine biological roles, considerable effort has

been exerted in order to obtain detailed knowledge of exact cellular and subcellular

locations of endogenous polyamines. Regular chemical analysis of tissue sample and

subcellular fractions was limited due to the diversity of cell types in most tissues and the








tendency of polyamines to redistribute during preparative procedures (Hougaard 1992).

Cytochemical methods showed advantages over the chemical analysis and clearly

demonstrated that in all proliferating cells and in certain differential cell types, polyamines

are mostly in the cytoplasmic compartments (Hougaard 1992). In a large number of

protein- or peptide-secreting cell systems, including exocrine, endocrine and

neuroendocrine cell types, the polyamines are mainly localized in secretary granules and

often co-localized with peptide growth factors in secretary granules. In addition,

polyamines are closely associated with DNA in condensed chromatin, demonstrated by the

necessity of DNase digestion prior to cytochemical staining. All of these cytochemical

locations of polyamines agree with the documented biological functions mentioned above

and indicate that the molecules may have many different functions during the cell cycle and

in nonproliferating differentiated cells.

Polyamine Biosynthesis and Catabolism


That the levels of polyamines are elaborately regulated in the cell has been

sufficiently documented. In mammalian cells, polyamines are derived from amino acids

arginine and methionine (Tabor & Tabor 1984, Figure 1-3). Putrescine serves as an

immediate precursor and also as a degradation and excretion product of the polyamines. In

certain microorganisms and plants, the arginine is decarboxylated to form agmatine which

is subsequently converted to putrescine (Luk & Casero 1987). Eukaryotes lack the enzyme

which catalyzes the decarboxylation, and thus the only source of putrescine is by

decarboxylating ornithine with ornithine decarboxylase (ODC) (Hayashi 1989). Ornithine

is derived from the plasma or intracellular arginine by the action of arginase (Williams-

Ashman & Canellakis 1979, Pegg 1986). In the cells, ornithine can accumulate to near

millimolar quantities (Porter 1988).

The biosynthetic and catabolic pathways for putrescine, spermidine and spermine

are well established (Tabor & Tabor 1984, Seiler 1990). There are two major pathways






9
along which polyamines are metabolized: the introconversion pathway and the so-called

"terminal polyamine" catabolism, in which the products can not he re-utilized and are

urinary excretory.

The interconversion pathway is a cyclic process which controls polyamine turnover

(Figure 1-3). Putrescine is converted to spermidine and spermine through two consecutive

actions of two aminopropyl transferases, spermidine synthase and spermine synthase.

Both of the enzymes use decarboxylated S-adenosyl-L-methionine (dcAdoMet) as an

aminopropyl donor under the action of S-adenosyl-L-methionine decarboxylase

(AdoMetDC). The dcAdoMet is formed from S-adenosyl-L-methionine. Unlike

putrescine, dcAdoMet is found in detectable quantities only in unperturbed cells (Porter &

Bergeron 1988). The degradation of spermidine and spermine occurs first by conversion

to monoacetyl derivatives, involving spermidine/spermine N1-acetyltransferase (SSAT) and

using acetyl CoA as the acetyl group donor. The acetyl derivatives are substrates for a

flavin dependent polyamine oxidase (PAO), which splits the NI-acetylpolyamines into an

aldehyde (3-acetamidopropanal) and a polyamine, containing one less propylamino group.

As a result, spermidine is formed from spermine and putrescine from spermidine.

Terminal polyamine catabolism is catalyzed by several kinds of amine oxidases

which act at the primary amino groups. These amine oxidases include monoamine oxidase,

diamine oxidase and plasma amine oxidase. By oxidative deamination of a primary amino

group, each intermediate of the interconversion cycle can be transformed into an aldehyde,

which is further oxidized to an amino acid or a y-lactam (Figure 1-4). The products of the

terminal catabolism are finally excreted in urinary form.

All of the amine oxidases involved can be divided into two categories, copper-

containing and flavin adenine dinucleotide (FAD)-dependent amine oxidases (as shown in

Table 1-1) (Morgan 1989). (1) Copper-containing amine oxidases (Cu2+-containing AOs)

deaminate oxidatively the primary amino groups) of histamine, monoamine and diamine to

the corresponding aldehydes (as shown blow). To this type of enzyme belong serum






10

amine oxidase and diamine oxidase. (2) The FAD-dependent amine oxidase is defined as a

flavinprotein acting on primary amines, and usually on secondary and tertiary amines with

small substituents (as shown blow). Monoamine oxidase and polyamine oxidase are FAD-

dependent oxidases. In both cases the enzyme is considered to act on a CH2-NH2 group of

the donor and oxygen is the electron acceptor.

Cleavage at the primary amino group

(operated by both Cu- and FAD-dependent AOs)
R-CH2-NH2 + H20 + 02 -+ R-CHO + NH3 + H202


Cleavage at the secondary amino group

(operated by the FAD-dependent AOs)

R-CH2-NH-CH2-R' + H20 + 02 -* R-CHO + H2N-CH2-R' + H202

or R-CH2-NH-CH2-R' + H20 + 02 R-CH2-NH2 + HOC-R' + H202

Enzymes in Interconversion Pathway


Three rate-limiting enzymes, ornithine decarboxylase (ODC, EC 4.1.1.17), S-

adenosyl-L-methionine decarboxylase (AdoMetDC, EC 4.1.1.50), and

spermidine/spermine NI-acetyl-transferase (SSAT), are the key enzymes that regulate

polyamine metabolism.

First, rate-limiting enzymes are recognized by their low activity. Studies show that

the three enzymes listed above have the lowest activity among all of the enzymes involved

in polyamine metabolism (Seiler 1990). The activity of these regulatory enzymes can

change rapidly under different circumstances by different mechanisms. Second, these three

enzymes are regulatory enzymes which turn over quickly, with biological half-lives

between 20 and 40 min (Russell & Snyder 1969, Seyfried et al. 1982, Matsui & Pegg

1981).






11
The ODC has been isolated from a considerable array of species. In all cases, it has

been found to be dependent on pyridoxal 5'-phosphate (PLP) for activity. Studies of ODC

from different mammalian sources showed over 90% identity in amino acid sequences

(McCann & Pegg 1992). In general, induction of ODC is either physiological (hormones,

growth factors) or non-physiological (tumor promoters, toxic agents). The induction of

ODC is mainly from enhanced transcription and protein synthesis (Porter and Bergeron

1988). However, the decreased enzyme activity is associated with post-transcriptional

regulation. The identified mechanisms involve either reduced translation of ODC mRNA or

increased activity of an ODC-directed protease.

The second regulatory enzyme, AdoMetDC, is an unusual decarboxylase in the way

that it uses the covalently bound pyruvate as a prosthetic group instead of using pyridoxal

phosphate (Pegg 1984). Then, the carbonyl group of pyruvate forms a Schiff base with

the substrate (Van Poelje & Snell 1990). Accelerated activities of AdoMetDC is considered

to be mediated by increases in enzyme synthesis, in half-life and in transitional efficiency of

mRNA. Alternatively, inhibition of mRNA translation is believed to be responsible for the

reduction of AdoMetDC related to higher levels of intracellular polyamines (Porter &

Bergeron 1988).

The third regulatory enzyme, SSAT, was recognized simply as a acetyltransferase

when monoacetyl derivatives of polyamines were first found in brain cells and urine 30

years ago (Libby 1980). The association of SSAT with the metabolic interconversion of

polyamines was not realized until NI-acetylspermine was found to be an extremely efficient

substrate for rat liver cytoplasmic polyamine oxidase (PAO) (Bolkenius & Seiler 1981).

The latter oxidatively eliminated acetylated aminopropyl units of SPD or SPM by their

conversion to 3-acetamido-propanal. In this way, PAO converted the NI-acetylspermine to

spermidine and NI-acetylspermidine to putrescine (Hilttii 1977). There are three forms

SSAT capable of the acetylation of polyamine: nuclear acetyltransferase (n-SAT),

noninducible cytosolic acetyltransferase and inducible cytosolic acetyltransferase (c-SAT)






12

(Seiler 1987). The first two are not inducible, while the last one, as its name suggests, is

highly inducible. The high inducibility of cytoplasmic SSAT was first reported in livers

exposed to carbon tetrachloride (Matsui & Pegg 1980). Numerous in vitro and in vivo

systems respond to a wide variety of physiological, pathological and pharmaceutical stimuli

(Seiler 1987). Among the pharmacological stimuli, the antiproliferative agent,

methylglyoxal bis(guanylhydrazone) (MGBG) (Persson & Pegg 1984, Nuttall & Wallace

1987) and some polyamine analogues are potent inducers of SSAT. i.e. diethylnorspermine

(DENSPM) induced SSAT activity 15 times over unstimulated level after a 48 h treatment

of murine leukemia cell at 2 pM (Bergeron et al. 1994, Libby et al. 1989).

In contrast to the regulatory decarboxylases and acetyltransferase, spermidine and

spermine synthase are usually present in cells and tissues at high levels of activity. They

are stable proteins with biological half lives of several days (Seiler 1990).

Polyamine oxidase (PAO) is also a stable enzyme. It contains a tightly bound FAD

prosthetic group and acts on the secondary amine group of Nl-acetylpolyamine derivatives

of spermidine and spermine yielding spermidine and putrescine, respectively, along with 3-

acetamido-propanal and hydrogen peroxide. Although this enzyme is also able to

transform spermine into spermidine, and spermidine into putrescine in the presence of

benzylaldehyde, it has demonstrated that N1-monoacetylspermidine, N1-

monoacetylspermine and N1, N12-diacetylspermine are much better substrates for PAO

than the non-acetylated polyamines (Bolkenius & Seiler 1981).

Enzymes in Terminal Pathway

Three types of amine oxidases have been found involved in the terminal pathway.

They are monoamine oxidase, diamine oxidase and serum oxidase. All of them act on the

primary amino groups and generate aldehyde, hydroxyperoxide and ammonia. These

products are believed to be related to both in vitro and in vivo toxicity of polyamines

(details in Chapter 2).








The monoamine oxidase is a flavin-dependent amine oxidase found in

mitochondria. It has been widely investigated for its primary involvement in the

metabolism of the biogenic monoamine neurotransmitters. It only plays a role of secondary

importance in the oxidation of acetylated polyamines (Mondovi et al. 1989). The

polyamine substrates of MAO are listed in Table 1-1.

Diamine oxidases (DAO) are Cu2+-containing enzymes localized in mitochondria

and microsomes. They are particularly active towards aliphatic diamines (e.g. putrescine,

cadaverine, Table 1-1) and histamine. The DAOs also catalyze the deamination of SPD,

SPM and their derivatives as long as there are primary amino groups available (Table 1-1).

Serum oxidase (SAO), the copper-containing amino oxidase, which is especially

rich in ruminant serum, was the first polyamine-selective oxidase to be described (Hirsch

1953). Among the SAOs, bovine serum amine oxidase (BSAO) and swine serum amine

oxidase (SSAO) are the best characterized (Morgan 1989, Mondovi et al. 1989). Both

enzymes catalyze the oxidative deamination of primary amino groups of several aliphatic

and aromatic monoamines, but while BASO oxidizes spermidine and spermine, SSAO

does not. The polyamine substrates of SAO is shown in Table 1-1.

Polyamine Transport


Polyamine transport system is sensitively regulated by polyamine levels in the cells.

Along with the polyamine biosynthesis and catabolism pathways, it maintains the

intracellular polyamine pools at a relatively constant level.

Polyamine transport system has been found in a wide range of mammalian cells and

it is believed to be protein nature (Kramer et al. 1993). Through kinetic studies in most

systems, the putative transport protein has been shown to be saturable and distinctly energy

dependent with separate sodium-dependent and sodium-independent components (Seiler &

Dezeure 1990).








Polyamine uptake is essential and sufficient to sustain cell growth fully in cells

where the biosynthetic pathway is genetically defective or drug inhibited (Pilz et al. 1990,

Kramer et al. 1989). Under such conditions, cells may up-regulate their polyamine

transport system in order to ensure that intracellular pools are maintained (Alhonen-

Hongisto et al. 1980). On the other hand, in order to avoid toxicity associated with

polyamine excess, the uptake mechanism may also be down regulated in response to a

surplus of polyamines (Kankinuma et al. 1988, Byers & Pegg 1990).

The substrate for the polyamine uptake system is not strictly limited to naturally

occurring polyamines. A wide variety of compounds are known to be transported by this

system, including synthetic molecules such as the herbicide, paraquat (Figure 1-5) (Byers

et al. 1987), the SSAT inhibitor, methylglyoxal-bis(guanylhydrazone (MGBG) (Dave &

Caballes 1973, Porter et al. 1982) and series of polyamine analogues (Bergeron et al. 1988

and 1989, Byers & Pegg 1990).

The polyamine transport system also responds to various physiological conditions.

The activity of the system increase during hormonal stimulation and proliferation and

decreases during differentiation (Rinehart & Chen 1984). While under conditions that

disfavor proliferation, for example, achievement of confluence, absence of growth factors

or existence of inhibitors, there is a substantial efflux of polyamines, particularly

spermidine, from the cell (Jinnd et al. 1983).

Like polyamine uptake, polyamine depletion also plays an important role in

maintaining the polyamine pool level. In vitro studies, a massive efflux of spermidine was

detected in a short 24 h period, when cells were exposed to polyamine analogues.

Spermine, different from spermidine, can not be excreted from cells as free amine. It needs

to be acetylated by the action of acetyltransferase before depletion. Acetylation reduces the

polyamine's charge and so results in weaker binding in various cellular compartments. In

whole animal studies, free polyamines account for about 75% of the polyamines found in








rat urine, while humans excrete the polyamines nearly exclusively as monoacetyl

derivatives (Seiler et al. 1981).

Polyamines and Cancer


As polyamines are related to proliferation, differentiation and regeneration of

normal cells, not surprisingly, they are also recognized in malignant transformation. As

early as 1853, it was reported that leukemic spleen was rich in spermidine (Chartot &

Robin). This observation was extended almost a hundred years later by HImaliinen who

found increased concentrations of spermine in the liver and bone marrow of patients that

had died of leukemia (Janne et al. 1978). In 1971, Russell et al. reported that patients with

a variety of tumors had elevated levels of polyamines in their urine. Since then,

polyamines and their acetylated-forms have been widely studied as markers for malignant

tumors. However, the usefulness of polyamines as markers is limited to only several

malignant diseases: breast, prostatic, bladder and colon cancers (Horn et al. 1982, Czuba &

Smith 1991). Although polyamines can not be used in screening for malignancy because

of nonspecific elevation, the polyamine assays are found valuable to monitor the long term

progression or regression of tumors and to evaluate short-term efficacy of therapy.

The polyamine biosynthetic pathway has attracted a lot of interest as a therapeutic

target. Enzyme inhibitors with the ability to interfere with polyamine biosynthesis were

first chosen as therapeutic agents. Numerous studies of ornithine decarboxylase and the

therapeutic effect of inhibition of ODC occurred in the late 1970s and 1980s. The most

widely used specific inhibitor of ODC, ao-difluoromethylomithine (DFMO) (Figure 1-6),

leads to a major reduction of putrescine and spermidine. As a result, growth is

significantly reduced (Pegg & McCann 1988b, Marton & Pegg 1995). However, in most

cases, DFMO only causes a small reduction in spermine. Like other specific inhibitors, the

effects of DFMO are cytostatic instead of cytotoxic. It is suggested that the lack of

cytotoxicity may be due to the residual spermine in the cells. The atom of DFMO as a





16

irreversible inhibitor starts with its recognition as a substrate by ODC, generating a highly

reactive intermediate by decarboxylation of DFMO, and finally formation of a covalent

bond with the protein (Metcalf et al.1978, McCann & Pegg 1992). The actual sites of

DFMO binding have been identified (Poulin et al. 1992). The major site of binding was the

cysteine residue at position 360 and is contained in a highly conserved region of ODC in

the peptide -GPTCD- found in all known eukaryotic ODC sequences. This site accounts

for about 90% of the total binding. It has been suggested that cysteine 360 may be located

close to the active site of the enzyme.

DFMO has shown significant efficacy as a single agent in slowing the growth of

tumors cells in vitro and in many animal models in vivo as a single agent (Sunkara et al.

1987). However, DFMO did not seem to significantly affect tumor growth or progression

of diseases in more than 500 patients with a variety of malignancies (Schecter et al. 1987)

The first specific inhibitor of AdoMet is MGBG (Figure 1-6), introduced by

Williams-Ashman and Schenone (1972). It exerted significant enzyme inhibition, both in

vitro and in vivo (Pegg 1988). However, its usefulness is limited because (1) it is a

reversible inhibitor; (2) it is not absolutely specific and also inhibits some other enzymes,

such as diamine oxidase; and (3) it is highly toxic (Pegg & McGill 1978). The discovery

of MGBG led to the design and synthesis of a number of specific AdoMetDC inhibitors,

which were expected to provide potent antiproliferative agents (Figure 1-6) (Pegg &

McCann 1992). However, like MGBG, the inhibition effects of these agents generally lead

to a large increase in putrescine and a decline in spermidine and spermine. As putrescine

content rises by more than the decrease in the spermidine and spermine levels, the total

supply of polyamines actually increases as the inhibitor is applied. In cell culture, the

growth inhibitory effects could be completely abolished by simultaneous administration of

spermidine (Kramer et al. 1989, Regenass et al. 1992). For this reason, the therapeutic

values of these inhibitors may be limited due to the reversal of growth inhibition by

exogenous spermidine.






17

There are significant problems with the use of inhibitor of polyamine biosynthetic

enzyme: (1) the key regulatory enzymes in the biosynthesis--ODC and AdoMetDC are

under the feedback regulation of polyamine levels. Depletion of polyamine level will

induce a large compensatory increase in these enzyme activities and uptake of exogenous

polyamines, which in turn will decrease the inhibition; (2) turnover rate of these enzymes is

so fast, the cancer cell is able to overcome the block quickly, thus making the inhibition

transient. This explains the limited clinical value of these enzyme inhibitors, at least when

they are used as a single agent.

Polyamine analogues, the laboratory constructed imitation polyamines, have

attracted considerable attention during recent years, as an alternative to direct enzyme

inhibitors. Although polyamine analogues can not substitute natural polyamines in cell

proliferation, they are accepted by the cell as the natural polyamines in many circumstances.

When facing a substantial amount of pseudo-spermidine or spermine, cells take in the

"ready-made nutrition" and at the same time shut down their own polyamine synthesis to

save energy. Intracellular natural polyamine are then excreted or degraded to keep the

balance of polyamine pool.

Compared to the enzyme inhibitors, polyamine analogues have several advantages

as a regulatory approach to interfere with polyamine biosynthesis: (a) polyamine analogues

use the specific cellular polyamine transport apparatus; (b) simultaneous regulation of

different enzymes can be managed, including suppression of biosynthetic enzymes and

induction of catabolic enzymes; (c) compensatory increases in related enzymes do not occur

as with enzyme inhibitors; (d) compensatory increase in cellular uptake of polyamines do

not occur; (e) depletion of all polyamine pools, including spermine, is possible; (f)

structure modification of polyamine analogues can make them immune from acetylation by

SSAT and have a much longer half-life than the natural polyamines. As a result, the

analogues can be concentrated up to 1000 times more in the cell than in the environment.








This is one of the reasons for high activity of polyamine analogues at a very low p.M

concentration (Porter & Bergeron 1988).

The first series of polyamine analogues was linear aliphatic triamines and

tetraamines homologues of spermidine and spermine (Israel et al. 1964). Antitumor

activities were found both in vitro against human epidermoid carcinoma (KB) and in vivo

against transplantable mouse tumors. Among the analogues, a spermine analogue, N,N'-

bis(3-aminopropyl)nonane-1,9-diamine was found to be most active. In 1981, Weinstock

et al. demonstrated that some of the homologues and acylated derivatives of spermidine and

spermine were active against B16 melanoma and human epidermoid carcinoma of the

nasopharynx.

It was observed that when polyamine biosynthesis was shut down in transformed

cells by treatment of DFMO, the cells incorporated spermidine at an accelerated rate relative

to untreated cells (Jiinn et al. 1978). This encouraged some researchers to consider the

spermidine uptake apparatus as a means of delivering antineoplastic drugs to transformed

cells (Porter et al. 1985). It was found that the uptake of MGBG (an AdoMet inhibitor), an

antineoplastic, also occurred via the polyamine transport apparatus and was enhanced by

pretreatment of the cells with DFMO (Casero et al. 1984). Noticing the limited structural

similarity between MGBG and spermine, Porter and Bergeron suggested that the

polyamine transport system might tolerate some structural modification on natural

polyamines. This reasoning led to the design and synthesis of a new series of polyamine

analogs, N4-spermidine and NI,N8-spermidine derivatives (Porter et al. 1982 and 1985).

It has been found that (1) terminal alkylated spermidine analogues have much higher

activity than the acetylated ones; (2) antiproliferative activity is higher when alkylation at

terminal nitrogens (N1 and N8) than when it is at the middle nitrogen (N4); (3) uptake is

dependent on the availability of the primary amines. Among all of the analogues, N1,N8-

diethylspermidine (DESPD) was found to be the most active. After treatment of 96 h at 10

^pM, DESPD depleted intracellular putrescine and spermidine, reduced spermine by 500%






19

and decreased omithine and S-adenosylmethionine decarboxylase by 98%. The regulation

mechanism of DESPD on ODC was the same as that of spermidine, which decreased the

enzyme activity by down regulation rather than direct bind and inactivate the enzyme

(Porter et al. 1986). Studies also indicate that while sharing the same cellular uptake

system, DESPD can not substitute for spermidine to support cell growth (Porter &

Bergeron 1988).

The diethyl derivatives of putrescine and spermine were also tested for their ability

to regulate polyamine biosynthesis and inhibit L1210 leukemia cell growth (Porter et al.

1987). The antiproliferative activities are in the order of diethyl spermine (DESPM) >

DESPD > diethyl putrescine (DEPUT). The DESPM also had significant effect on

polyamine depletion and enzyme inactivation. As the tetraamine analogues showed the

highest activity, more diethyl alkylated tetraamines were investigated as antineoplastics

(Bergeron et al. 1989). The order of antiproliferative activity in vitro was shown to be

DEHSPM > DESPM > DENSPM. In vivo studies, in human pancreatic adenocarcinoma

models, DENSPM showed greater antitumor activity than either DEHSPM or other

conventional agents (Chang et al. 1992). The DENSPM also demonstrated remarkably

high inhibition to the growth of MALME-3M and SN-I melanomas; A549 lung

adenocarcinoma; and A121 ovarian carcinoma (Marton & Pegg 1995). It is believed to be a

very promising anticancer drug, and its phase II trials are in process.

A systematic investigation is needed in order to first define the minimal structural

requirements of a polyamine analogue necessary for antineoplastic behaviors. Once having

done this, we can then modify nonessential components of the molecule possibly to

minimize toxicity. It is clear that the structural changes in polyamine analogues are closely

related to their differences in biological activities. In order to study the relationship

between structure and activity, a library of polyamines, especially tetraamines, has been

constructed in Dr. Bergeron's lab. Extensive studies have been conducted to find the

correlation between biological activities and the following structure factors: (1) charge






20

distribution, (2) distance (methylene backbone) between the nitrogen atoms, (3) changes of

size in terminal alkylation groups (Bergeron 1995a, 1989 and 1994).

In this study, a number of tetraamines were synthesized and tested for their

antiproliferative activities, their impact on polyamine pools, the effect on polyamine

regulatory enzymes (including ODC, AdoMet and SSAT) and their ability to compete with

radiolabelled spermidine for polyamine uptake apparatus. All of the tetraamine analogues,

including the ones synthesized in current and from previous work, were used to draw a

clear picture of structure and activity relationship by computer-assisted molecular modeling.

In light of the low toxicity of triamine analogues compared to tetraamine analogs, a series

triamine analogues were designed, synthesized and tested for antiproliferative activity.

Computer-assisted molecular modeling was used to guide the design.









H2N ,NH2






H2N NH2






H2NN ., -. NH2






H
H2N NNH2


1,3-propanediamine






putrescine







cadaverine







spermidine


H
H2N N, N '.-,, NH2
H


spermine


Figure 1-1: Structure of natural polyamines.










elF-5A Precursor


H
~N o NH2

0


H2N N N NH2
H


Lysine Residue

I


1,3-diaminopropane


Spermidine



Deoxyhypusine Synthase


H
N N H2


eIF-5A Intermediate


Deoxyhypusine


Deoxyhypusine Hydroxylase


H OH
N N HNH2

0)


Hypusine Residue


Figure 1-2: Formation of hypusine.


I~;I








CH3 COOH
I I
S(CH2)2CHNH2
(L-methionine)


HOC CH3 Adenine


HO OH
(S-Adenosylmethionine)

S-adenosylmethionine
decarboxylase
(AdoMetDC)


CH3 Adenine


HO OH
(Decarboxylated
S-Adenosylmethionine)







Adenine
HaC S 0

HO OH
(Methylthioadenine)


NH
H2N A 2,. NH2
n COOH
(L-arginine)

I Arginase

H2N NH2
COOH
(L-ornithine)

I Ornithine Decarboxylase
(ODC)


Spermidine
Synthase


Spermine
Synthase



I.


N N NH2
H
N1 -Acetylspermine
ISSAT


Figure 1-3: Polyamine biosynthesis and the interconversion pathway of catabolism.








O

H2N
0






H
"HO NNH2

0


H2N S N p

0

H
H3C HNN

0 0


H
NH2N 0 OH
H O


4-Aminohutyric acid



Pyrrolidin-2-one




Putreanine


Isoputreanine lactam




N-Acetyl isoputreanine lactam





N8-(2-Carboxyethyl) spermidine


0

HO',,.^ N
H


,,, OH

0


Spermic acid


Figure 1-4: Metabolites of polyamines formed in the terminal polyamine catabolism
pathway through oxidative deamination.






25










CH3 +N N+---CH3 2CI



Figure 1-6: Structure of paraquat, a kind of herbicide which transported by
polyamine transport apparatus.








ODC Inhibitor


F HC COOH
H2N N H2


a-Difluoromethylomithine
(DMFO)


AdoMetDC Inhibitors


NH CH3

H2N N N2
H NH


NH2


H2N


H2N


MGBG






CGP-39'937




'2

N

MHZPA




NH2



N
AbeAdo


Figure 1-6: Structure of ODC inhibitor and AdoMet inhibitors.










Table 1-1: Different polyamine substrates of mammalian amine oxidases.


Polyamines and Flavin-dependent Proteins Copper-dependent Proteins
Acetylated Polyamines
MAO PAO DA( SAO
Putrescine +
Acetylputrescine +
Cadaverine +
Acetylcadaverine +
Spermidine + +
N1-Acetylspermidine + +
N8-Acetylspermidine + +
Spermine + +
N1-Acetylspermine + +


+ Represents the polyamine or acetylated polyamine is a substrate of the amine
oxidase.













CHAPTER 2
TRIAMINES AS ANTIPROLIFERATIVE AGENTS


Design Concept

In this study, a group of triamine analogues is designed as a new series of

anticancer agents. The rational is based on the following facts: (a) triamine analogues have

high antiproliferative activities; (b) triamine analogues have significantly lower host toxicity

than tetraamine analogues.


Antiproliferative Activity of Triamines

During the 1980s, several of spermidine analogues were investigated for their

antiproliferative activity by Porter et al. (1982, 1985). The two dialkylated spermidine

NI,N8-bis(ethyl)spermidine and NI,N8-bis(propyl)spermidine appeared be the most active

analogues. At 48 h, a 50% inhibition of growth (IC50) was reached at 40 and 50 pLM.

They also significantly reduced the polyamine pools and biosynthetic enzyme activities at

10 to 30 piM. Among different kinds of modification, terminal dialkylation was shown to

be the most effective one in the terms of increasing antiproliferative activity.

The spermidine homologues, norspermidine (NSPD), homospermidine (HSPD)

and 4,5-triamine (Figure 2-1) are found in some plants and bacteria, but not in mammalian

cells (Hamana et al. 1994, Kuttan et al. 1971, Hamana & Matsuzaki 1990, Fulihara et al.

1995). In vitro these unusual triamines are able to stimulate polypeptide synthesis of E.

coli at suboptimal Mg2+ concentrations in a manner comparable with that of spermidine

(Koumoto et al. 1990). In eukaryotic cells, Porter and Bergeron found that several

spermidine homologues (including NSPD and HSPD) could be transported into the cells

and






29
substrate SPD for cell proliferation, except NSPD because of inherent toxicity (1983).

Later on, Sunkara et al. investigated the antitumor activity of NSPD (1988). It is obvious

that these natural triamines, along with the synthetic triamine, 5,5-triamine (Figure 2-1),

could be good candidates for a source of the triamine backbones for modification. As in

the previous study, terminal dialkylation is used here as an effective derivatization of

triamines to provide potent antiproliferative agents with minimal toxicity. In order to

investigate the relationship between the size of the terminal groups and activity, alkylating

groups were designed to vary from methyl, ethyl to propyl.


Design of Polyamine Analogues with Low Toxicity


In order to design new polyamine analogues with lower toxicity, the historical

studies of polyamine toxicity, both in vitro and in vivo were reviewed. A new series of

polyamine analogues were designed and anticipated to have lower toxicity than other

polyamine analogues.

Previous study of polyamine toxicity in vitro

Although polyamines are essential for growth, at high concentrations, spermine and

spermidine are toxic to several species of bacteria, bacteriophage (Hirsch 1953, Kimes &

Morris 1971) and a number of different mammalian cell lines (Gaugas & Dewey 1978, Ali-

Osman & Maurer 1983, Parchement et al. 1990). The toxicity, often found with the

presence of ruminant sera in the culture medium, is believed to depend on the activity of

serum amine oxidases (SAO) which break the polyamines into highly active fragments and

cause the damage to the cell. Since ruminant is a component of many tissue culture media,

oxidative deamination of the polyamines by SAO was believed to be a frequent source of

erroneous interpretation of experiments with cultured cell.

As mentioned in the introduction, amine oxidase is a family of enzymes that include

monoamine oxidase, diamine oxidase histaminasee), polyamine oxidase and serum amine

oxidase. Amine oxidases have the generation of hydrogen peroxide and aldehyde in






30

common. Diamine oxidase and serum amine oxidase are Cu2+-dependent amine oxidases.

Monoamine oxidase and polyamine oxidase are FAD-dependent amine oxidase.

The oxidation of polyamines by serum amine oxidase is represented as in Figure 2-

2. These enzymes oxidize spermine and spermidine only at terminal-CH2NH2 groups with

the formation of the corresponding aldehydes plus NH3 and H202. The aldehydes are then

transformed into lower order amines by spontaneous |3-elimination of acrolein or turned

into amino acids by oxidation. The aminoaldehydes, peroxide and acrolein are postulated

to be responsible for the toxicity of extracellular polyamines in cell culture.

Toxicity mediated by aminoaldehyde. In 1964, Tabor et al. first reported the

identification of aminoaldehyde produced by oxidation of spermine and spermidine with

purified plasma amine oxidase (1964). Dioxidized spermine (tl/2 2.3 h in tissue culture

medium) led to a potent non-cytotoxic arrest of cell proliferation confined at the GI phase

of the cell cycle (Gaugas & Dewey 1978). The aminomonoaldehyde [NI-(4-aminobutyl)-

aminopropionaldehyde)] and the dialdehyde [N1, N4-bis-(3-propionaldehyde)-1,4-

diaminobutane] are found to interfere with DNA replication and cross-link DNA in vitro

(Bachrach 1973, Eilon & Bachrach 1969). By cross-linking a double-strand DNA, the

aminoaldehydes may prevent the strand separation necessary for transcription and

replication, in a manner similar to other bifunctional alkylating agents. Aminoaldehydes

may also exert their toxicity by interaction with proteins. Aldehydes are known to bind

reversibly to the amino group of amino acids, to the basic residues of proteins, and to

combine irreversibly with the sulphydryl groups of cysteine (Schauenstein 1978).

Toxicity mediated by peroxide. Henle et al. observed the random induction of

DNA strand breaks as a result of spermidine oxidation by serum amine oxidase in Chinese

hamster ovary cells (1986). Damage from SPD can be reduced when culture media are

supplemented thiourea (15 mM) or catalase (1000 units/mL) (the enzyme evolved in

removal of H202).






31

Toxicity of acrolein. Acrolein was found as a product of enzymatically oxidized

spermidine or spermine by Alarcon in 1970. At low levels (0.01-0.02 mM), acrolein was

able to inhibit nucleic acid and protein synthesis in E. coli. Alarcon suggested that acrolein,

the highly active aldehyde, was the oxidation product largely responsible for the

polyamine's cytotoxicity on mammalian cell based on its high cytotoxicity and high affinity

for SH groups (important in cell division and preferential inhibition of nucleic acids).

Acrolein is formed from spermidine and spermine but not putrescine (Alarcon 1972). The

fact that acrolein was formed from spermidine and spermine instead of putrescine could be

correlated with the observation that in rat cerebellar cultures, the neurotoxicity of

spermidine and spermine were much more potent than putrescine by about two orders of

magnitude (Gilad & Gilad 1987).

Early studies showed that replacement of calf serum by horse serum and addition of

aminoguanidine could abolish the toxicity of polyamine in cell culture (Gahl & Pitot 1978).

Aminoguanidine, as a copper chelator, is a highly effective inhibitor of serum amine

oxidase and diamine oxidase present in fetal calf serum. Ruminant sera, especially bovine

sera, contain high titers of serum amine oxidase (Bachrach 1970). Horse sera have lower

levels of amine oxidases. Rat serum is known to contain very low amine oxidase activity

(Seiler et al. 1980). Human sera may have no or very low levels of amine oxidase.

Although in many cases, the toxicity was understood to be due to the extracellular

metabolism of spermine by the amine oxidases found in bovine serum, some studies have

shown that the toxicity is not entirely dependent on extracellular oxidation, but may be due

in part of an intracellular event. Higgins et al. suggested that other factors were involved in

the toxicity by studying the growth of KB cells in different types of sera (1969). It was

found that the inhibition was decreased as calf and bovine sera concentration increased,

while an increase in horse serum concentration led to higher inhibition. Other studies show

that polyamines are also toxic in the absence of ruminant serum. Smith et al. found that in

a serum-free system, the oxidase activity on the cell surface reacted with polyamines lost








from the cell, and this was attributable to the growth inhibition of T-lyphocytes and

granulocytes (1983). The inhibition was reversible in the presence of polyamine oxidase

inhibitor 3-hydroxybenzyloxyamine. The cellular enzyme, the FAD-dependent polyamine

oxidase, is believed to be responsible for the intracellular oxidation of polyamines. It acts

on spermidine and spermine to produce 3-aminopropionaldehyde, H202 and putrescine,

although the preferred substrates of this enzyme are the acetylated polyamines (Morgan

1989).

Brunton et al. studied the dose-dependent inhibition of spermine in BHK-21/C13

cells grown in medium supplemented with horse serum (1989a, 1989b). At toxic levels

DNA synthesis was decreased and depletion of intracellular glutathione was observed. It

was postulated that the toxicity may result from the loss of intracellular GSH. But at this

point, it's not quite clear whether the loss of GSH is simply due to the presence of oxidants

or due to the block of GSH biosynthesis.

Polyamine analogues are less toxic than the parent polyamines in vitro. Porter et al.

found that in the cell culture with presence of fetal bovine serum, the terminal dialkylated

polyamines analogues did not show the potent nonspecific cytotoxicity as their parent

polyamines (1987b). Thus, amino alkyl substitution was suggested to be an effective

means of eliminating nonspecific host toxicity otherwise associated with enzymatic

oxidation of polyamines.

Previous study of polyamine toxicity in vivo

Compared to the in vitro studies, there are fewer descriptions of the toxicity of

spermidine and spermine and their mechanism on the intact animal. In 1956, the first

pharmacological study of polyamines in laboratory animals was reported (Tabor &

Rosenthal 1956). It was found that the acute toxicity of spermine and spermidine was

primarily related to renal tubular necrosis. Spermidine is approximately one-twentieth as

nephrotoxic as spermine. Less charged polyamines like putrescine and cadaverine are not

nephrotoxic. The order of toxicity is tetraamine > triamine > diamine. The same trend was






33
also found in the acute toxicity of polyamines in mice (norspermine > norspermidine > 1,3-

diaminopropane) (Bergeron et al. 1995b), which agreed with the result found in the cell

culture of resting lymphocytes (toxicity of spermine > spermidine > putrescine) (Nishida &

Miyamoto 1986).

The pharmacological properties of spermidine and spermine was further explore by

Shaw and the dose dependent toxicity was carefully studied in mice and rat (1972). The

acute toxicity cause by intravenous injection of spermine were ataxic, sedation, ptosis,

piloerection, hyperthermia, cardiorespiratory failure and lethargy. The toxic effect of

spermidine was similar to that of spermine but larger doses were required to produce equal

effects. From the acute LD5o value, spermidine was at least three times less toxic than

spermine.


Design Concepts


From all of the studies above, triamines consistently showed lower toxicity both in

vitro and in vivo. This phenomenon can be explained by the fact that under the effect of

polyamine regulatory system, excess diamines and triamines can be excreted from the cell

easily, while tetraamines are only poorly exported from cells unless acetylated and oxidized

to SPD under the successive actions of acetylase--SSAT and polyamines oxidase. It is

possible that longer half-life of spermine make it more toxic to the host than spermidine.

Compared with the natural polyamines, the terminal alkylated analogues have lower

acute toxicity. It was reported that DESPM was about 50% less toxic than spermine for

mice after a subcutaneous injection (Igarashi et al. 1990). The absence of primary amine in

the alkylated analogues may reduce the host toxicity associated with enzyme oxidation of

polyamines (Porter et al. 1987b).

The studies of chronic toxicity of polyamine analogues focused mainly on

tetraamine analogues. Among them, DEHSPM is the most toxic one with chronic ID50 of






34
37.5 mg/kg[chronic multiple-dose ip (t.i.d. x 6 days)] (Bergeron et al. 1994). Part of its

toxicity is due to its ability to initiate inhibition on gastrointestinal tract mobility.

It was found that the alkylated tetraamines were preferentially accumulated over the

native polyamines in L1210 cells (Bergeron et al. 1989). While significantly inhibited the

growth of cancer cells, the tetraamine analogues may have some disturbance on the regular

cell proliferation due to their prolonged existence in the host system.

Based on the previous studies of polyamine toxicity, triamines analogues are

expected to show lower toxicity than the tetraamine analogues. The triamine analogue with

the highest antiproliferative activity will be studied for its acute and chronic toxicity, along

with its effect on GI tract.


OSAR Study of Polyamine Analogues

Quantitative structure-activity relationship (QSAR) has been used in drug-design

studies since the 1960s (Martin 1978). Typically, a number of possible independent

variables, usually physiochemical parameters relating to a series of compounds, are

evaluated for correlation with activity values using multiple-regression analysis.

Previous studies have generated some relationship between the structural property

and the biological activity of polyamine analogues, such as terminal alkylation and the

required presence of positive charged nitrogens (Bergeron et al. 1994). Further studies are

needed to explain the dramatic activity difference caused by small variations in structures.

An obvious example is that although structurally DEHSPM, DESPM and DENSPM are

different only in one or two methylene bonds, they show great difference in bioactivity and

toxicity (Bergeron et al. 1989 and 1994). Based on a large number of polyamine analogs

synthesized in our group and some from other researchers, a library of diverse analogues is

established for this QSAR study. I believe that a computer-assisted QSAR study could

serve as a guide to the mechanism of the drug action and could assist in the design of new

analogues.






35
A comparative molecular field analysis (CoMFA) was considered as the approach to

define overall QSAR (Plummer 1990). The idea underlying a CoMFA is that differences in

a target property are often related to differences in the shapes of the non-covalent fields

surrounding the tested molecules (SYBYL 1995). To put the shape of a molecule field into

a QSAR table, the magnitudes of its steric (Lennard-Jones) and electrostatic fields

(Coulombic) are sampled at regular intervals throughout a region. While there are many

possible adjustable parameters in CoMFA, the most important is the relative alignment of

the individual molecules when their fields are computed.

Method of Data Analysis

Partial least squares (PLS), in CoMFA, like multiple regression (MR) used in

classical QSAR, is used to derive linear equations which describe the differences in the

values of targeted properties based upon the differences in the values of their corresponding

parameters (SYBYL 1995). As implemented in SYBYL/QSAR, PLS might be described

as a major extension of MR. In any MR analyzed QSAR, there exists a potential error

called chance correlation, a correlation which is only an accidental correspondence among

numbers which in fact are not related. PLS could omit this risk and has less restrictions on

the number of compounds and parameters compared to MR.

Methods of Analysis Validation

CoMFA uses several statistical tools to examine the accuracy and stability of the

derived equation. They are crossvalidation and bootstrapping. Crossvalidation examines

the accuracy of QSAR predictions. The crossvalidation technique involves random

elimination of one or more analogues from the original data set with subsequent equation

development and activity prediction for the eliminated analogues in an interactive manner.

This develops a QSAR equation that is generally of greater predictive value than that

derived from conventional regression analysis. The result of crossvalidation is reported in








the form of crossvalidated r2. Crossvalidated r2 is defined as the fraction of original
"variance" (squared differences among target values) predicted by QSAR. The r2 ranges

from 1 to 0, sometimes include negative numbers. When r2 = 1, a perfect model has been

found. If r2 = 0, then there is absolutely no relationship between structure and activity.

Negative r2 doesn't necessarily imply "no model" at all, but highlights that additional work

is needed. In this type of study, when r2 is greater than 0.4, useful and statistically

significant results are implied.

Another modem validation method used in CoMFA is bootstrapping.

Bootstrapping evaluates the stability of the equation derived by the regression analysis.

The name is derived from the adage for self-advancement occurring when you pull yourself

up by your own bootstraps. The idea is to simulate a statistical sampling procedure by

assuming that the original data set is the true population and generating many new data sets

from it. These new data sets (called bootstrap samplings) are of the same size as the

original data. Repeated selection of the same row is allowed. The difference between the

parameters calculated from the original data set and the average of the parameters calculated

from multiple bootstrap sampling is a measure of the bias of the original calculation. The

calculated variance of the parameter estimates reflects the accuracy with which any of the

parameters can be estimated from the input data.


Computer Modeling


In early studies, only terminal alkylated tetraamines were included in the database

(Bergeron et al. 1996b). A more accurate CoMFA can be obtained by expanding the

database with more structurally diverse compounds. So, primary polyamines along with

the relatively inactive polyamine derivatives reported before are included. In this study, a

total of 99 polyamine analogues, including the triamine analogues synthesized in this study

were used in the final database (Table 2-1). All of the polyamine analogues listed in Table






37
2-1 are numbered differently from the other chapters. In other chapters where compound

numbering are referred (Chapter 3, 4, 5 and 7), the same numbering system is applied.

The biological activity parameter used in this QSAR study is Ki. The Ki values are

the concentration of drugs required to inhibit the uptake of radioactive spermidine transport

by 50%. The Ki values of diamines (60-65), triamines (56-59) and triamine analogues

(39,40,42-52,54,55) and part of the tetraamine analogues (2-5, 7-11,14-37) are from

other studies (Porter and Bergeron 1983, Porter et al. 1985, Bergeron et al. 1994). While

the Ki values of norspermidine, spermidine, homospermidine and spermine had been

reported long time ago, the Ki values presented in this study were repeated as positive

control, and most of them agree well with the previous values.

On the basis of pKa values measured (Bergeron et al. 1995a) for DENSPM (4),

DESPM (11), DEHSPM (17), NI,NI2-bis(2,2,2-trifluoroethyl)spermine (FDESPM 26),

PIP(4,4,4) [N,N'-bis(4-piperidinylmethyl)-1,4-diaminobutane (32)], PYR(4,4,4) [N,N'-

bis(4-pyridylmethyl)-1,4-diaminobutane (33)] and PIP(5,4,5) [N,N'-bis {2-(4-

pyridylethyl)}-1,4-diaminobutane (35)] (Table 2-1), the following assumption about the

protonation state have been made. All of the nonaromatic, nontrifluoroethylated

tetraamines are largely in the from of tetracation at physiological pH 7.2. The spermine

analogues (9-13) and the homospermine analogues (15-21) and homologues (22-25)

should be 85 and 97%, respectively, in the form of tetracations. The norspermine

analogues (2-7) should at least 74% in the form of the tetracations. Compounds 26, 28,

30, 33 and 35 should be almost exclusively in the form of internal dication at

physiological pH. Compound 27 should also be a dication but because of the resonance it

should be a terminal dication. Compound 31 should be a dication with N1 and N3

charged. All of the unacetylated diamines and triamines are treated as dications and

trications, while the acetylated nitrogens are not charged. Thus the calculation of atomic

charges and conformational search of the analogues are carried out based on the anticipated

protonated structures. Two non-polyamines, N4-benzylspermidine nitrile 98 and MGBG






38
99 are also included in the database for their ability to compete for polyamine transport

systems (Porter et al. 1982).

The low energy conformations of polyamine analogues were obtained by systematic

conformation searching supported by Tripos Associates, Inc. in SYBYL 6.2 molecular

modeling program. The energies of steric and electrostatic interaction between each of the

analogues were calculated as parameters. A probe atom was placed at the various

interactions of regular three-dimensional lattice, which was large enough to include all of

the analogues in the database and with a 2.0 A lattice space. The probe atom had the shape

of sp3 carbon and a charge of + 1.0. The Van der Waals values taken from the standard

Tripos force field and the atomic charges were calculated by the method of Gastereiger and

Marsili (Gastereiger & Marsili 1980, SYBYL 1995). Wherever the probe atom experiences

a steric repulsion greater than "cutoff" (30 kcal/mol in these studies), the steric interaction is

set to the value "cutoff", and the electrostatic interaction was set to the mean of the other

molecular electrostatic interactions at the same location.

As mentioned before, the initial alignment of the molecules, the positioning of a

molecular model within the fixed lattices, is by far the most critical step in developing a

successful CoMFA model, since the relative interaction energies depend strongly on

relative molecular positions. In this study, polyamines were aligned by using DESPM

(11) as a template. In the tetraamines, both the end nitrogen atoms of the analogues and

the third nitrogen atom were used as atom pairs between two molecules for performing a

best fit. All of the triamines were aligned by using the three nitrogen atoms. Two nitrogen

and one carbon atoms were used for molecular match in diamines. The resulting structures

will be used to generate the CoMFA analysis.

The resulting database as described above with steric and electrostatic field values,

on a regularly spaced lattice around the analogues, were then correlated with the Ki data

for the polyamine transport system by a very efficient statistical method mentioned above,

PLS (Cramer 1988). Statistical techniques, bootstrapping and crossvalidation, are used to






39
determine the quality of the correlation. The implementation of PLS also rotates the PLS

solution back into the original data space, thus generating a "conventional" QSAR equation

showing r-square, F test, and the standard error S.


Results


The QSAR CoMFA studies by using a total of 99 polyamine analogues provided a

model with a crossvalidated r2 = 0.810 (optimum component 5) and the conventional

QSAR equation (R2 = 0.959, F test = 431.506, S = 0.199). The molecular modeling

shows that the relative contribution of the steric and electrostatic term in the QSAR equation

are 0.708 and 0.292, respectively, supporting the importance of the geometry of the groups

fixed to the nitrogen and of charge.

All of the polyamine analogues in the database, including diamines, triamines and

tetraamine analogues are listed in Table 2-1. The calculated Ki values of polyamine

analogues by the QSAR CoMFA equation, compared with the actual, are also listed (Table

2-1). The differences between the predicted and actual Ki values are generally small.

Figure 2-3 graphically demonstrates the significant correlation that exists between the

experimental versus the calculated Ki values.

This CoMFA model of 99 compounds was generated using different kinds of

structures of polyamine derivatives with linear, cyclic, and aromatic substituents. As the

Ki values used in the model are widely distributed, from 0.7 to 4500 pM, it is expected that

this CoMFA model would be very powerful in the evaluation of Ki values of new

structures.












H2N N NH2
H

Norspermidine (3,3-Triamine)




H
H2N N NH2



Homospermidine (4,4-Triamine)




H
H2N N NH2


4,5-Triamine


Figure 2-1: Structure of unusual triamines found in plants and bacteria.








H
H2N N N NH2
H
Spermine


Acrolein


0


VH202, NH3


H 0
H2 N N

Oxidized spermine





O
0
DixN id N sr0n
0
Dioxidized spermine







O ,


_H
H2N N N NH2

Spermidine





H
rOo', N NH2

Oxidized spermidine







X'.)O


H2N NH2 Putrescine




Figure 2-2: Oxidative deamination of spermidine and spermine by serum amine oxidase.
Ammonia, hydrogenperoxide, acrolein and aldehydes were generated.































-4 -3 -2 -1 0 1


Actual value







Figure 2-3: Measured versus calculated activity values for polyamine analogues.
The values presented here were expressed in the form of -logKi, directly used
in the molecular modeling. Each point represented a polyamine analogue,
and 99 polyamine analogues were included in this database.












Table 2-1: Comparison of actual and calculated Ki values of polyamine analogues.


Structure Ki (pM)

Abbreviate Actual Calculated

Norspermines
1 H2N N N NH2 NSPM 4.2 4.1
H H
2 N N N NN DMNSPM 5.6 3.0
H H H H

3 N N -' N-' NH, MENSPM 7.7 9.1
H H H
4 W'- N"-' N N"' DENSPM 17 17
H H H H
5 N -- N -N N DPNSPM 11 26
H H H H

6 JN^ N N'' \ DIPNSPM 40 39
H H H H


H H H



Spermine
H
8 H 2N/' N N NH2 SPM 0.7 0.9
H
H H
9 "N -N N -N -N DMSPM 1.1 1.0
H H
H
10 N'N_.'-N'- N NH2 MESPM 1.7 0.9
H H
H H
11 .N-/ --N, N'-, DESPM 1.6 2.3
H H
H H
12 NN -- NN DPSPM 2.3 2.7
H H
13 N N NN N DECroSPM 1.9 4.4
I H H














Table 2-1--Continuted


Structure


Abbreviate


Homospermines
H
14 H2N N NH2
H
H H
15 N NN N---" N,
H H
H
16 NH N- N N NH NH2
H H
H H
1N N N NW
H H

18 N

H H

H H
120 NHNN
H H
20 N---N-''- N >< N
OH H H
21 'N ,', N ', N N(" N,-
H H OH


HSPM


DMHSPM


MEHSPM


DEHSPM


DIPHSPM


ETBHSPM


DTBHSPM


(20H)DEHSPM(R,R)


Ki (pM)
Anhisl


Abbreviat


Homospermine Homologues
H
22 N-- / N N DE(3,4,4) 8 7
H H H
H H
23 N'/\N N N -. N DE(4,3,4) 4 6.5
H H
H H
24 -, vNN N NNN DE(4,5,4) 6.0 4.8
H H
25 HN 'N DE(5,4,5) 16 15
H H
H H
26 F NcF'- N-N_^ -__ N. C. FDES 285 245
H H










Table 2-1--Continuted

Structure Ki (pM)

Abbreviate Actual Calculated

Piperidine and Pyridine Derivatives

27 ,--'"N N PIP(3,3,3) 60.5 69.1
HN"' H H NH

28 'N N -'N PYP(3,3,3) 3940 3548
H H
HN
29 PIP(3,4,3) 24.6 31.6
NH
N

30 H PYP(3,4,3) 4480 5495
N H N

31 N N PYP(3,4,4) 608 661
H
HN H
32 N PIP(4,4,4) 4.9 4.4
HNH

33 N /N PYP(4,4,4) 3750 2399
H

34 PIP(5,4,5) 18.1 19.1

H
H0N









Table 2-1--continuted


Structure


Abbreviate


Cyclohexane Derivatives

36 HN- -
H H p

HH H


H H

H


39 H H
H H
HH




H
39 -





H H
H H


Norspermidine Analogues

42 H
H2N N NH2


43
H2N N NH2
44 H H H
Boc' N N N',Boc
45 H H H

46 H H
.N N N NH2
47 H H H

48 H H
49 N NH2
49 H H H


BAHSPM


CHX(4,4,4)-trans


CHX(3,4,3)-trans



CHX(4,4,4)-cis


CHX(3,4,3)-cis


CHX(3,3,3)-cis


NSPD


N4-Benzyl-NSPD

Bis(Boc)-NSPD

DMNSPD

MENSPD

DENSPD

MPNSPD

DPNSPD


2 1.5


3.5 2.7


7.9 7.2



1.5 1.6


1.8 1.6


30 35


7.2 7.2


135

1100

60

34

250

33

125


Ki (pM)

Actual


Calculated









Table 2-1--continuted


Structure


Abbreviate


Spermidine Analogues

50 H
H2N N NH2

51 INH
NH2

52 H2N NNH2


53

H2N N,_N NH2

54


NH2








H
55 H2N ./N,.NH2










58 H --N-NH2
0
0


59 O
H H
60 BOC'N N N BOC
H
HO OH 0

1 NHH



62 H NH2
63 H

NH2
64 H2N_ N^ NH2

65 H
H2N N NH2

66 H H H
HN. N NH2

67 H
2 NH2
68 H
N2N N NH2


SPD



N4-Methyl-SPD


N4-Propyl-SPD


N4-Hexyl-SPD



N4-Benzyl-SPD


N4-Acetyl-SPD


N4-Hexanoyl-SPD



N4-Benzoyl-SPD


Bis(acetyl)-SPD


Bis(propionyl)-SPD

Bis(BOC)-SPD



Bis(benzoyl)-SPD
DMSPD

MESPD(N1)

MESPD(N8)

DESPD

MPSPD(N1)

MPSPD(N8)

DPSPD


Ki (pM)

Actual


Calculated


>500


>500 1259


>500


1698


521 295


256


8.6

7

193

3.0

8.5

25.6


sm ,


2.2 2.0



2.6 2.6


3.1 4.5


43 32



39 47


115 135


151 257


F- I










Table 2-1--continuted


Structure


Abbreviate


Homosperidine Analogues

69 H
HN _-,N "-I NH2


70
H2N N NH2
H
71 BOC N N N-BOC
H H
72 H
H2N NH2
73 H
H2N -- N -N NH2
74 H
H2N NH2
75 H
H2N- N NH2


HSPD


3.4 3.6


N4-Benzyl-HSPD

Bis(BOC)-HSPD
DMHSPD

DEHSPD

MPHSPD

DPHSPD


Spermidine Homologues


76 H Triamine(3,
H2N _,-,_,N NH2
77 H
77 H2N'_,N NH2 Triamine(3,
78 H
H2N ^^.N /-NH2 Triamine(3,
79 H
7 H2N NNH2 Triamine(3.

80 H2N N NH2
H Triamine(4,
81 H
H2N N N NH2 Triamine(5,
H
82 ,N
H H DM(4,5)
H
83 NN "N N
H H DE(4,5)
H
84 -N N N
H H DP(4,5)
85 H
.NN N DM(5,5)
86 H
SDE(5,5)
87 H
**N /\, N DP(5,5)


5)


6)

7)

8)

5)

5)


21 24.6

64 63


Ki (pM)

Actual


Calculated











Table 2-1--continuted


Structure


Abbreviate


Ki (pM)

Actual


Calculated


Diamine Analogues

88 H2N 00NH2

89 H2N NH2
90 H2N 59NH22
91 H2N NH2
92 H2N 2 NH2
93 H2N NH2

H H
94
H
95 N N
H

96 E-1E13 6
H
97 H


DA3

DA4

DA5
DA6
DA7

DA8

Et-9-Et

Et-10-Et

Et- l-Et

Et-12-Et


Other Compounds



98 N4-Benzyl-SPD nitrile 163 219
NH H
99 H2N Ni NN J N NH2
H NH MGBG 53 87


>500
171

459

63
18

22

721

540

1230

232


72.4
68

229

79
27.5

16.6

339

692

692

347













CHAPTER 3
SYNTHESIS

Previous Studies on Polyamine Synthesis


The early work on polyamine analogue synthesis started with the modification of

linear methylene bridges between the nitrogens. A series of spermidine and spermine
homologues with structure of H2N(CH2)x(CH2)2NH2 and H2N(CH2)3NH-

(CH2)xNH(CH2)3NH2 (x = 2-6, 9, 10,12) were synthesized by mono or symmetrical

dicyanoethylation of the appropriate a,co-alkylene diamines, followed with catalytic

reduction of nitriles (Israel 1964).

The modification of different nitrogens of polyamines was once to be the major

tasks of chemists. Direct regioselective functionalization of nitrogens is difficult because

the reactivity difference between primary and secondary amines are quite small. So, the

synthetic strategy is to block the other nitrogens while leaving one free to be functionalized.

Functionalization with (tert-Butoxycarbonyl) Group

N2,N3-Diprotected spermidine was synthesized by masking N2 and N3 with tert-

butoxycarbonyl (BOC) group (Humora & Quick 1979). Monodiaminonitrile obtained by

the method of Israel was treated with 2-(tert-butoxycarbonyloxyimino)-2-phenylacetonitrile

(BOC-ON), a reagent previously utilized in peptide chemistry. The nitrile group was then

reduced to amine by LiA1H4 in a 70% yield. Then the N1 nitrogen of spermidine could be

functionalized and the BOC groups could be removed under acid condition, in this case,

hydrogen chloride saturated methanol was used. This method can be applied not only to

obtain a series of N1-functionalized triamines, but also to reach the terminal functionalized

spermine and its homologues.









Functionalization with Cycling Reagents

Another series of reagents offered functionalization at N1 and N8 of SPD by

masking the spermidine as either cyclic urea or hexahydropyrimidine (Figure 3-2, McManis

& Ganem 1980, Chantrapromma & Ganem 1981). The urea was produced in 95% yield

from spermidine by exhaustive acylation (CICO2CH3) and then hydrolyzed by Ba(OH)2.

One application of this intermediate is in a three-step synthesis of spermine from

spermidine. Following the alkylation of cyclourea at N8 (relative to spermine) with

acrylonitrile, reduction and hydrolysis, spermine was released in a yield over 50%. The

pyrimidine system could be functionalized and opened with ethyl hydrogen malonate and

piperidine or pyridine. When regioselective acylations and alkylations reagents are

available, N1 or N8-derivatives can be obtained.

The methodology was also extended to the symmetrical terminal modification of

spermines (Figure 3-2). Treatment of spermine with aqueous formalin solution gave a

crystalline bis-hexahydropyromidine in 95% yield. After acylation of the N1 and N12 with

the corresponding chloride, the gemdiamine heterocycles were removed by a Knoevenagel

reaction (ethyl hydrogen malonate, piperidine, ethanol, heat) with a yield above 80%.

Benzyl group as a protecting group

The secondary nitrogen-modified spermindine and its analogs were synthesized by

using benzylamine (1) as the starting material, which provides the central block--the

secondary nitrogen of the triamines (Figure 3-3) (Bergeron 1986). In proceeding to the

norspermidine reagent, benzylamine (1) can either be directly reacted with excess

acrylonitrile under pressure (step a) to produce the bis(nitrile) (2) or can be reacted at room

temperature with a mole of acrylonitrile (step d) to produce the mononitrile (3) which can

be reacted at room temperature with a second mole of acrylonitrile (step e) to generate (2).

The spermidine reagent (4) can be prepared by alkylating the mononitrile (3) with 4-






52
chlorobutyronitrile in butanol in the presence of sodium carbonate (step f). Synthesis of

the homospermidine reagent involves bisalkylation of benzylamine (1) with 4-

chlorobutyronitrile (step c) as before to produce nitrile (5). Each of the nitriles (2, 3, 4, 5)

gave a over 80% yield. Reduction of these nitriles with either lithium aluminum hydride or

with W-2 Raney nickel in ethanol in the presence of sodium hydroxide (step d) produced

the corresponding amines in a high yield about 70-90%. The Raney nickel procedure is

found to be more effective in this case.

The secondary-N-benzylated reagents were used in the synthesis of a large number

of terminally N-substituted triamines, both acylated and alkylated systems. i.e. N1, N8-

diethyl (or dipropyl) spermidine and NI, N8-diacetyl (or dipropionyl) spermidine. The

alkylated compounds were generated by reduction of the corresponding acyl compounds

with lithium aluminum hydride. The N4 (relative to sperimidine) benzyl group is removed

by hydrogenolysis over palladium chloride at atmospheric pressure.

The first threefold protected spermine, namely N3-benzyloxycarbonyl-N1-

phthaloyl-N2-tosylspermine was made in eight steps (Figure 2-4 (A), Eugster 1978)

Obtained from phthalamiding of 1,4-dibromobutane with potasium phthalimide (2) (step

b), N-(4-bromobutyl)phthalimide (3) was treated with NaN3 (step c) to give N-(4-

azidobutyl)phthalimide (4). Followed the reduction of (4) in alkohol/HCI with 10% Pd/C

(step d), the amine (5) was acylated with benzyl chloroformate (step e)to get diprotected

diamine (6). The phthalimide was removed by hydrazinolysis (step f), and the resulting

amine (7) was converted to diamide (8) with p-toluenesulfonychloride (step g). Alkylation

of (8) with N-(3-bromopropyl)phthalimide in DMF/K2CO3 (step h) generated the final
triprotected spermidine (9) with a yield of 38%. The protecting groups at N1, N2 and N3

can be removed selectively by hydazineolysis (H2NNH2, EtOH, reflux), catalytic

hydrogenolysis (Pt/C) and reductive treatment with sodium/liquid ammonia respectively.
The systemetic synthesis of norspermidine, spermidine and homospermidine
triprotected analogues was introduced in 1984 (Bergeron et al.). It generated tris-protected








spermidine in five high-yield steps [Figure 3-4 (B)]. The synthesis of the polyamine

reagents starts from the appropriate N-benzyl protected amine. Nitrile (1) is obtained in

high yield from benzylamine alkylatedwith acrylonitrile (step a), and reduced to diamine

(2a, n = 3) by Raney nickel (step c). With the presence of excess putrescine, the analog

diamine (2b, n = 4) was obtained by monobenzylation of putrescine with benzaldehyde

under reductive amination conditions (formic acid) in high yields (80%) (step b). The

diamines (2) are further protected by reacting them with 1 equiv. of 2-[[(tert-

butoxycarbonyl)oxy]-imino]-2-phenylacetonitrile (BOC-ON) at low temperature (0 *C)

(step c), give mono-acylation regioselectively at the primary amine site, resulting in 3a and

3b. Cyanoethylation of (3a) with acrylonitrile (step d) gave the nitrile (4a) in quantitative

yield. Alkylation of (3a) and (3b) with 4-chlorobutyronitrile (step e) gave the nitriles (4b)

and (4c), respectively, also in high yield (95%). The nitriles were then reduced with

Raney nickel (step f) and acylated with trifluoroacetic anhydride (step g) to give the desired

triprotected spermidine and its analogues. The conditions for the selective deprotections are

rather mild. The debenzylation of benzylspermidine required only mild hydrogenolysis

conditions, removal of the BOC groups only need brief exposure to trifluoroacetic acid

(TFA), and removal of the N-trifluoroacetyl group can be simply finished by refluxing the

corresponding amide with methanolic sodium carbonate.

The introduction of the p-tolulenesulfonyl protecting group facilitated the synthesis

of a series of terminal alkylated polyamine analogues (Bergeron et al. 1988). The NI,N12

terminally alkylated spermine compounds were synthesized by first mono-sulfonating all of

the spermine nitrogens with p-toluenesulfonyl chloride, leaving only the terminal nitrogens

to be alkylated. The resulting sulfonamides were next treated with sodium hydride in

dimethylformamide followed by excess alkylating agents, e.g. methyl iodide, ethyl iodide

or n-propyl iodide. The sulfonamides can be unmasked by treatment with sodium in liquid

ammonia for 12 h. One disadvantage of this reduction is that it produces various yields

(30-70%) and some insoluble byproducts. Later on, the mesitylenesulfonyl group was








reported to be a better protecting group than p-tolulenesulfonyl (Bergeron 1994). The

mesitylenesulfonyl groups could be cleanly removed under reductive conditions with 30%

HBr in HOAc/PhOH. The deprotection gave a predictable yield of 70-80%.

In order to meet the need of synthesizing a series of asymmetrical tetraamine

analogues, an asymmetrically protected diamine (1) was introduced (Figure 3-5). The

protecting groups, a mesitylenesulfonyl at each end and a BOC, can be selectively

removed. This triprotected diamine can be deprotonated on the monosubstituted end and

alkylated with fragmenting reagent (2). After the BOC group selectively removed with

trifluoroacetic acid (TFA), the terminal amide can be alkylated with another polyamine

fragmenting reagent (3). The resulting tetrasulfonamide (4) can be demasked with the

method of HBr/HOAc as mentioned above to provide the corresponding tetraamine

analogues.

Synthetic Methods

Synthesis of Tetraamines

Three different families of tetraamines were synthesized (1) those with linear

methylene backbones and symmetrical alkyl groups at each terminal nitrogen, (2) those

with linear methylene backbones but only one substitution at one end, (3) those with

symmetric but branch methylene backbones. The linear polyamines dimethyl-(1) and

monoethyl-(2) norspermine and dimethylspermine 7 were accessed by alkylation of the

corresponding tetrasulfonamide dianion (Figure 3-6, step b). However, linear symmetric

tetraamine diisopropylnorspermine 5 as well as the dimethyl-(12) and diisopropyl-(15)

homospermine were accessed via the "fragment synthesis" (Figure 3-6). In this case, the

central diamine segment as its N,N'-bis-sulfonamide 51-53 was alkylated at each end with
the appropriate fragmenting reagent N-(Go-halomethylene)-N-alkanesulfonamide (Figure 3-

6, step d). This less direct method was empolyeed for these compounds because isopropyl

bromide can not be used to alkylate tetrasulfoamide successfully, and homospermine along






55
with its homologue 1,6,12,17-tetraazaheptadecane (4,5,4-tetraamine) are not commercially

available. The unsymmetric linear tetraamine, monoethylnorspermine 2, was obtained by

monoalkylation of tetrasulfonylnorspermine dianion 49 (Figure 3-6, step b).

The synthesis of polyamine-segmenting reagents 65-71, Figure 3-7, began by

reacting a primary amine, even if hindered, with methylenesulfonyl chloride (step a). N-

Alkylsulfonamides 61-64 were deprotonated (NaH/DMF) followed by alkylation with the
appropriate dihalide in excess to generate synthons 65-67 (step b). Thus N-(co-

halomethylene)-N-alkylsulfonamides of any length can be prepared. These polyamine

synthons eliminate limitations associated with the availability of the starting tetraamines. In

addition, while earlier methods were limited to accessing polyamine analogues with

terminal primary alkyl groups, this approach allows for fixing primary, secondary, and

tertiary alkyl groups to the terminal amines of polyamines.

In this study, an entirely new synthetic method was introduced to access the

symmetric tetraamines with branching methylene backbones, diethylcrotyl-norspermine

(DECroNSPM 6) and -spermine (DECroSPM 11) (Figure 3-8). The tetraamine backbones

were obtained by fixing the protected branching segments 74 symmetrically to both ends
of 3 or 4 carbon bridges provided by a,o)-dibromides. The synthesis of this segment

began with crotononitrile (mixture of cis and trans), which reacted with excess amount of

ethylamine in 50% NaOH aqueous solution (step a). The obtained ethylamino nitrile 72

was reduced to 73 with Raney nickel in methanolic ammonia (step b), and masked with

mesitylenesulfonyl chloride to give disulfonamide 74 (step c). Alkylation of 2 equivalent

of 74 with I equivalent of the appropriate a,wt-dibromide provided the tetrasulfonamide

75 or 76 (step d), which was then deprotected by 30% HBr in HOAc with presence of

phenol to generate DECroNSPM (6) and DECroSPM (11), respectively (step e).

Synthesis of Triamines

Two different families of triamines were synthesized (1) those with symmetrical

methylene backbones, with one alkyl group at a terminal nitrogen or an alkyl group at






56
both nitrogens, (2) those with unsymmetrical methylene backbones, with a single alkyl

group at a terminal nitrogen or an alkyl group at both nitrogens. These include

norspermidines, spermidines, homospermidines, (4,5) and (5,5) triamines, the numbers

referring to the number of methylenes separating the nitrogens. In the case of the NaNw-

disubstituted norspermine and spermidine analogues the commercially available

triamines, norspermidine [NSPD (22)] and spermidine ([SPD (28)], were sulfonated by

mesitylenesulfonyl chloride (3 equiv.) under biphasic conditions (CH2Cl2/dilute NaOH)

(Figure 3-9, step i). Triprotected amines 85 and 86 were then converted to their

corresponding dianions (NaH/DMF) and alkylated with an excess of the appropriate

primary alkyl idodine (step f). Finally the mesitylenesulfonyl blocking groups were

cleanly removed under reductive conditions utilizing 30% HBr in HOAc and phenol to

give terminal dimethyl-(23) and dipropyl-(27) analogues of NSPD and

dimethylspermidine (29), which were isolated as their trihydrochloride salts (step g).

In order to prepare the symmetrically dialkylated derivatives of 4,4- (m = 2, n =
2) and 5,5-(m = 3, n = 3) triamines (Figure 3-9), mesitylenesulfonamide 77 was first

dialkylated with 2 equivalents of either 4-bromobutyronitrile or 5-bromovaleronitrile in

the presence of NaH to provide the dinitriles 79 and 81, respectively (step h). The

cyano groups were then hydrogenated with Raney nickel in methanolic ammonia (step d).

The resulting primary diamines 82 or 84 were converted to their corresponding

mesitylenesulfonamides 87 and 89 (step e). After alkylated with the appropriate

primary halide (step f) and the HBr-promoted deprotection (step g), the dimethyl-

(37,46) and dipropyl-(40,48) series of 4,4-, 5,5-triamines and diethyl 5,5-triamine

[DE(5,5) (47)] were obtained.

Diethyl homospermidine [DEHSPD (38)] was made following another synthetic
route (Figure 3-9). Alkylation of mesitylenesulfonamide 77 with two equivalents of N-
(4-bromobutyl)-N-ethylmesitylenesulfonamide 66 (step c) resulted the triprotected HSPD
analogue 95, which was then deprotected with HBr/HOAc to give DEHSPD 38 (step g).








The diethylatd unsymmetrical 4,5-triamine analogue, DE(4,5) (22), was

produced from N-(tert-butoxy)-N-mesitylenesulfonamide 102, a diprotected ammonia

synthon (Figure 3-10). This reagent was first alkylated with N-(5-bromopentyl)-N'-

ethylsulfonamide 66 (step b) to give triprotected diamine 103. The BOC group of 103

was then removed with TFA (step c), and the obtained disulfonamide 104 was alkylated

with N-(4-bromobutyl)-N'-ethylsulfonamide 67 (step d) to complete the triamine

framework. Unmasking of the amino groups of 105 with HBr led to the diethylated

analogue of the 4,5-triamine [DE(4,5) (43)] (step e).

The other two dialkylated 4,5-triamines, DM(4,5) (42) and DP(4,5) (44), were

produced by another route, Figure 3-9. Two consecutive monoalkylations of

mesitylenesulfonamide 77 with 4-bromobutyronitrile (step a) and 5-bromovaleronitrile

(step b) generated dinitrile 80. The cyano groups of 80 were reduced with Raney nickel

in methanolic ammonia, resulting in primary amine 83 (step d). Treatment of 83 with 2

equivalents of mesitylenesulfonyl chloride (step e), terminal dialkylation with

iodomethane or 1-iodopropane (step f), and then unmasking of the amino groups led to

DM(4,5) (42) and DP(4,5) (44) (step g), respectively.

Monopropyl norspermidine, MPNSPD (26), can be made directly by mono
alkylating trimesitylenesulfonyl NSPD 85 (step f), derived from commercially available

NSPD (22) (step i). However, synthesis of the SPD and HSPD monopropyl analogues,

involves a "fragment synthesis" (Figure 3-11), because the asymmetric backbone of SPD

makes the direct alkylation poorly selective and HSPD is not commercially available. The

bulky triphenylmethyl group can only be introduced selectively at one of the terminal amino
groups of either diaminopropane or diaminobutane (Figure 3-11). Triphenylmethyl

chloride was stirred at room temperature with either diamine in a 5-fold molar excess (step

a). The primary nitrogen of 106 or 107 was selectively sulfonated (step b) and alkylated

with the appropriate segmenting polyamine segment reagent 68 or 69 (step c). The
alkylation at this anion was highly selective. All of the protecting groups can be removed








with HBr in HOAc with presence of phenol (step e) resulting in NI-monopropyl

spermidine [MPSPD(NI), (33)], N8-monopropyl spermidine [MPSPD(N8), (34)], and

N -monopropyl homospermidine [MPHSPD, (39)].

Homospermidine and 4,5-triamine has been previously obtained by nonspecific
alkylation of diamines with 2 equivalents of O-bromoalkylphthalimide, followed by

hydrolysis with 12 M HC1 (Figure 3-12 (A), Okada et al. 1979). The obtained mixture

of diamine, triamine and tetraamine were purified on a ion-exchange column to give a

yield of 37% for HSPD. While the detailed synthesis of 4,5 triamine was not mentioned

and its yield was not specified in the study, it is assumed that 4,5 triamine was obtained

by an further alkylation of N-butylphthalimide diamine with 5-bromopentaphthalimide.

Clearly, direct alkyation of diamines dose not provide easy access to triamines.

A better synthetic route was introduced to get HSPD in the form of triacetate salt

at high yield (Figure 3-12 (B), Bergeron et al. 1981). The 5-benzyldinitrile was obtained

from dialkylation of benzylamine with 4-chlorobutanenitrile under vigorous condition-

115 C over 20 h. Treatment of dinitrile with platinum oxide in glacial acetic acid under a

hydrogen atmosphere gave HSPD triacetate in a yield over 85%.

In this study, an alternative synthetic route was chosen for its mild condition, easy

workup and high yield (Figure 3-13). In this method, use of the aromatic imide to protect

the primary amines avoided the solubility problems during attempted hydrogenation (Raney

nickel, methanolic NH3) of bis(3-cyanopropyl)mesitylenesulfonamide 82 (Figure 3-9).

The N,N'-di(butylamino)mesitylenesulfonamide 113 was synthesized by dialkylation of

anionized mesitylenesulfonamide 77 with 2 equivalents of 5-chlorobutylphthalimide in

DMF (Figure 3-13, step a), followed by hydrazinolysis (step b) to give 82. The final

treatment of sulfonamide 82 with 30% HBr in HOAc released HSPD 36 in good yield

(step c).

The synthesis of the parent 4,5- (41) and 5,5-(45) triamines involved

hydrogenation (Raney nickel, methanolic NH3) of the corresponding dinitriles without






59
solubility problems (Figure 3-9). Hydrogenation of the N-4-cyano-N'-5-cyano-

sulfonamide 80 (Figure 3-9, step d), followed by unmasking of sulfonyl group with
HBr/HOAc (step j) has produced 4,5-triamine (41). The 5,5-triamine (45) was

synthesized in a similar manner by reduction of the symmetrical dinitrile 81 (step d),
followed by treatment with 30% HBr in HOAc gave 5,5-triamine (45) in good yield

(Figure 3-9, step j).








H2N NH2

I CN


H
N C N^. NH2

BOC-ON

BOC
NCNN cNHBOC


LiAIH4


BOC
H2NN -NHBOC


(1) Modification at N1
(2) removal of BOC


R
R H
HNN NH2




Figure 3-1: Humora and Quick's modification of spermidine at N1 with
tert-butoxycarbonyl group (BOC) as the protecting group.









0


2N NH

H2N f


Cyclic Urea







Hexahydropyrimidine
(derived from SPD)







Hexahydropyrimidine
(derived from SPM)


Figure 3-2: Ganem's masked polyamines as cyclic urea or hexahydropyrimidines.













N-C N N CEN
(2)

\fe


b

H2N N NH2

(6)


b


N-C N C E-N


H2N N NH2
(6)


b

H2N NNH2


Figure 3-3: Modification of spermidine and its homologues at the secondary nitrogen.
(a) Acrylonitrile (excess); (b) H2/W-2 Raney nickel/NaOH/EtOH or LiAIH4;
(c) (f) 4-Chlorobutyronitrile/butanol/Na2CO3; (d) (e) Acrylonitrile (1 equiv.).


H2C NH2
(1)


NN C









CCH2NH(CH H2)2CN


(1) O 0
(b) 1,4-dibromobutane
0

c N-(CH2)4R (3)R=Br

| (c) NaN3
C6HHsC
(4) R = N3
(d) Acr
(d) Pd/C, H2

(5) R =NH2 4a

(e) CICO2CH2Ph


cIIV N-(CH2)4NHC02CH2Ph (6)

0

(f) (H2N)2

H"N-NHCO2C-1OCH Ph (7)

(g) p-Toluenesulfony
chloride
H
TOS -N NHCO2CH Ph (8)


(h) N-(3-Bromc
phthalimide


(a) Acrylonitrile

2an=3


0
C6HsCH

/(b) PUT
2b n=4


C6H5CH2NH(CH2)n-NH2


(c) BOC-ON


2NH(CH 2)n-NHC02C(CH3 ) (3a,b)


ylonitril

n=3,x=3


\(e) Cl-(CH2)3CN

4b n= 3, x=4
4c n =4, x =4


z (CH2)n-NHCOC(CH3)3
C6HsN
(CH2)x-I-CN


(f) H2/Ra Ni

0
(CH2)n-NHCOC(CH 3)
C (CH2)x-NH2


propyll)


0
I N ,NNNHCO2CH2Ph
0 TOS (9)
A


(5a,b,c)
(g) I Trifluoro
acetic anhydride
0
II
/ (CH2)n-NHCOC(CHa)3
(CH2)x-NHCCFa
0
(6a,b,c)
B


Figure 3-4: Synthesis of triprotected triamines.
(A) Eugster's eight-step synthesis of triprotected SPD. TOS = tosyl.
(B) Bergeron's five-step synthesis of triprotected NSPD (n = 3, x = 3);
SPD (n = 3, x = 4) and HSPD (n = 4, x = 4).










a
MesSO2NHBOC


SO2Mes

CN CN


b SO2Mes
C NI
B0C 'N NH2


SO2Mes

BOC" N NHSO2Mes


SO2Mes
d I
R (CH2)nX
SOgMes

BO CN (CH2)n ,R
I I
SO2Mes SO2Mes


SO2Mes
I
R N(CH2)mX


SO2Mes SO2Mes
I I
R',N (CH2)m N (CH2)n ..R
SO2Mes SO2Mes

H H
R' "N "(CH2 ^-^m (C2)n \ NR


Synthesis of triprotected putrescine reagent and its application in synthesis
of asymmetric spermine homologues. X = Cl or Br.
(a) 4-Chlorobutyronitrile/NaH/DMF; (b) H2/Ra Ni/NH3/CH3OH;
(c) Mesitylenesulfonyl chloride/NaOH(aq)/CH2CI2; (d) NaHIDMF/2;
(e) TFA/CH2C12; (f) NaH/DMF/3; (g) HBr in HOAc (30%)/PhOH, HCI.


c
---- k


Figure 3-5:








H H
H2N N ,,N,(},NH2

m =1, n= 1, NSPM
m =1, n = 2, SPM

I a

SO22MesSOgMes SO2Mes SO2Mes
H N N NH
m'r n\r m/r


49m=1,n=1
50 m=1, n=2


H2N nNH2


n = 1, 1,3-diaminopropane
n = 2, 1,4-diaminobutane
n = 3, 1, 5-diaminopentane

cI

SO2Mes S02Mes
HN{fI. NH


51 n=1
52n=2
53n=3


Sb
d
S02Mes SO2Mes SO2Mes SO2Mes
R'N'4 N nN N'R'


'X


54 m =1, n = 1, R = R'= Me
55 m =1, n = 1, R = Et, R'= H
56 m =1, n = 1, R = R' = i-Pr
57 m =1, n =2, R= R'= Me
58 m = 2, n =2, R =R'= Me
59 m = 2, n= 2, R = R'= i-Pr
60 m = 2, n =3, R = R' = Et


SO2Mes
I- NR


70 m = 1, R = i-Pr, X = CI
65 m = 2, R = Me, X = Br
66 m =2, R = Et, X = Br
71 m =2, R = i-Pr, X = Br


I e

R*- ir- f ,1.


1 m =1, n = 1, R = R' = Me, DMNSPM
2 m =1, n = 1, R = Et, R' = H, MENSPM
5 m =1, n = 1, R = R'= i-Pr, DIPNSPM
7 m =1, n = 2, R = R' = Me, DMSPM
12 m = 2, n = 2, R = R' = Me, DMHSPM
15 m = 2, n = 2, R = R' = i-Pr, DIPHSPM
20 m = 2, n = 3, R = R' = Et, DE(4,5,4)

Figure 3-6: Synthesis of linear tetraamine analogues.
(a) (c) Mesitylenesulfonyl chloride/I N NaOH(aq)/CH2Cl2;
(b) NaH/DMF/iodomethane or iodoethane; (d) NaH/DMF/co-haloalkyl-
sulfonamide (65, 66, 70 or 71); (e) HBr in HOAc(30% )/PhOH/CH2CI2, then
HCI.




















a b
RNH2 RNHSO2Mes --



61 R = Me

62 R = Et

63 R = i-Pr

64 R = n-Pr


SO2Mes




65 R = Me, m= 2, X = Br

66 R = Et, m = 2, X = Br

67 R = Et, m = 3, X = Br

68 R = n-Pr, m =1, X = Br

69 R = n-Pr, m = 2, X = Br

70 R = i-Pr, m = 1, X = CI

71 R= i-Pr, m=2, X= Br


Figure 3-7: Synthesis of polyamine segmenting reagents.
(a) Mesitylenesulfonyl chloride/NaOH(aq)/CH2Cl2;
(b) NaH/DMF/a~,-dihaloalkane.











a CH3 CN

NHEt


R
I
b,c EtN NHR

CH3


73R=H
74 R = SO2Mes


Br. Br
(CH2) /


SO2Mes SO2Mes SO2Mes SO2Mes

EtNy N'( CH2 n Et
CH3 CH3

75 n = 3
76 n =4


H H
EtNH,,,N, N(NCH2 N. NHEt 4HCI
^T ^^ CH2 n "


CH3


CH3


6 n=3
lln=4


Figure 3-8: Synthesis of DECroNSPM and DECroSPM.
(a) EtNH2/50% NaOH/CH2CI2; (b) H2/Ra Ni/NH3/CH3OH;
(c) Mesitylenesulfonyl chloride/NaOH(aq)/CH2C12; (d) NaH/DMF/
a,to-dibromoalkane; (e) 30% HBr in HOAc/PhOH/CH2CI2, then HC1.


CH3 CN

















-
Io I



c
0.


Z



EE


0

0
co-Z




0-Z


Sc I




E E










x Z
z





















\ 2 V








1 6
Z-
0c

cm 1


II II II II II


II II II II II
EEEEE
isCD ;ag CO


EEE
II II


















c c
II I 11 f
EE
G 0C0 a


a)


0-


cl



co


0




U)-Z








cr-Z




cc
-2


M -Z








M-z


(-Z


Z

0
o


11 f 1 11 1 11 11 ,I i
II II II II II II II II II II
cc r cccc cr ccccc a:



i- l C si C C C CO c6
II II II II II II i II II Ii II
EEEEEEEEEEEE







a.






,-_ w "
ccd- e


SII 11 IfI iI 11 11 11 11
CccCCccCcecC
II II i | II II II II 1g II II |I

EEEEEEEEEEEEc
m cm CM m CO v "O t
awwwww qww










a
--- SO a
/ SO2NH2

77


BOC.N H


102Mes
icI"


SO2Mes c SO2Mes
N s H ----- BN BOC
SO2Mes SO2Mes
ind 103


B r N
SO2Mes


SO2Mes

I s
SO2Mes SO2Mes


H
\N N -N 3 HCI
I I
H H
43



Figure 3-10: Synthesis of DE(4,5).
(a) Oxalyl chloride/tert-butyl alcohol; (b) NaH/DMF/66; (c) TFA/CH2C12;
(d) NaH/DMF/67; (e) 30% HBr/HOAc/PhOH, then HC1.


SO2Mes










R
H-N NNHCPh,

106 n = 1, R = H
107 n =2, R =H
108 n = 1, R = SO2Mes
109 n = 2, R = SO2Mes
c I-


SO2Mes
NN /--mBr

68 m =1
69 m=2


SO2Mes SO2Mes H
S m N'NCPh3


110m=1,n=2
111m=2,n =1
112 m=2, n =2


1d


HN H ,2
m n
N mN nNH2


S3 HCI


33 m = 1, n =2, MPSPD(N1)
34 m =2, n = 1, MPSPD(N8)
39 m = 2, n = 2, MPHSPD


Figure 3-11: Synthesis of monopropyl SPD and HSPD analogues.
(a) Ph3CCl/CH2C12; (b) Mesitylenesulfonyl chloride/1 N NaOH(aq)/
CH2Cl2; (c) NaH/DMF/68 or 69; (d) 30% HBr in HOAc/PhOH/CH2CI2,
then HC1.


a,b


H2N nNH2










H2N NH2


H2 N '


2 equi.


I

I 0


Hydrazinolysis


Mixtrue of HSPD and HSPM


HSPD 3HCI

(A)




85% PtO2/H2/AcOH
HSPD triacetate


H2N ,NN NH2

(Obtained as in Figure 3-3)


Figure 3-12: Synthesis of homospermidine by previous methods. (A) Okada's synthesis
of HSPD. (B) Bergeron's synthesis of HSPD.










-/- SO2NH2


a


113

b

MesSO2N NH2


H2N NNNH2 3 HC
H
36 Homospermidine




Figure 3-13: Synthesis of homospermidine in this study.
(a) N-(4-Bromobutyl)-phthalimide/NaH/DMF; (b) (H2N)2, H20/EtOH;
(c) 30% HBr in HOAc/PhOH/CH2CI2, then HC1.









Table 3-1: Polyamine analogues synthesized in this study.

Number Structure


Abbreviation


Tetraamines
1

2


5


6


7


11


12


17

20


N N N N
H H H H

N N'-" N.-' N, '-' NH2
H H H


H H H H

-'N J` N -^-` N N N
H H H H
H H

H H
H H
H H H
H H

H H



H H

H H


DMNSPM

MENSPM


DIPNSPM


DECroNSPM


DMSPM


DECroSPM


DMHSPM


DIPHSPM

DE(4,5,4)


H H H


N N N NH2
H H
N N NH2N
H H H
H
H H
SN N NH2
H H
H
H2N N HN -
H
H
NH N '- NH2
H
-N Ns N ,' N
H H


DMNSPD

MENSPD

MPNSPD

DPNSPD

DMSPD

MPSPD(N1)

MPSPD(N8)

HSPD

DMHSPD


Triamines

23

24

26

27

29

33

34

36

37











Table 3-1--continuted


Triamines

38

39

40

41

42

43

44

45

46

47


48


DEHSPD

MPHSPD

DPHSPD

4,5-Triamine

DM(4,5)

DE(4,5)

DP(4,5)

5,5-Triamine

DM(5,5)

DE(5,5)


DP(5,5)


The numbers of the compounds are as same as the ones assigned in the synthetic
methods and in the assays of biological activity.


H
H H
H
HN H
H

H
H H


H
H H
H
H H
H

H
H2N N NH2
H H H

H H H
H H H













CHAPTER 4
MATERIALS AND METHODS

Synthetic Methods

Chemical reagents were purchased from Aldrich Chemical Co. (Milwaukee, WI).

Reactions using hydride reagents were run in distilled DMF under a nitrogen atmosphere.

Fisher optima grade solvents were routinely used, and organic extracts were dried with

sodium sulfate. Silica gel 60 (70-230 mesh) obtained from EM Science (Darmstadt,

Germany) or silica gel 32-63 (40 mM "flash") from Selecto, Inc. (Kennesaw, GA) was

used for column chromatography. Melting points were determined on a Fisher-Johns

melting point apparatus and are uncorrected. Proton NMR spectra were run at 90 or 300

MHz in CDCI3 (not indicated) or D20 with chemical shifts given in parts per million

downfield from tetramethylsilane or 3-(trimethylsilyl)propionic-2,2,3,3-d4 acid, sodium

salt, respectively. Coupling constants (J) are in hertz. FAB mass spectra were run in a

glycerol/trifluoroacetic acid matrix. Elemental analyses were performed by Atlantic

Microlabs, Norcross, GA.

Synthesis of Tetraamine Analogues

N-Isopropylmesitylenesulfonamide (63). A solution of mesitylenesulfonyl

chloride (36.5 g, 0.167 mol) in CH2Cl2 (200 mL) was added dropwise to isopropylamine

(9.87 g, 0.167 mol) in 1 N NaOH (200 mL) at 0 C. After the mixture was stirred at room

temperature overnight, the bilayer was separated and the aqueous layer was extracted

with CH2Cl2 (3 x 100 mL). The combined organic fraction was then washed with H20

(100 mL), saturated NaCI solution (100 mL) and dried over anhydrous Na2SO4 for 12 h.

After the solvent was removed, the crude product was recrystallized from aqueous EtOH
to give 33.5 g (83%) of 63 as colorless prisms: mp 97-98 C; 1H NMR 8 1.00 (d, 6 H, J=

75






76
7), 2.20 (s, 3 H), 2.55 (s, 6 H), 3.16-3.55 (m, 1 H), 4.35 (d, I H, J = 7), 6.85 (s, 2 H).

Anal. Calcd. for C12H19N02S: C, 59.72; H, 7.94; N, 5.80. Found: C, 59.66; H, 7.97; N,

5.81.
N-(4-Bromobutyl)-N-methylmesitylenesulfonamide (65). NaH (80%, 3.73 g,

0.124 mol) was added to a solution of 61 (Schreinemakers 1897) (20.4 g, 95.7 mmol) in

DMF (200 mL) at 0 -C. The mixture was stirred at room temperature for 30 min, and
1,4-dibromobutane (137 mL, 1.15 mol) was added at 0 *C. After stirred at 70 C

overnight, the reaction was quenched by adding dist. water (10 mL). The solvents were
removed under high vacuum and the obtained mixture was taken up in CHC13 (80 mL)

and H20 (80 mL), followed by extraction of aqueous phase with chloroform (3 x 60 mL).
After washed with brine (100 mL), the combined organic portion was then dried over

anhydrous sodium carbonate. The solvent was removed by rotvap. Purification by

column chromatography with 7:1 hexane/EtOAc produced 9.55 g (30%) of 65 as an oil:
1H NMR 8 1.54-1.85 (m, 4 H), 2.25 (s, 3 H), 2.56 (s, 6 H), 2.68 (s, 3 H), 3.00-3.35 (m, 4

H), 6.88 (s, 2 H). Anal. Calcd. for CI4H22BrNO2S: C, 48.28; H, 6.37; N, 4.02. Found:
C, 48.50; H, 6.45; N, 3.99.

N-(4-Bromobutyl)-N-ethylmesitylenesulfonamide (66). Sodium hydride (80%,

3.71 g, 0.124 mol), 62 (Schreinemakers 1897) (21.6 g, 95.0 mmol) in DMF (100 mL),
and 1,4-dibromobutane (140 mL, 1.17 mol) in DMF were combined, and the reaction was
worked up by the method of 65. Column chromatography (6:1 hexane/EtOAc) resulted
in 33.2 g (96%) of 66 as an oil: IH NMR 8 1.04 (t, 3 H, J= 7), 1.4-1.8 (m, 4 H), 2.25 (s, 3

H), 2.55 (s, 6 H), 3.03-3.38 (m, 6 H), 6.87 (s, 2 H). Anal. Calcd. for C s5H24BrNO2S: C,
49.73; H, 6.68; N, 3.87. Found: C, 49.78; H, 6.72; N, 3.88.
N-(3-Chloropropyl)-N-isopropylmesitylenesulfonamide (70). NaH (80%, 1.7

g, 57 mmol), 63 (9.46 g, 32.9 mmol) in DMF (120 mL), and 1,3-dichloropropane (38 mL,
were reacted and worked up following the method of 65. Column chromatography (12:1
hexane/EtOAc) produced 7.48 g (60%) of 70 as an oil: 1H NMR 8 1.10 (d, 6 H, J = 7),






77
1.72-2.02 (m, 2 H), 2.22 (s, 3 H), 2.52 (s, 1 H), 3.80 (septet, 1 H, J = 7), 6.80 (s, 2 H).

Anal. Calcd. for C15H24CIN02S: C, 56.68; H, 7.61; N, 4.41. Found: C, 56.74; H, 7.59;
N, 4.46.
N-(4-Bromobutyl)-N-isopropylmesitylenesulfonamide (71). NaH (80%, 1.54

g, 51.2 mmol), 63 (9.5 g, 39 mmol) in DMF (120 mL), and 1,4-dibromobutane (56.5 mL,
0.473 mol) were mixture and worked up following the procedure of 65. Column
chromatography (10:1 hexane/EtOAc) furnished 11.34 g (77%) of 71 as an oil: 1H NMR
8 1.16 (d, 6 H, J= 7), 1.50-1.85 (m, 4 H), 2.27 (s, 3 H), 2.50 (s, 6 H), 3.00-3.40 (m, 4 H),

3.75-4.05 (septet, 1 H), 6.90 (s, 2 H). Anal. Calcd. for C16H26BrNO2S: C, 51.06; N,
6.96; N, 3.72. Found: C, 51.17; H, 7.01; N, 3.66.
N1,N4,N8,Nll-Tetrakis(mesitylenesulfonyl)norspermine (49).
Mesitylenesulfonyl chloride (24.2 g, 0.110 mol) and N, N'-bis(3-aminopropyl)-1,3-
propanediamine (NSPM, 5.2 g, 28 mmol) in CH2C12 (150 mL) and 1 N NaOH (150 mL)
were combined and worked up by the method of 63. Recrystallization from 50%
EtOAc/hexane afforded 20.1 g (80%) of 49 as a crystalline solid: mp 135-136 *C; 1H
NMR8 1.45-1.80 (m, 6 H), 2.25 (s, 12 H), 2.50 and 2.55 (2 s, 24 H), 2.63-3.27 (m, 12 H),

4.97 (t, 2 H, J = 7), 6.90 (s, 8 H). Anal. Calcd. for C45H64N408S4: C, 58.93; H, 7.03; N,
6.11. Found: C, 58.84; H, 7.05; N, 6.16.
NlN4,N8,N11-Tetrakis(mesitylenesulfonyl)-NlN 11-dimethylnorspermine
(54). Sodium hydride (80%, 2.5 g, 0.083 mol), 49 (24.3 g, 0.026 mol) in DMF (200 mL)
and iodomethane (8.1 g, 57 mmol) were combined and worked up by the method of 65.
Flash chromatography (5:1 hexane/EtOAc) produced 15.0 g (60%) of 54 as a white solid:
1H NMR 8 1.55-1.81 (m, 6 H), 2.25 (s, 12 H), 2.50 and 2.56 (2 s, 30 H), 2.85-3.15 (m, 12

H), 6.87 (s, 8 H). Anal. Calcd. for C47H68N408S4: C, 59.72;H, 7.25; N, 5.93. Found: C,
59.78; H, 7.26; N, 5.90.
N1,N11-Dimethylnorspermine Tetrahydrochloride (1). HBr (30%) in HOAc

(300 mL) was added slowly to a solution of 54 (14.8 g, 15.7 mmol) and phenol (58.9 g,






78
0.626 mol) in CH2C12 (150 mL) at 0 TC. After the reaction mixture was stirred for 1 day
at room temperature, H20 (200 mL) was added followed by extraction with CH2CI2 (3 x

200 mL). The aqueous portion was evaporated under high vacuum, and the residue was
taken up in 1 N NaOH (10 mL) and 19 N NaOH (10 mL) followed by extraction with
CHC13 (14 x 50 mL). After removal of CHCl3, the residue was taken up in EtOH (100

mL) and acidified with concentrated HCI (10 mL). After the solvents were removed, the
solid was recrystallized from aqueous EtOH to give 4.84 g (85%) of 1 as crystals: IH
NMR (D20) 8 1.80-2.35 (m, 6 H), 2.70 (s, 6 H), 3.05-3.30 (m, 12 H). Anal. Calcd. for

CllH32Cl4N4: C, 36.48; H, 8.90; N, 15.47. Found: C, 36.20; H, 8.84; N, 15.19.
Nl,N4,N8,Nll-Tetrakis(mesitylenesulfonyl)-Nl-ethyl norspermine (55).

Sodium hydride (80%, 0.20 g, 6.5 mmol), 49 (6.0 g, 6.5 mmol) in DMF (50 mL) and

iodoethane (0.52 mL, 6.5 mmol) were combined and worked up by the method of 65.
Flash chromatography (5:3:2 hexane/CH2Cl2/EtOAc) produced 2.17 g (30%) of 55 as a
white foam: IH NMR 8 0.85 (t, 3 H, J = 7), 1.50-1.70 (m, 6 H), 2.23 (s, 12 H), 2.47-2.50

(2 s, 24 H), 2.80-3.20 (m, 14 H), 4.80-5.00 (br s, 1 H), 6.85 (s, 8 H). Anal. Calcd. for

C47H68N4O0S4-H20: C, 58.60; H, 7.32; N, 5.82. Found: C, 58.61; H, 7.25; N, 5.57.
N1-Ethylnorspermine Tetrahydrochloride (2). HBr (30%) in HOAc (50 mL),
55 (0.86 g, 0.91 mmol) and phenol (4.7 g, 50 mmol) in CH2C12 (100 mL) were mixed at
0 C. The reaction mixture was stirred overnight at 74 C and worked up following the
procedure of 1. Recrystallization from aqueous EtOH generated 60.5 mg of 2 (18%) as
crystals: 1H NMR (D20) 8 1.25 (t, 3 H, J = 7), 1.90-2.30 (m, 6 H), 2.95-3.30 (m, 14 H).

Anal. Calcd. for CllH32Cl4N4: C, 36.48; H, 8.90; N, 15.47. Found: C, 36.55; H, 8.90; N,
15.39.
N,N'-Bis(mesitylenesulfonyl)-1,3-propanediamine (51). Mesitylenesulfonyl
chloride (28.4 g, 0.130 mol) in CH2C12 (150 mL) and 1,3-diaminopropane (4.84 g, 65
mmol) in 0.5 N NaOH (300 mL) were combined and the biphasic mixture was stirred at
room temperature overnight. The obtained white solid was filtered and washed with 1 N






79
HCI (3 x 25 mL) and H20 (3 x 50 mL). Recrystallization from EtOAc furnished 26.1 g
(91%) of 51 as crystals: mp 214-215 C; 1H NMR 8 2.20 (s, 6 H), 2.45 (s, 12 H), 2.60

(quartet, 4 H), 3.25 (s, 2 H), 6.93 (s, 4 H). Anal. Calcd. for C21H30N204S2: C, 57.51; H,
6.89; N, 6.39. Found: C, 57.43; H, 6.92; N, 6.30.
NlN4,N8,Nll-Tetrakis(mesitylenesulfonyl)-N1,Nll-diisopropylnorspermine
(56). Sodium hydride (80%, 0.74 g, 23.6 mmol), 51 (3.11 g, 7.14 mmol) in DMF (40
mL) and 70 (5 g, 15.7 mmol) in DMF (40 mL) were combined and worked up following
the procedure of 65. Column chromatography (12:1 toluene/EtOAc) produced 5.3 g
(74%) of 56 as an oil: 1H NMR 8 1.05 (d, 12 H, J = 7), 1.50-1.70 (m, 6 H), 2.25 (s, 12

H), 2.50 (s, 24 H), 2.80-3.05 (m, 12 H), 3.75 (septet, 2 H, J = 7), 6.90 (s, 8 H). Anal.
Calcd. for C5lH76N4O0S4: C, 61.17; H, 7.65; N, 5.60. Found: C, 61.26; H, 7.67; N, 5.54.
N1,Nll-Diisopropylnorspermine Tetrahydrochloride (5). A solution of 56 (5.2

g, 5.19 mmol) and phenol (18.4 g, 0.196 mol) in CH2C12 (80 mL) was combined with
30% HBr in HOAc (100 mL) and worked up following the procedure of 1. Recrystalli-
zation from aqueous EtOH gave 1.5 g (69%) of 5 as crystals: 1H NMR (D20) 8 1.30 (t,

12 H, J = 7), 1.90-2.30 (m, 6 H), 3.05-3.60 (m, 14 H). Anal. calcd. for C15H40Cl4N4: C,
43.07; H, 9.64; N, 13.39. Found: C, 42.92; H, 9.65; N, 13.25.
N,N'-Bis(mesitylenesulfonyl)-1,4-butanediamine (52). Mesitylenesulfonyl
chloride (54.4 g, 0.249 mol) in CH2C12 (300 mL) was added to 1,4-diaminobutane (11.34
g, 0.129 mol) in 1 N NaOH (300 mL), and the mixture was stirred for 1 day at room
temperature. Organic solvent was evaporated, and 2.4 N HCl (250 mL) was added. Solid
was filtered, washed with water (250 mL), and recrystallized from aqueous EtOH to give
50.46 g (90%) of 52 as needles: mp 156.5-157.5 C; 1H NMR 8 1.36-1.60 (m, 4 H), 2.27

(s, 6 H), 2.57 (s, 12 H), 2.69-2.96 (m, 4 H), 4.65 (t, 2 H, J= 6), 6.89 (s, 4 H). Anal.
calcd. for C22H32N204S2: C, 58.38; H, 7.13; N, 6.19. Found: C, 58.31; H, 7.19; N, 6.14.
N1,NlN10,N14-Tetrakis(mesitylenesulfonyl)-N1,N14-dimethylhomospermine
(58). Sodium hydride (80%, 1.19 g, 39.7 mmol), 52 (5.60 g, 12.4 mmol), and 65 (9.55 g,






80
27.4 mmol) in DMF (100 mL) were combined and worked up following the procedure of
65. Purification by flash chromatography (5:1 toluene/EtOAc) gave 8.61 g (70%) of 58
as a white solid: IH NMR 8 1.25-1.50 (m, 12 H), 2.25 (s, 12 H), 2.53 and 2.59 (2 s, 30

H), 2.90-3.15 (m, 12 H), 6.90 (s, 8 H). Anal. Calcd. for C50H74N408S4: C, 60.82; H,
7.55; N, 5.67. Found: C, 60.92; H, 7.56; N, 5.66.
N1,N14-Dimethylhomospermine Tetrahydrochloride (12). A solution of

58 (8.58 g, 8.61 mmol) and phenol (32.6 g, 0.347 mol) in CH2CI2 (70 mL) was
mixed with HBr (30%) in HOAc (170 mL) and worked up by the method of 1.
Recrystallization from aqueous EtOH generated 2.54 g (72%) of 12 as white
crystalline plates: 1H NMR (D20) 8 1.62-1.90 (m, 12 H), 2.68 (s, 6 H), 2.95-3.20 (m,

12 H). Anal. Calcd. for C 14H38Cl4N4: C, 41.59; H, 9.47; N, 13.86. Found: C, 41.66;

H, 9.51; N, 13.81.
N,N5,Nl1,N4-Tetrakis(mesitylenesulfonyl)-N1,Nl4-diisopropyl-
homospermine (59). Sodium hydride (80%, 1.21 g, 40.3 mmol), 52 (5.70 g, 12.6
mmol) in DMF (50 mL), and 71 (11.34 g, 30.1 mmol) in DMF (75 mL)
werecombined and worked up by the method of 65. Column chromatography (10:1
toluene/EtOAc) gave 7.61 g (58%) of 59 as a white solid: IH NMR 8 1.10-1.45 (m,

24 H), 2.25 and 2.35 (2 s, 12 H), 2.55 (s, 24 H), 3.65-3.95 (m, 2 H), 6.85 (s, 8 H).
Anal. Calcd. for C54H82N408S4: C, 62.16; H, 7.92; N, 5.37. Found: C, 62.38; H,
7.97; N, 5.22.
N1,N14-Diisopropylhomospermine Tetrahydrochloride (15). A solution of

59 (7.57 g, 7.25 mmol) and phenol (27.2 g, 0.289 mol) in CH2C12 (100 mL) was
combined with 30% HBr/HOAc (140 mL) and worked up by the method of 1,
recrystallization from aqueous EtOH gave 2.52 g (75%) of 15 as crystals: 1H NMR
(D20) 8 1.25 (d, 12 H, J = 7), 1.70-1.90 (m, 12 H), 2.95-3.15 (m, 12 H), 3.40 (septet,

2 H, J= 7). Anal. Calcd. for C18H46C14N4: C, 46.96; H, 10.07; N, 12.17. Found: C,
46.69; H, 10.14; N, 12.10.






81
N,N'-Bis(mesitylenesulfonyl)-1,5-pentanediamine (53). Mesitylenesulfonyl

chloride (12.6 g, 57.6 mmol) in CH2CI2 (150 mL) and 1,5-diaminopentane
dihydrochloride (5.05 g, 28.8 mmol) in 1 N NaOH (150 mL) were combined and
worked up by the method of 52. Recrystallization from aqueous EtOH gave 12.21 g
(91%) of 53 as crystals: mp 117-120 'C; 1H NMR 8 1.18-1.49 (m, 6 H), 2.25 (s, 6 H),

2.58-2.97 (m + s, 16 H), 4.66 (t, 2 H, J = 6), 6.87 (s, 4 H). Anal. Calcd. for

C23H34N204S2: C, 59.20; H, 7.34; N, 6.00. Found: C, 59.13; H, 7.35; N, 6.00.
3,8,14,19-Tetrakis(mesitylenesulfonyl)-3,8,14,19-tetraazaheneicosane
(60). NaH (80%, 0.35 g, 12 mmol), 53 (2.29 g, 4.90 mmol) and Nal (63 mg, 0.42
mmol) in DMF (90 mL) was combined with 66 (Bergeron et al. 1994) (5.90 g, 16.3
mmol) in DMF (10 mL) and worked up by the method of 65. Column
chromatography (1.5% CH3OH/CHCl3) gave 4.40 g (87%) of 60: 1H NMR 8 0.97 (t,

6 H, J= 7), 1.2-1.5 (m, 14 H), 2.25 (s, 12 H), 2.52 (s, 24 H), 2.8-3.2 (m, 16 H), 6.87
(s, 8 H). Anal. Calcd. for C53H80N408S4: C, 61.84; H, 7.83; N, 5.44. Found: C,
61.55; H, 7.81; N, 5.30.
3,8,14,19-Tetraazaheneicosane Tetrahydrochloride (20). A solution of 60
(4.29 g, 4.17 mmol) and phenol (14.8 g, 0.157 mol) in CH2Cl2 (60 mL) and 30%
HBr/HOAc (80 mL) were combined and worked up by the method of 1.
Recrystallization from aqueous EtOH gave 1.31 g (70%) of 20 as plates: 1H NMR
(D20) 5 1.25 (t, 6 H, J = 7), 1.55-1.85 (m, 14 H), 2.90-3.20 (m, 16 H). Anal. Calcd.

for C17H44Cl4N4: C, 45.74; H, 9.94; N, 12.55. Found: C, 45.80; H, 10.00; N, 12.46.
3-Methyl-4-azahexyl nitrile (72). To a 0 C mixture of ethylamine
hydrochloride (50 g, 0.613 mol) and NaOH (aqueous 50%w/w, 44 mL, 0.613 mol)
solution, crotononitrile(cis & trans) (28.5 mL, 0.35 mol) was added dropwise. After
stirred at room temperature for 1 h and 60 *C overnight, the reaction was worked up by
adding ethyl ether (100 mL) and 1 N NaOH (50 mL) to the mixture to make the aqueous
layer pH >13. Then the water layer was separated and extracted with Et20 (3 x 100 mL).








The organic fractions were combined, dried with anhydrous Na2SO4 and purified by

short path distillation. 25.40 g (64.6%) 72 was obtained as clear liquid.

4-Methyl-1,5-diazaheptane Dihydrochlodide (73). W-2 grade Raney nickel

(18.0 g) and concentrated NH40H (60 mL) were successively added to a solution of 72

(25.40 g, 0.226 mol) in CH3OH (200 mL) in a 500 mL Paee bottle, and a slow stream of
NH3 was bubbled through the mixture for 50 min at 0 'C. After hydrogenation was

carried out at 50-60 psi for 18 h, the suspension was filtered through Celite and the

solvents were removed in vacuo. The residue was taken up in abs. ethanol and acidified

to pH =1 with conc. HC1. 39.25 g (92%) 73 was generated as very hydroscopic white
solid: 1H NMR (D20) 8 1.28-1.38 (m, 6 H), 1.88-2.01 (m, 1 H), 2.14-2.25 (m, 1 H), 3.10-

3.18 (m, 4 H), 3.40-3.46 (m, 1 H). Anal. Calcd. for C6H18Cl2N2: C, 38.10; H, 9.59; N,

14.81. Found: C, 38.14; H, 9.63; N, 14.78.

4-Methyl-1,5-bis(mesitylenesulfonyl)-1,5-diazaheptane (74). To a solution of

73 (5 g, 26.4 mmol) in 1 N NaOH (120 mL), 11.6 g (0.0528 mol) 2-mesitylenesulfonyl

chloride in CH2C12 (100 mL) was added dropwise at 0 C in 45 min. After stirred at

room temperature overnight, worked out following the procedure of 63 and purified by

recrystallization from EtOAc. 10.26 g (81%) 74 was obtained as shinning plates: m.p
134-136 C. 1NMR 6 0.90-1.10 (m, 6 H), 1.45-1.70 (m, 2H), 2.20 (s, 6 H), 2.50-2.55 (2s,

12 H), 2.80-3.30 (m, 4 H), 3.70-3.90 (m, 1 H), 5.05 (br t, 1 H), 6.85 (s, 4 H). Anal.
Calcd. for C24H36N204S2: C, 59.97; H, 7.55; N, 5.83. Found: C, 59.93; H, 7.57; N, 5.86.

4,14-Dimethyl-3,7,11,15-tetrakis(mesitylenesulfonyl)-3,7,11,15-

tetraazaheptadecane (75). Sodium hydride (80% in oil, 0.9 g, 30 mmol), 74 (10.0 g, 21

mmol) in dry DMF (100 mL).and 1,3-dibromopropane (1 mL, 9.85 mmol) in DMF (10
mL) were combined and worked up by the method of 65. Column chromatography (3:1
Hexane/EtOAc) produced 5.5 g (56%) 75 as a white foam. IH NMR 8 0.85-1.05 (m, 12

H), 1.45-1.70 (m, 6H), 2.25 (s, 12 H), 2.50-2.55 (2 s, 18 H), 2.85-3.20 (m, 12 H), 3.35-






83
3.60 (m, 2 H), 6.85 (s, 8 H). Anal. Calcd. for C51H76N408S4: C, 61.17; H, 7.65; N, 5.60.

Found: C, 61.27; H, 7.65; N, 5.54.
4,14-Dimethyl-3,7,11,15-tetraazaheptadecane Tetrahydrochloride

(DECroNSPM, 6). HBr (30%) in HOAc (100 mL), 75 (5.50 g, 5.5 mmol) and phenol

(19.8 g, 0.21 mol) in CH2Cl2 (90 mL) were combined and worked up following the

procedure of 1. Recrystallized from aqueous EtOH to give 1.59 g (70%) of 6 as crystals:
1H NMR (D20) 8 1.15-1.35 (m, 12 H), 1.60-2.30 (m, 6 H), 2.95-3.50 (m, 14 H). Anal.

Calcd. for C 15H40N4Cl4: C, 43.07; H, 9.64; N, 13.39. Found: C, 42.97; H, 9.65; N,

13.32.
4,15-Dimethyl-3,7,12,16-tetrakis(mesitylenesulfonyl)-3,7,12,16-
tetraazaoctadecane (76). Sodium hydride (60% in oil, 0.48 g, 12 mmol), 74 (4.81 g, 10
mmol) in dry DMF (80 mL), and 1,4-dibromobutane (1.06 g, 4.92 mmol) in DMF (10
mL) were combined and worked up by the method of 65. Column chromatography (3:1
Toluene/EtOAc) produced 4.0 g (80%) 76 as a white foam: 1H NMR 8 0.85-1.70 (m, 20

H), 2.25 (s, 12 H), 2.50 (s, 18 H), 2.85-3.60 (m, 14 H), 6.85 (s, 8 H). Anal. Calcd. for

C52H78N408S4: C, 61.51; H, 7.74; N, 5.52. Found: C, 61.35; H, 7.70; N, 5.44.
4,15-Dimethyl-3,7,12,16-tetraazaoctadecane Tetrahydrochloride

(DECroSPM, 11). HBr (30%) in HOAc (75 mL), 76 (3.98 g, 3.92 mmol) and phenol

(13.9 g, 0.15 mol) in CH2C12 (75 mL) were worked up following the procedure of 1.
Recrystallized from aqueous EtOH to give 1.23 g (73%) of 11 as crystals: 1H NMR
(D20) 8 1.15-1.45 (m, 12 H), 1.65-2.30 (m, 8 H), 2.95-3.50 (m, 14 H). Anal. Calcd. for

C16H42N4CI4: C, 44.45; H, 9.79; N, 12.96. Found: C, 44.37; H, 9.72; N, 12.98.

Synthesis of Triamines

NlN4,N7-Tris(mesitylenesulfonyl)norspermidine (85). A solution of
norspermidine (2.62 g, 0.02 mol) in 1 N NaOH (80 mL) and 14.0 g (0.064 mol) 2-
mesitylenesulfonyl chloride in CH2Cl2 (50 mL) were combined and worked up by the







method of 63. Recrystallization from EtOAc-hexane gave 85 (11.63 g, 86%) as
crystalline solid : 1HNMR 8 1.55 (t, 4 H, J= 7), 2.25 (s, 9 H), 2.50-2.55 (2 s, 18 H),

2.70-3.20 (m, 8 H), 4.20 (br t, 2 H), 6.87 (s, 1 H). Anal. Calcd for C33H47N306S3: C,

58.47; H, 6.99; N, 6.20. Found: C, 58.36; H, 6.95; N, 6.18.
NlN4,N7-Tris(mesitylenesulfonyl)-Nl1,7-dimethylnorspermidine (90). NaH

(60% in oil, 0.44 g, 11 mmol), 85 (3.39 g, 5 mmol) in DMF (70 mL), and iodomethane

(1.63 g, 11.5 mmol) were combined and worked up using the method of 65, and purified

by flash column chromatography with toluene/EtOAc (8:1) as eluant. 2.47 g (70%) 90
was obtained as a thick oil: 1H NMR 8 1.73 (quintet, 4 H), 2.29 (s, 6 H), 2.54-2.55 (2 s,

18 H), 2.60 (s, 6 H), 2.99-3.06 (2 t, 8 H, J = 7), 6.94 (s, 6 H). Anal. Calcd for

C35H51N306S3: C, 59.55; H, 7.28; N, 5.95. Found: C, 59.50; H, 7.33; N, 5.88.
N1,N7-Dimethylnorspermidine Trihydrochloride (23). A solution of 90 (2.13
g, 3.02 mmol) and phenol (12.27 g) in CH2Cl2 and HBr (30% in HOAc, 30 mL) were

combined and worked up by the method of 1. Recrystallization from aqueous EtOH
afforded 0.298 g (37%) of 23 as plates: 1H NMR (D20) 8 2.08-2.19 (m, 4 H), 2.75 (s, 6

H), 3.13-3.22 (m, 8 H). Anal. Calcd for C8H24N3C13: C, 35.77; H, 9.00; N, 15.64.

Found: C, 35.91; H, 8.96; N, 15.69.

NlN4,N7-Tris(mesitylenesulfonyl)-Nl-monopropylnorspermidine (91). A

solution of 85 (2.57 g, 3.8 mmol) in DMF, NaH (60% in oil, 0.52 g), and 1-iodopropane

were combined and worked up with the method of 65. Flash chromatography
(hexane/EtOAc 3:1) generated 1.16 g (23%) of 91 as an oil: 1H NMR 8 0.66 (t, 3 H, J=

7), 1.23-1.31 (m, 2 H), 1.58-1.62 (m, 4 H), 2.262-2.271 (2 s, 9 H), 2.507-2.519 (2 s, 12

H), 2.59 (s, 6 H), 2.82-2.99 (m, 8 H), 3.21 (t, 2 H, J = 7), 4.85 (br t, 1 H), 6.89-6.92 (m,
6 H). Anal. Calcd for C36H53N306S3: C, 60.05; H, 7.42; N, 5.84. Found: C, 59.79; H,
7.32; N, 5.70.

N1-Monopropylnorspermidine Tihydrochloride (26). HBr (30% in HOAc, 30
mL), 91 (1.14 g, 1.58 mmol) and phenol (6.4 g) in CH2C12 were reacted and worked up


F--






85
following the procedure of 1. Recrystallization from aqueous EtOH gave 26 (87 mg,
20%) as crystals: 1H NMR (D20) 8 0.98 (t, 3 H, J= 7.5), 1.67-1.75 (m, 2 H), 2.07-2.16

(m, 4 H), 3.01- 3.21 (m, 10 H). Anal. Calcd for C9H26C13N3: C, 38.24; H, 9.27; N,
14.87. Found: C, 38.15; H, 9.32; N, 14.75.
Nl,N4,NT-Tris(mesitylenesulfonyl)-N1,N7-dipropylnorspermidine (92). NaH
(60%, 0.44 g, 11 mmol), 85 (3.39 g, 5 mmol) and 1-iodopropane (1.95 g, 11.5 mmol) in

DMF (70 mL) were combined and worked up following the procedure of 65. Column
chromatography (3:1 hexane/EtOAc) gave 3.54 g (93%) of 92 as an thick oil: 1H NMR 8

0.7 (s, 6 H), 1.20-1.65 (m, 8 H), 2.25 (s, 9 H), 2.50 (s, 18 H), 2.80-3.05 (m, 12 H), 6.87
(s, 6 H). Anal. Calcd for C39H59N306S3: C, 61.55; H, 7.68; N, 5.52. Found: C, 61.52;
H, 7.79; N, 5.55.
NI,N7-Dipropylnorspermidine Trihydrochloride (27). HBr in HOAc (30%, 80
mL), 91 (3.475 g, 4.57 mmol) and phenol (15.8 g) in CH2Cl2 (30 mL) were reacted and
worked up following the procedure of 1. Recrystallization from aqueous EtOH gave 1.26
g (85%) of 27 as plates: 1H NMR (D20) 8 0.87 (t, 6 H, J = 7), 1.60 (m, 4 H), 2.01 (m, 4

H), 2.93 (t, 4 H, J = 7), 3.06 (m, 8 H). Anal. Calcd for CI2H32Cl3N3: C, 44.38; H, 9.93;
N, 12.94. Found: C, 44.42; H, 9.89; N, 12.88.
N1,N4,A8-Tri(mesitylenesulfonyl)spermidine (86). 2-Mesitylenesulfonyl

chloride (6.87 g, 31.4 mmol) in CH2Cl2 (30 mL) was combined with spermidine
trihydrochloride (28) and worked up with the method of 63. Flash column
chromatography (4:3 hexane/EtOAc) gave 3.73 g (55%) 86 as a white foam: 1H NMR 8

1.30 (m, 2 H), 1.44 (m, 2 H), 1.66 (m, 2 H), 2.30 (s, 9 H), 2.46 (s, 6 H), 2.60 (s, 12 H),
2.76 (quartet, 2 H), 2.84 (quartet, 2 H), 3.04 (t, 2 H, J = 7), 3.24 (t, 2 H, J = 7), 4.56 (br t,
1 H), 4.92 (br t, 1 H), 6.90 (s, 2 H), 6.95 (s, 4 H). Anal. Calcd for C34H49N306S3: C,
59.02; H, 7.14; N, 6.07. Found: C, 58.74; H, 7.12; N, 5.99.
NN4,AN8-Tri(mesitylenesulfonyl)-N1,N8-dimethylspermidine (93). NaH
(60%, 0.41 g, 10.3 mmol), 86 (2.15 g, 3.11 mmol), and iodomethane (1.41 g, 9.95 mmol,






86
620 mL) in DMF (60 mL) were combined and worked up by the method of 65. Flash
chromatography (5:3 hexane/EtOAC) gave 2.24 g 93 (100%) as a thick oil: 1H NMR 8

1.39-1.43 (m, 4 H), 1.69-1.78 (m, 2 H), 2.28 (s, 3 H), 2.30 (s, 6 H), 2.55 (s, 12 H), 2.57
(s, 6 H), 2.60 (s, 3 H), 2.62 (s, 3 H), 2.96-3.13 (m, 8 H).
Nl,NS-Dimethylspermidine Trihydrochloride (29). HBr (30%) in HOAc (60

mL), 93 (2.24 g, 3.11 mmol) and phenol (12.3 g) in CH2Cl2 (30 mL) were combined and
worked up by the procedure of 1. Recrystallization from aqueous EtOH provided 0.658 g
(75%) 29 as crystals: 1H NMR (D20) 8 1.77-1.82 (m, 4 H), 2.06-2.18 (m, 2 H), 2.73 (s, 3
H), 2.75 (s, 3 H), 3.06-3.19 (m, 8 H). Anal. Calcd. for C9H26C13N3: C, 38.24; H, 9.27;

N, 14.86. Found: C, 38.19; H, 9.28; N, 14.79.
N-Bis(3-cyanopropyl)mesitylenesulfonamide (79). Sodium hydride (60%, 2.6

g, 66 mmol), 77 (6 g, 30 mmol) in DMF (100 mL) and 3-bromobutyronitrile (9.77 g, 66
mmol) were combined and worked up by the method of 65, column chromatography ( 1:1
hexane/EtOAc) produced 7.48 g (75%) of 79 as an oil: 1H NMR 8 1.87 (m, 4 H), 2.27 (t,

4 H, J= 7), 2.31 (s, 3 H), 2.60 (s, 6 H), 3.34 (t, 4 H, J = 7), 6.99 (s, 2 H). Anal. Calcd for
C17H23N302S: C, 61.24; H, 6.95; N, 12.60. Found: C, 61.08; H, 6.96; N, 12.55.
NI,N5,N9-Tris(mesitylenesulfonyl)homospermidine (87). 2-Mesitylene-
sulfonyl chloride (6.71 g, 30.7 mmol) in CH2C12 (30 mL) was reacted with a solution of
82 in 1 N NaOH (35 mL) and worked up by the method of 63. Column
chromatography (4:1 toluene/EtOAc) produced 3.06 g (31%) 87 as a white foam: IH
NMR 8 1.32-1.38 (m, 4 H), 1.44-1.54 (m, 4 H), 2.28-2.29 (2 s, 9 H), 2.54 (s, 6 H), 2.60

(s, 12 H), 2.79 (quartet, 4 H), 3.09 (t, 4 H, J = 7), 4.70-4.80 (br s, 2 H), 6.90 (s, 2 H),
6.92 (s, 4 H). Anal. Calcd. for C35H51N306S3: C, 59.55; H, 7.28; N, 5.95. Found: C,
59.34; H, 7.29; N, 5.92.
N1,N5,N9-Tris(mesitylenesulfonyl)-Nl,N9-dimethylhomospermidine (94).
Sodium hydride (60%, 0.17 g, 4.17 mmol), 87 (1.28 g, 1.8 mmol) in DMF (50 mL), and
iodomethane (0.565 g, 0.25 mL, 4.0 mmol) were combined and worked up by the